Nvidia, the chipmaker, revealed its latest financial statements on Wednesday, with revenue reaching $30.04 billion in the last three months. This is a significant increase of 122% compared to the previous year, indicating sustained growth in their artificial intelligence investments.
Despite analysts’ projections of $28.7 billion in sales, the company’s shares dropped more than 3% in after-hours trading.
Nvidia’s founder and CEO, Jensen Huang, announced plans to ship a greater number of chips and hardware next year than in the company’s 31-year history during an earnings call.
Huang highlighted the importance of fast development due to the increasing complexity of their models. He stated that the company aims to lower costs while scaling AI models to unprecedented levels for the next industrial revolution.
Analysts, while optimistic about the results, acknowledged signs that Nvidia’s exceptional revenue growth might be slowing down. Major tech companies’ aggressive AI investments are driving demand for Nvidia chips, but these companies are also investing in their own silicon development.
The company informed customers about a delay in the launch of their next-generation AI chip, known as Blackwell. Early samples have already been sent to a limited number of customers. Despite this, the current graphics processing unit, Hopper, continues to sell well according to CEO Jensen Huang.
Nvidia reported record revenue with a 154% increase in data center revenue year over year, amounting to $26.3 billion, reflecting the demand for accelerated computing and generative AI in data centers globally.
Nvidia’s earnings results hold great significance on Wall Street, as the company accounts for 6% of the total value of the S&P 500 and is the third-largest company globally with a market capitalization of $3.1 trillion.
Recent reports from major tech customers such as Microsoft, Amazon, Meta, and Google, show increased capital spending as they utilize Nvidia chips to develop and train their AI models.
The company’s earnings per share were $0.68, and they announced a $50 billion share repurchase. Profit is expected to rise to $15.1 billion, up from approximately $6.2 billion in the same period last year.
Ives, a Wedbush analyst, emphasized the importance of Nvidia’s earnings report on the stock market, estimating that every dollar spent on Nvidia’s GPU chips contributes $8 to $10 to profits across the tech sector.
The market’s focus on Nvidia’s performance stems from the belief that AI advancements will boost global productivity for years to come.
Comparisons to the Internet bubble of the late 1990s have emerged, with concerns that the AI boom might peak if Nvidia’s results disappoint investors.
Regulators are closely monitoring Nvidia, following an antitrust investigation launched by the Department of Justice after allegations from rival chipmakers. The investigation claims Nvidia is using its market power to monopolize markets and compel customers to continue buying its products.
THoused inside a glass box in the chapel on the campus of Rensselaer Polytechnic Institute in Troy, New York, the “Quantum Chandelier” is the symbolic centerpiece of an ambitious effort to transform upstate New York into a tech hub — something like Silicon Valley for social media or Cambridge, Massachusetts, for biotechnology.
The silvery sci-fi object, named for the internal gold lattice that mounts, cools and isolates the processors, will be the heart of a “quantum computing system” that will herald a new era of computing. It’s the heart of Curtis Prime’s dream, co-founder of Nvidia, a $2.8 trillion artificial intelligence hardware and software company, to transform Rensselaer (RPI) into an advanced computing hub, remaking this part of upstate New York into a new Silicon Valley.
Priem has invested a significant amount of his wealth into building the Curtis Priem Quantum Constellation, a workshop where RPI students can envision the future of quantum computing. Just as his partners at Nvidia, where he served as the company’s first chief technology officer, allowed him to freely imagine the graphics chip architecture that will power the AI revolution, he hopes his investment will spark a new era of computational innovation in the region.
Prime believes the area along the Hudson Valley, from Yorktown Heights, home to IBM’s Quantum Research Institute, to Troy, home to the RPI/SUNY nanotech complex, to Syracuse, where Micron is building a massive $100 billion fab complex, will be the future home of U.S. computer technology.
To that end, he’s thinking beyond concerns about artificial intelligence and the success of Nvidia’s H100 graphics processing unit (GPU), which powers 90% of generative AI systems.
There are two RPI students on campus. Photo: Gregory Sherin
Wall Street has become skeptical of technology. AI has caused billions of dollars of losses, and Wall Street is disheartened by the idea that new technology is going to change the world. But the same thing happened with the internet overbuild of the 1990s, which went through booms and busts before eventually paying off.
In theory, quantum technology could solve in seconds problems that take today’s supercomputers decades to solve, unlocking secrets about the behavior of molecules, the genetic code, weather forecasting and, of more recent concern, cracking the encryption systems that underpin the internet.
Astronomers Murchison Widefield Alley Researchers in Western Australia conducted a search for extraterrestrial signals emanating from around 2,800 galaxies pointing towards the Vela supernova remnant with a spectral resolution of 10 kHz.
This diagram shows what a Kardashev Type III civilization might operate like. Containing stellar energy in so-called Dyson spheres is one way to harness the enormous energy on a galactic scale. The resulting waste heat products should be detectable with telescopes. Image by Danielle Futselaar / ASTRON.
“When we think about the search for extraterrestrial intelligence, we often consider the age and advancement of technology that could produce signals that we could detect with telescopes,” said Dr Chenoa Tremblay from the SETI Institute and Professor Steven Tingay from Curtin University.
“In popular culture, advanced civilizations are depicted as having interstellar spacecraft and the means to communicate.”
“In the 1960s, astrophysicist Nikolai Kardashev proposed a scale for quantifying the degree of technological advancement of extraterrestrial intelligence.”
“The Kardashev scale has three levels. A Type I civilization uses all the energy available on its planet (1016 W); Type II civilizations can consume stellar energy directly (1026 W) and a Type III civilization could consume all the energy emitted by the galaxy (1036 “W)”
“Civilizations at the higher end of the Kardashev scale could generate vast amounts of electromagnetic radiation detectable at galactic distances.”
“Some of the ideas that have been explored in the past have been to harness the light of stars in our galaxy, to colonize the solar system, and to use pulsars as a communications network.”
“Radio waves' ability to penetrate space over long distances and even planetary atmospheres makes them a practical tool for searching for interstellar communications.”
The authors used the Murchison Widefield Array (MWA), focusing on low radio frequencies (100 MHz), to look for signs of alien technology in galaxies beyond the Milky Way.
They observed about 2,800 galaxies in one observation, and determined the distances to 1,300 of them.
“This research represents a major step forward in efforts to detect signals from advanced extraterrestrial civilizations,” Dr Tremblay said.
“The MWA's wide field of view and low-frequency range make it an ideal tool for this type of study, and the limits we set will guide future research.”
CD Tremblay & SJ Tingay. 2024. An extragalactic wide-field search for technosignatures with the Murchison Wide Field Array. ApJ 972, 76;doi:10.3847/1538-4357/ad6b11
Condé Nast and OpenAI have announced a long-term partnership to feature content from Condé Nast’s brands such as Vogue, Wired, and The New Yorker in OpenAI’s ChatGPT and SearchGPT prototypes.
The financial details of the agreement were not disclosed. OpenAI, backed by Microsoft and led by Sam Altman, has recently signed similar deals with Axel Springer, Time magazine’s owner, Financial Times, Business Insider, Le Monde in France, and Prisa Media in Spain. This partnership allows OpenAI to access extensive text archives owned by publishers for training large language models like ChatGPT and real-time information retrieval.
OpenAI launched SearchGPT, an AI-powered search engine in July, venturing into Google’s long-dominant territory. Collaborations with magazine publishers enable SearchGPT to display information and references from Condé Nast articles in search results.
OpenAI’s Chief Operating Officer, Brad Lightcap, expressed the company’s dedication to collaborating with Condé Nast and other news publishers to uphold accuracy, integrity, and respect for quality journalism as AI becomes more assimilated in news discovery and dissemination.
Condé Nast CEO Roger Lynch mentioned in an email reported by The New York Times that this partnership will help offset some revenue losses suffered by publishers due to technology companies. He emphasized the importance of meeting readers’ needs while ensuring proper attribution and compensation for the use of intellectual property with emerging technologies.
On the contrary, some media companies like The New York Times and The Intercept have taken legal action against OpenAI for using their articles without permission, indicating an ongoing legal dispute.
“I Last week, I sent out TechScape to our readers, hoping to take a break from writing about Elon Musk. However, my news editor had other plans: “Can you keep an eye on Elon Musk’s Twitter feed this week?”
Reading Musk’s tweets, I felt like my brain was melting. Even though I’d covered him for years, his online presence surprised me. From promoting Tesla and SpaceX to sharing cheesy nerdy jokes and diving into right-wing politics, his chaotic behavior had a new twist.
His briefest overnight break came on Saturday night, when he retweeted controversial content and then dived back in hours later with more questionable tweets.
Musk’s involvement in UK politics pushed him further into the far-right spectrum. Engaging with controversial figures like Lauren Southern and supporting Britain First’s co-leader, Musk’s online persona was more divisive than ever. Now, they are his supporters.
Well, that’s fine.
Today I’ll give you a good example from the world of AI that shows the difference between a scientific press release and a scientific paper. University of Bath’s press release claims AI poses no existential threat to humanity, but the actual research paints a different picture.
The study questions the capabilities of large-scale language models, suggesting that they are not as groundbreaking as claimed.
While the press release version is attention-grabbing, the scientific paper delves deeper into the limitations of AI capabilities. It highlights the challenges of ensuring AI safety in the face of emergent capabilities.
The paper reveals that emergent AI capabilities might not be as groundbreaking as they seem and are more controllable than believed. This sheds light on the complexity of AI safety in the face of evolving technology.
The pain of training
Nvidia’s use of YouTube data to train AI has led to legal troubles. A federal lawsuit alleges Nvidia stole videos from YouTube creators to train its AI, sparking a debate about intellectual property rights.
This lawsuit underscores the challenges AI companies face when sourcing training data. While some companies openly disregard copyright restrictions, others like Nvidia face legal battles over data usage.
On the other hand, companies like Google have a different approach due to their dominant position in the tech industry. Many websites allow Google to use their content for AI training to maintain visibility in search results.
Ask Me Anything
In my final TechScape after 11 years at the Guardian, I’m answering readers’ questions. Feel free to ask me anything tech-related, and I’ll do my best to provide insights and recommendations.
Google, the creator of Android, is set to release a range of new devices including smartphones, smartwatches, and earbuds featuring advanced AI technology like Gemini Live. This move is aimed at surpassing competitors like Apple and Samsung.
The new Pixel products unveiled at the event in California showcase Google’s commitment to integrating AI into its devices, showcasing their superiority over the competition.
Pixel 9 Series
The Pixel 9 Pro is Google’s first “pro” phone with a smaller screen size. Photo: Google
The new Pixel 9, Pixel 9 Pro, and Pixel 9 Pro XL feature Google’s advanced Tensor G4 chip with a faster AI processor, setting them apart from their competitors.
An exciting new AI feature, Gemini Live, promises natural conversations with AI assistants, reminiscent of sci-fi movies like Iron Man’s Jarvis.
Other notable features include the Pixel Studio image generator and the camera’s “Add Me” function, merging two consecutive images to include the photographer in group photos.
The Pixel 9 comes with a 6.3-inch screen and dual-camera system, the Pixel 9 Pro adds a telephoto camera with 5x optical zoom, and the Pixel 9 Pro XL features a 6.8-inch screen similar to last year’s model.
The “Pro” smartphones are direct competitors to Apple’s iPhone 15 Pro series, offering similar features and a year of access to Gemini Advanced.
All models include two years of free satellite SOS messaging in the U.S., akin to Apple’s latest iPhones.
Prices for the Pixel 9 start at £799, the 9 Pro XL at £1,099, and the 9 Pro at £999, available for shipping in August and September.
Pixel 9 Pro Foldable
The Google Pixel 9 Pro Fold is equipped with a large foldable screen. Photo: Google
Google’s Pixel 9 Pro Fold boasts a thinner, lighter design with an 8-inch flexible internal screen and a 6.3-inch external screen.
Featuring the Tensor G4 chip and advanced AI capabilities, the Pro Fold surpasses its predecessor and rivals other foldable phones on the market.
With 5x optical zoom, the triple camera system outperforms competitors like Samsung’s Galaxy Z Fold 6, offering innovative features like “Made You Look” animations.
The Pixel 9 Pro Fold is priced at £1,749 and will be available for shipping in September.
Pixel Watch 3 and Pixel Buds Pro 2
The Pixel Watch 3 and Pixel Buds Pro 2 offer advanced AI technology. Photo: Google
Google also revealed new accessories including the Pixel Watch 3 and Pixel Buds Pro 2. The smartwatch features improved health and fitness tracking, longer battery life, and advanced health monitoring capabilities.
The Pixel Buds Pro 2 offer enhanced noise cancellation and support for Gemini Live, allowing users to interact naturally with their AI assistant.
The Pixel Watch 3 is priced at £349 and the Pixel Buds Pro 2 at £219, set to ship in late September.
IAs is often the case, this incident began with drama in a WhatsApp group. In 2021, a loose coalition of my friends and acquaintances, in order to survive repeated lockdowns, started playing a much maligned online game called “Excuse.” Betrayal and backstabbing are all part of the game.
Except this time people went too far, someone got so upset that he left the group, and to get him back, my friends came up with a dubious plan: they got Nigel Farage, of all people, to make a sarcastic apology video, urging his comrades to rejoin the group.
This was easy to do, thanks to Cameo, which Farage had recently signed up for. For around £100, Farage had 24 hours to record a 60-second video clip and distribute it through the app. The ploy worked, and peace was restored within the friend group, thanks to Cameo and Farage’s rather bewildered recitation of a series of inside jokes that he completely failed to understand.
It’s one of the many uses for Cameo, a service that lets anyone request custom videos for friends from celebrities and influencers and have them delivered within 24 hours. The site became an unexpected hit during the COVID-19 lockdown, helping it attract investment that put it at a $1 billion valuation in 2021.
The front page of the Cameo website, featuring some of the celebrities registered with the site. Photo: Cameo
But since then, things have been tough. The video is still being made, and it’s still… Eclectic. A number of celebrities and influencers (more on that later) are on board, but they seem to be struggling to pay the relatively small bills.
last month, Business Insider Reported Cameo reaches settlement with 30 U.S. states. The company was found to have violated the Federal Trade Commission’s rules regarding celebrity endorsements.
Cameo was fined $600,000, a modest sum for a billion-dollar company in theory, but court filings attested that it couldn’t afford that amount and instead settled for $100,000 (split among 30 states). So what went wrong?
Day to day, the site operates much the same as before, offering custom videos to those seeking them from a range of actors, comedians and influencers, including British names like actor Miriam Margolyes (£134 a video), former footballer John Terry (£197), singer Gareth Gates (£47) and presenter and environmental activist Ben Fogle (£71).
In the US, NSync’s Lance Bass (235 pounds) and Dean Norris (235 pounds), who played DEA agent Hank Schrader, Breaking Bad (193 pounds), and former supporters of the U.S. version. Office Kevin Malone, aka Brian Baumgartner (154 pounds), or “Drunk Meredith” Kate Flannery (150 pounds).
While each has its own fanbase, sometimes with cult followings, Cameo has struggled to attract or retain A-list celebrities and influencers with top-tier followings, in part because using the app poses significant risks to a celebrity’s reputation: Some have used the app for more nefarious purposes, such as scripting videos to trick celebrities into making racist or other insults.
Last year, the “Russian group” A series of videos was commissioned and stitched together. Video recordings from Cameo and similar services made it appear as if various prominent U.S. figures were calling Ukrainian President Volodymyr Zelensky a corrupt drug addict, creating the illusion that they were helping to stage an intervention.
The disinformation campaign also included cameos from actor Elijah Wood and convicted rapist boxer Mike Tyson, though both have since withdrawn their cameos, and both Norris and Flannery are still active on the site.
Cameo takes 30% of creators’ fees for handling sales, transaction fees, and hosting and submitting the videos. Larger creators appear to have decided to forgo this potential revenue stream because it would require making lots of individual videos that never get published, or because of the risks. But smaller creators say it’s a convenient, hassle-free source of funding with few downsides.
CrowdStrike, the cybersecurity company that caused a massive global computer outage in July, has been sued for misleading investors.
A class action lawsuit filed in Texas by the Plymouth County Retirement Association, a pension fund, alleges that CrowdStrike misled investors by representing its technology as “verified, tested and certified,” when in fact, the investors allege, CrowdStrike's software was anything but.
“Defendants failed to disclose that: (1) CrowdStrike implemented insufficient controls over its Falcon update procedures and did not adequately test Falcon updates before deploying them to customers; (2) this improper software testing created a significant risk that the Falcon updates would cause widespread outages for many of the company's customers; and (3) such outages could, and ultimately did, result in significant reputational damage and legal risk for CrowdStrike.” As a result, the lawsuit alleges, “CrowdStrike's stock price was traded artificially inflated until the widespread outages allowed its stock price to recover.”
“We believe this lawsuit is without merit and will vigorously defend the company,” a CrowdStrike spokesperson said.
Securities fraud lawsuits typically arise after an adverse event has occurred for a company. If the reasons for a decline in a stock price were not clearly disclosed to investors in advance, a defendant may be able to prevail by arguing that the lack of disclosure constituted a fraudulent sale of the relevant shares.
CrowdStrike also faces more general legal liability for the outage. Delta Air Lines Chief Executive Ed Bastian estimated on Wednesday that the outage would force the cancellation of more than 5,000 flights and ultimately cost the company $500 million (£391 million). He said airlines had “no choice” but to seek damages as a result.
“To get priority access to the Delta ecosystem on the technology side, we need to test how it works. We can't just walk into a mission-critical operation that runs 24/7 and say there's a bug,” Bastian added. “We have to protect our shareholders. We have to protect our customers and employees, not just from costs but from damage to our brand and reputation.”
The outage, which crashed roughly 1% of Windows PCs worldwide, was estimated to have cost the Fortune 500 companies in the U.S. alone $5 billion. Nevertheless, the company's most visible response, aside from its efforts to restore service, was to thank “teammates and partners” who helped resolve the outage by sending $10 UberEats gift cards, though Uber quickly blocked the gift cards due to fears of possible fraud.
“Where did CrowdStrike go wrong?” is, if anything, a slightly overly generalized question.
You can also think about it the other way around: if you push an update to every computer on your network at the same time, by the time you find a problem, it’s too late to contain the impact. Alternatively, with a phased rollout, the update is pushed to users in small groups, usually accelerating over time. If you start updating 50 systems at once and then they all immediately lose connection, you hope you notice the problem before you update the next 50 million systems.
If you don’t do a staged rollout, you need to test the update before pushing it to users. The extent of pre-release testing is usually up for debate; there are countless configurations of hardware, software, and user requirements, and your testing regime must narrow down what’s important, and hope that nothing is overlooked. Thankfully, if 100% of computers with the update installed experience crashes and become inoperable until you manually apply a tedious fix, it’s easy to conclude that you didn’t test enough.
If you’re not doing a staged rollout and testing the update before it ships, you need to make sure that: Not broken.
Broken
Many flights at Orlando, Florida’s airport were canceled or delayed amid the CrowdStrike crisis. Photo: Miguel J. Rodriguez Carrillo/Getty Images
In CrowdStrike’s defense, I can understand why this happened. The company offers a service called “endpoint protection,” which if you’ve been in the Windows ecosystem for a few years, might be easiest to think of as antivirus. It’s built for the enterprise market, not the consumer market, and not just protects against common malware, but also tries to prevent individual computers used by companies from gaining a foothold on the corporate network.
This applies not only to PCs used by large corporations that need to provide every employee with a keyboard and mouse, but also to any other business with large amounts of cheap, flexible machines. If you left your house on Friday, you know what that means: advertising displays, point-of-sale terminals, and self-service kiosks were all affected.
The comparison is relevant because CrowdStrike is in a space where speed is crucial. The worst-case scenario, at least until last week, is a ransom worm like WannaCry or NotPetya, malware that not only does significant damage to infected machines but also spreads automatically in and out of corporate networks. So its first line of defense operates quickly: Rather than waiting for a weekly or monthly release schedule for software updates, the company pushes out files daily to address the latest threats to the systems it protects.
Though limited, even a phased rollout could cause real damage. WannaCry destroyed many NHS computers during the few hours it spread unchecked, before being accidentally halted by British security researcher Marcus Hutchins while trying to figure out how it worked. In this scenario, a phased rollout could result in loss of life. Delays in testing could be even more costly.
That means updates shouldn’t cause this kind of problem: rather than new code that runs on each machine, updates are more like dictionary updates that tell already-installed CrowdStrike software what new threats to look out for and how to recognize them.
At the loosest level, you can think of it as something like this article: You’re probably reading it through some application, like a web browser, an email client, or the Guardian app. (If you’ve arranged for someone to print this and deliver it to you with your morning coffee, congratulations!) We haven’t done a staged rollout or full testing of the article, because nothing would happen there.
Unfortunately, the update pushed out on Friday actually did something. High-level technical details remain unclear, and until CrowdStrike reveals the full details, we’ll just take their word for it. The update, which was meant to teach the system how to detect a specific type of cyberattack that had already been seen in the wild, actually “introduced a logic error, causing the operating system to crash.”
I’ve been covering this sort of thing for over a decade now, and my guess is that this “logic error” boils down to one of two things: Either an almost incomprehensible failure condition occurs in one of the most complex systems mankind has ever built, causing a catastrophic event through an almost unthinkable combination of bad luck, or someone does something incredibly stupid.
Sometimes there are no classes
Consumer self-service kiosks operated by Britain’s South Western Railway were also affected. Photo: Anadolu/Getty Images
There have been a lot of comments over the past few days.
This is an inevitable evil that results from the concentration of power in the technology sector in just a few companies.
This is an inevitable consequence of the EU prohibiting Microsoft from restricting antivirus companies’ ability to tamper with basic levels of Windows.
This is the inevitable harm of cybersecurity regulation that focuses more on checking boxes than on actual security.
This wasn’t a security issue because no one was hacked – it was just a bug.
None of it worked. CrowdStrike, despite the disruption it caused, doesn’t wield much power. It’s one of the big players in the space, but it’s installed on only about 1% of PCs. Microsoft says: They claim that the failure happened only because of regulations.Meanwhile, in the alternative where third-party security companies can’t operate on Windows, with Microsoft setting itself up as the only line of defense, it looks like we’ll be in a world where the first big failure actually affects 100% of PCs.
Cybersecurity regulations have actually benefited companies that have adopted CrowdStrike, making complicated certification processes into a simple checkbox check, and maybe that’s a good thing: “Buy a product to be safe” is the only reasonable request for the vast majority of companies, and CrowdStrike has delivered, except for that one unfortunate time.
But unfortunate or not, it was definitely a security issue. The golden triangle of information security has three goals: confidentiality (are the secrets kept secret?), integrity (is the data correct?), and availability (can the system be used?). CrowdStrike could not maintain availability, which meant they could not protect their customers’ information security.
In the end, the only lesson I can take comfort in is that this is going to happen more. We’ve managed so well with so many of our society’s failures that the ones that hit us from now on will be more unexpected, more severe, and less prepared for. Just as a driver can become so confident in their cruise control that they lose control right before an accident, we’ve managed to make catastrophic IT failures so rare that recovering from them is a marathon effort.
Yay?
The Wider TechScape
Social media automatically distributes problematic content to young men with little oversight. Illustration: Nash Weerasekera/The Guardian
“A complete river of rubbish”: Josh Taylor of The Guardian Australia Facebook and Instagram Algorithms The blank account fueled sexism and misogyny.
Is the world’s largest search engine broken? Tom Faber asks Google It is losing momentum.
Is this the end? The Story of Craig Wright? Post The Court’s Full Decision Post on your Twitter feed that you feel like the last decade of your career is final.
Parents have even more reason to worry, as AI technology overwhelms capture efforts. Child Abuser.
and Roblox Back in the spotlight Child sexual abuse failureCritics say the company’s privacy stance makes things worse.
AI could help us predict the weather more accurately
LaniMiro Lotufo Neto/Alamy
Google researchers have developed an artificial intelligence that they say can predict weather and climate patterns as accurately as current physical models, but with less computing power.
Existing forecasts are based on mathematical models run by extremely powerful supercomputers that deterministically predict what will happen in the future. Since they were first used in the 1950s, these models have become increasingly detailed and require more and more computer power.
Several projects aim to replace these computationally intensive tasks with much less demanding AI, including a DeepMind tool that forecasts localized rainfall over short periods of time. But like most AI models, the problem is that they are “black boxes” whose inner workings are mysterious and whose methods can’t be explained or replicated. And meteorologists say that if these models are trained on historical data, they will have a hard time predicting unprecedented events now being caused by climate change.
now, Dmitry Kochkov The researchers, from Google Research in California, and his colleagues created a model called NeuralGCM that balances the two approaches.
Typical climate models divide the Earth's surface into a grid of cells up to 100 kilometers in size. Due to limitations in computing power, simulating at high resolution is impractical. Phenomena such as clouds, turbulence, and convection within these cells are only approximated by computer codes that are continually adjusted to more closely match observed data. This approach, called parameterization, aims to at least partially capture small-scale phenomena that are not captured by broader physical models.
NeuralGCM has been trained to take over this small-scale approximation, making it less computationally intensive and more accurate. In the paper, the researchers say their model can process 70,000 days of simulations in 24 hours using a single chip called a Tensor Processing Unit (TPU). By comparison, competing models, called X-Shield A supercomputer with thousands of processing units is used to process the simulation, which lasts just 19 days.
The paper also claims that NeuralGCM performs predictions at a rate comparable to or better than best-in-class models. Google did not respond to a request for an interview. New Scientist.
Tim Palmer The Oxford researcher says the work is an interesting attempt to find a third way between pure physics and opaque AI approximations: “I'm uncomfortable with the idea of completely abandoning the equations of motion and moving to AI systems that even experts say they don't fully understand,” he says.
This hybrid approach is likely to spur further discussion and research in the modeling community, but time will tell whether it will be adopted by modeling engineers around the world, he says. “It's a good step in the right direction and the type of research we should be doing. It's great to see different alternatives being explored.”
circlehat Intention How will AI affect jobs? After “Will AI destroy humanity?”, this is the most important question about technology and it remains one that is extremely difficult to pin down, even as the frontier moves from science fiction to reality.
At one extreme there is the somewhat optimistic assertion that new technologies will simply create new jobs. At the other extreme there are fears that companies will replace their entire workforce with AI tools. The debate is often about the speed of the transition rather than the end state. A cataclysmic change that is completed in a few years is devastating to those caught in the middle, whereas a cataclysmic change that takes 20 years may be survivable.
Even the parallels with the past are not as clear-cut as we would like: the internal combustion engine eventually put an end to horse labor, but the steam engine, on the other hand, had a much bigger impact. increase Number of draft animals employed in the UK. Why? The arrival of the railways increased freight traffic in the country, but deliveries could not be completed from warehouse to doorstep. Horses were needed to do the things that steam engines could not do.
Until it isn’t.
Steam power and the internal combustion engine are examples of general-purpose technologies, breakthrough technologies that revolutionize the entire structure of society. There are not many such technologies, even if you count from writing, or even before that, from fire itself. It is pure coincidence that the initial letters of the term “Generative Pretrained Transformer” are the same, which is why GPT looks like GPT.
That’s not a job, idiot
Humans are not horses, and AI tools are not humans.
Humans are not horses [citation needed]It seems hard to believe that AI technology will be able to do everything humans can do. Becoming HumanThis is an inconveniently circular argument, but an important one: horses still race, because if you replace horses with cars, it’s no longer a horse race. [citation needed]people will still provide the services they want for one reason or another, and as culture warps around the rise of AI, some of those services will teeth You might be surprised. For example, AI in healthcare is underrated because for many people, the “human touch” is bad The problem is the doctor who worries they are judging your drinking, or the therapist who lies to you because they want you to like them.
As a result, many people like to think in terms of “tasks” rather than jobs: take a job, define it in terms of the tasks it contains, and ask whether an AI can do them. In doing so, we can identify some jobs that are at risk of being completely cannibalized, some jobs that are perfectly safe, and a large intermediate group of jobs that will be “impacted” by AI.
It’s worth pointing out an obvious fact: this approach results in a higher number of jobs that are mechanically “influenced” and a lower number of jobs that are “destroyed.” (Even the jobs most influenced by AI are likely to have some tasks that the AI finds difficult.) That may be why the technique was pioneered by OpenAI, who in a 2023 paper wrote: The researchers in the lab:“80% of workers are in occupations where at least 10% of the work requires a law degree, and 19% of workers are in occupations where more than half of the work requires a law degree.”
The report claimed between 15 and 86 professions were “completely at risk”, including mathematicians, legal secretaries and journalists.
I’m still here. But a year on, the idea is trending again, thanks to a paper from the Tony Blair Institute (TBI). The giant think tank, powerful and influential even before Labour’s landslide victory two weeks ago, is now seen as one of the architects of Starmerite thought. And it believes the public sector is ripe for disruption through AI. According to the TBI paper: The potential impact of AI on the public sector workforce (pdf):
More than 40% of the tasks performed by public sector workers could potentially be partially automated through a combination of AI-based software, such as machine learning models and large-scale language models, and AI-enabled hardware, ranging from AI-enabled sensors to advanced robotics.
Governments will need to invest in AI technology, upgrade data systems, train employees to use the new tools and cover the redundancy costs of early retirement – costs that are estimated to amount to £4 billion under ambitious implementation plans.That averages $1 billion a year for the term of this Congress.
Over the past few weeks TechScape has been keeping a close eye on the new Government’s approach to AI. Tomorrow, the King’s Speech is expected to announce the AI Bill, and we will hear more. The TBI paper makes one takeaway worth watching: Will investment in transformation approach £4 billion a year? There is a lot that can be done for free, but much more could be done with more money. The institute estimates that spending would return more than nine times, but a £20 billion bill would be hard to get through Parliament without question.
AI Geek
Prime Minister Tony Blair spoke at the Tony Blair Institute’s Britain’s Future conference on 9 July. Photo: Yui Mok/PA
The report drew renewed attention over the weekend as critics took issue with its methodology. From 404 Media:
Breaking down work into tasks is already done by a huge database created by the US Department of Labor. But with 20,000 such tasks, describing which ones should be exposed to AI is a daunting task. In a similar paper from OpenAI, “the authors personally labeled a large sample of tasks and DWAs, and hired experienced human annotators who reviewed the output of GPT-3, GPT-3.5, and GPT-4 as part of OpenAI’s tuning efforts,” but they also had the then-new GPT-4 perform the same tasks and found a 60-80 percent match between robots and humans.
circle
What if AI isn’t as great as we thought? What if its potential has been exaggerated to the point of being dangerous? Leading cancer experts in the NHS are raising concerns that the health service’s excessive focus on new technology is jeopardizing patient safety. As reported in our recent article:
Cancer experts warn that “innovative solutions” like new diagnostic tests are being falsely hailed as a “magic bullet” for the cancer crisis, but fail to address the root cause of cancer as a systemic issue.
The authors note that a common misconception among NHS leaders is that new technologies can eliminate inequalities, when in reality, technologies like AI can create additional hurdles for those with low digital and health literacy.
“We advise against adopting a technology-centric approach without a thorough evaluation of equity,” concludes the paper.
The Lancet Oncology publication argues for a back-to-basics strategy for cancer treatment. Their suggestions include increasing staff numbers, refocusing research on traditional areas such as surgery and radiation therapy, and establishing dedicated units for technology transfer to ensure proven treatments are integrated into routine care.
AI may divert attention from these critical improvements. While the technology holds much promise for the future, experts are concerned that hype about future advancements may detract from necessary present-day changes.
The paper describes AI as the latest iteration of “bionic duckweed,” a concept introduced by Stian Westlake in 2020 to highlight the use of future technologies to oppose current investments.
(Read More)
At a test site in Amsterdam, self-cooling artificial turf was tested against regular artificial turf.
Joris Forten
By absorbing rainwater and allowing it to evaporate on hot days, artificial turf stays much cooler than regular artificial grass. Self-cooling turf protects athletes from burns and heat stroke and helps cities manage stormwater.
Such surfaces are already in use in Amsterdam, London and Kobe, Japan, he said. Marjolein van Huygevoort At the KWR Water Resources Institute in Nieuwegein, The Netherlands.
“A normal lawn stays cool because the grass itself evaporates,” she says, “so this system mimics that natural situation by letting the water rise and evaporate.”
Van Huygevoort says many cities and sports venues are installing artificial turf fields because overuse can damage natural grass. Even in warm climates like those in Northern Europe, the sun can heat the plastic surface to about 70°C (158°F), not only damaging the health of athletes but also warming the air in urban areas, a situation that will only get worse with global warming, he says.
Inspired by A “blue-green roof” that collects rainwater for a cooling effectVan Huygevoort and her colleagues built a toy-sized mock-up of the playground in a climate-controlled lab, with a water-retention unit below the surface topped with a two-centimeter-thick cushioning pad with thin irrigation cylinders embedded inside.
Filled with mineral wool fibres that act like a sponge, these “capillaries” slowly wick water up into a thin layer of sand at the bottom of the artificial turf surface. In a heated laboratory environment, the irrigated water evaporates from the sand, creating a cooling effect on the surface.
Encouraged by these results, the team built a 25-square-metre test site in Amsterdam containing irrigated natural grass. During a heatwave that recorded a maximum temperature of 29.8°C (86°F), the conventional artificial grass reached 62.5°C (145°F). The researchers found that Self-cooling lawnBut the temperature never rose above 37°C (99°F), just 1.7°C warmer than on a natural grass field. Even the air above the water-cooled turf was cooler, which helped keep temperatures down across the city, van Huygevoort said.
The system is based on a design in which the rate at which the water level rises and the evaporative cooling process depend on various natural factors, such as weather conditions. “So water only evaporates when there is a demand for cooling,” says van Huygevoort.
A reservoir beneath the grass can store about 512,000 litres of rainwater beneath the floor of a standard 100-metre by 64-metre football pitch, she says, and capillaries in the buffer pad can store another 96,000 litres, meaning the field should be useful for absorbing large amounts of water during storms, van Huygevoort says.
The latest Surface tablet from Microsoft is expected to bring significant improvements, moving beyond just being faster, quieter, and more efficient. The key change is the switch to an Arm-based processor at its core.
This shift to Arm chips is not new for Microsoft, but this particular machine, the Surface Pro 11, has been touted as the most successful one yet, outperforming its predecessors like the Surface Pro X from 2020 and Surface Pro 9 5G from last year.
Powered by Qualcomm’s Snapdragon X chipset, the Surface Pro 11 offers improved performance and efficiency. While the new Arm chips offer advantages, there are still some compromises in terms of software and accessories.
The exterior of the new Surface Pro remains largely unchanged from its predecessor. Photo: Samuel Gibbs/The Guardian
Despite the new chip, the Surface Pro 11 retains key features that made its predecessor successful, including a robust built-in kickstand, high-quality aluminum frame, improved speakers, and faster Windows Hello facial recognition. The 13-inch OLED display on the high-end model is top-notch, offering an exceptional viewing experience.
However, the Surface Pro 11 comes at a premium price, starting at £1,049 (€1,199/$999/AU$1,899) without the keyboard. The higher-end model with an OLED screen and faster Snapdragon X Elite chip is even pricier, starting at £1,549 (€1,799/$1,499/AU$2,699).
The Flex Keyboard continues to function even when detached, offering versatility for users. Photo: Samuel Gibbs/The Guardian
The Surface Pro 11 works well as a tablet, but a detachable keyboard is essential for maximizing its potential as a laptop. Microsoft offers various keyboard options, including the high-end Flex keyboard priced at £340.
Specification
Screen: 13″ LCD or OLED 2880×1920 (267 PPI) 120Hz
Processor: Qualcomm Snapdragon X Plus or X Elite
RAM: 16 or 32 GB
Storage: 256, 512GB or 1TB
Graphics: Qualcomm Adreno
Operating System: Windows 11 Home
Camera: 10.5MP rear, 12.2MP front, Windows Hello
Connectivity: Wi-Fi 7, Bluetooth 5.4, 2 x USB-4, Surface Connect
Size: 287 x 209 x 9.3mm
Weight: 895g (without keyboard)
Snapdragon Power
The tablet includes two USB4 ports and a Surface Connect port for charging and accessories, but lacks a headphone jack. Photo: Samuel Gibbs/The Guardian
The transition to the Arm-based Snapdragon X Elite chip offers improved efficiency and performance over traditional Intel chips. Performance comparisons have shown promising results, with the Surface Pro 11 performing on par with top Intel chips and Apple’s M1.
In everyday use, the Surface Pro 11 is fast and responsive, running quietly most of the time. Battery life is comparable to its Intel-powered predecessor, lasting around 8 hours. However, under heavy workloads, the Surface Pro 11 with Snapdragon X Elite chip offers better battery life.
While many apps have been updated to work well on the new chip, there are still compatibility issues for some legacy software, resulting in slower performance. Additionally, certain Windows apps and games may not work at all on the new system, highlighting the need for further updates from software developers.
Paint’s Cocreator uses AI to enhance manual drawing, helping you turn rough outlines into impressive works of art. Photo: Samuel Gibbs/The Guardian
The Surface Pro 11 also introduces new AI tools from Microsoft, such as Paint’s Cocreator system, offering unique creative capabilities. While these features add value, some AI functions may be limited by current implementation.
Sustainability
The tablet’s removable SSD is accessible through a small door on the back. Photo: Samuel Gibbs/The Guardian
The Surface Pro 11 demonstrates Microsoft’s commitment to sustainability, with 72% recycled materials incorporated into its design, including aluminum and rare earth metals. The company also offers repair services and recycling programs for old devices, contributing to a more environmentally-friendly approach.
Price
Starting prices for the Microsoft Surface Pro 11 range from £1,049 (€1,199/$999/AU$1,899) for the base model with Snapdragon X Plus and an LCD screen. The higher-end model with Snapdragon X Elite and an OLED screen starts at £1,549 (€1,799/$1,499/AU$2,699).
Keyboard options are available starting at £139.99 (€159.99 / $139.99 / AU$239.95), with the premium Flex keyboard priced at £340.
Verdict
The Surface Pro 11 raises the bar for Arm-based Windows tablets, offering a sleek, powerful, and quiet device. However, app compatibility remains a crucial factor in determining the overall user experience.
While the performance is commendable, the Surface Pro 11 falls short in delivering promised battery life improvements. The premium features like the OLED screen and Flex keyboard come at a steep price, which may deter some potential buyers.
Although Arm chips show promise for lightweight devices, issues with app compatibility and AI features indicate that there’s still work to be done to fully embrace this technology.
Cons: High price, lack of included keyboard, app and accessory compatibility issues with Arm chips, absence of USB-A port, microSD card slot, or headphone jack, underwhelming AI features.
The All England Lawn Tennis Club has become the first to use artificial intelligence to protect Wimbledon players from online abuse.
The AI-driven service monitors players’ public social media profiles and automatically flags death threats, racist and sexist comments in 35 languages.
High-profile athletes who have been targeted online, including former US Open champion Emma Raducanu and four-time Grand Slam champion Naomi Osaka, have previously said they had to delete Instagram and Twitter (now named X) from their phones.
Britain’s number two, Harriet Dart, said she sometimes uses social media just because of the “hate” she sees online.
“I think there’s a lot of positives to take from this match,” Dart said after her win over British number one Katie Boulter on Thursday. [social media] But there was also a lot of negativity. If I opened the app today, I think I’d get a lot of hate, whether I won or not.”
Tournament director Jamie Baker said Wimbledon had deployed social media monitoring service Threat Matrix, developed by AI company Signify Group, which will also be rolled out to the US Open.
Baker said: “This is not something that would be found in the public domain. It’s not something that we would be shouting about, but we basically scroll through social media looking for this type of content and it means we have access to information that we wouldn’t have had access to before.”
“We’re not just going to rely on players to tell us what happened to them, but if there’s anything that we feel is of concern, then essentially our security team will step in and actually help address that.”
He said the AI-driven service is also supported by people monitoring accounts, and players can opt for a more robust service that scans for cheating and blackmail via private direct messages.
Baker, a former British number two, said Wimbledon would discuss the abuse with players and then report it to technology companies for removal or, if necessary, to police.
Explaining how the service works, Baker said: “If there’s an issue that we feel is of concern or worth reporting, we’ll ultimately communicate with the player and then work through the next steps. The benefit of this service is that it allows us to officially register the situation with the appropriate personnel.”
“But we can’t take those steps without actually engaging with the players and their teams and finding out what’s going on.”
World Rugby also uses the service, and in April an Australian was charged after a referee and his wife received threatening and abusive messages via Facebook during the Rugby World Cup.
Wimbledon said Threat Matrix conducted an investigation, monitoring more than 1.6 million public posts from X and 19,000 Instagram comments sent to 454 players competing in various professional tennis tournaments in 2022, and found that one in four players had been the target of abuse. 546 offensive posts were identified from 438 accounts.
The rise of artificial intelligence has propelled the stock prices of major tech companies to new heights, but this growth has come at the expense of the industry’s environmental efforts.
Google recently admitted that AI technology poses a challenge to its sustainability objectives. The company disclosed that its data centers, crucial for its AI infrastructure, have caused a 48% increase in greenhouse gas emissions since 2019. Google cited “significant uncertainties” in achieving its goal of net-zero emissions by 2030, particularly due to the complex and unpredictable environmental impacts of AI.
As the tech industry races ahead with AI advancements, the question arises: can technology mitigate the environmental impact of AI, or will the pursuit of cutting-edge innovation overshadow these concerns?
Why is AI a threat to tech companies’ environmental goals?
Data centers play a critical role in developing and operating AI models like Google’s Gemini and OpenAI’s GPT-4. These centers house complex computing equipment that require substantial electricity, leading to CO2 emissions both from energy sources and the manufacturing processes involved. According to the International Energy Agency, data centers are projected to double their electricity consumption by 2026, equivalent to Japan’s energy demand. Additionally, studies suggest that AI’s water consumption could reach significant levels by 2027, potentially straining resources equivalent to England’s annual consumption.
What do experts say about the environmental impact?
Government-sponsored reports in the UK have highlighted the importance of energy sources in determining the environmental cost of technology. Some experts caution that the reliance on fossil-fuel-powered energy sources for training AI models remains a significant challenge. While tech companies are increasing their use of renewable energy to meet sustainability goals, concerns persist that the lack of clean energy may push other users towards fossil fuels.
Alex de Vries, founder of Digiconomist, notes the dual challenge of rising energy consumption in AI and the struggle to secure sustainable energy sources.
Will there be enough renewable energy?
Global efforts to triple renewable energy resources by the end of the decade face challenges due to surging energy demands from AI data centers. The International Energy Agency warns that current plans may only double renewable energy capacity by 2030, potentially impacting climate goals.
Technology companies may need to invest heavily in new renewable energy projects to meet the escalating electricity needs driven by AI.
How quickly can new renewable energy projects be built?
While renewable energy projects like wind and solar farms can be developed relatively quickly, bureaucratic hurdles and grid connectivity issues can delay the process for years. The pace of building offshore wind and hydroelectric schemes faces similar challenges, posing concerns about whether renewable energy can keep up with the expansion of AI.
The reliance on existing low-carbon sources by tech companies may divert clean energy away from other users, potentially increasing fossil fuel consumption to meet growing demands.
Will AI’s power demands keep growing?
The escalating energy needs of AI systems could lead to higher energy costs, prompting cost-saving measures in the industry. However, the competitive landscape and the push for cutting-edge AI technologies may result in excessive electricity consumption despite rising costs.
The pursuit of state-of-the-art AI systems has fueled a “winner takes all” mentality among tech giants, compelling heavy investments in the development of advanced AI. The pressure to remain at the forefront of AI innovation, including the race towards achieving AGI, threatens to escalate energy consumption and costs.
Despite advancements in AI efficiency, the industry’s drive for innovation may offset potential energy savings, akin to the economic concept known as “Jevons’ Paradox.”
Won’t AI companies learn to use less electricity?
While AI breakthroughs continue to enhance efficiency, the industry’s relentless pursuit of cutting-edge models may counteract potential energy savings. The growth in AI capabilities does not necessarily translate to reduced energy consumption, leading to a paradox similar to historical instances of technological advancements increasing use rather than conserving resources.
Top row: Original image. Second row: AI-reconstructed image based on macaque brain recordings. Bottom row: Image reconstructed by the AI system without the attention mechanism.
Thirza Dado et al.
Artificial intelligence systems can currently create highly accurate reconstructions of what a person sees, based on recordings of brain activity, and these reconstructed images improve significantly as the AI learns which parts of the brain to pay attention to.
“As far as I know, these are the most accurate and closest reconstructions.” Umut Güçül Radboud University, Netherlands.
Güçül's team is one of several around the world using AI systems to understand what animals and humans see through brain recordings and scans. In a previous study, his team used a functional MRI (fMRI) scanner to record the brain activity of three people while they were shown a series of pictures.
In a separate study, the team used an implanted electrode array to directly record the brain activity of a single macaque monkey as it viewed AI-generated images — an implant done by a different team and for a different purpose, Güçül's colleagues say. Sarza Dado“We didn't put implants in macaques to restructure their perception,” she says. “That's not a good argument against doing surgery on animals.”
The research team has now reanalyzed the data from these earlier studies using an improved AI system that can learn which parts of the brain to pay most attention to.
“Essentially, the AI is learning where to pay attention when interpreting brain signals,” Gyuklüh says, “which of course in some way reflects what the brain signals pick up on in the environment.”
By directly recording brain activity, some of the reconstructed images were very close to the images seen by the macaques, as generated by the StyleGAN-XL image-generation AI. But accurately reconstructing AI-generated images is easier than real images, because aspects of the process used to generate the images can be incorporated into the AI training to reconstruct those images, Dado said.
The fMRI scans also showed a noticeable improvement when using the attention guidance system, but the reconstructed images were less accurate than those for the macaques. This is partly because real photographs were used, but Dado also says that it is much harder to reconstruct images from fMRI scans. “It's non-invasive, but it's very noisy.”
The team's ultimate goal is to develop better brain implants to restore vision by stimulating the higher-level parts of the visual system that represent objects, rather than simply presenting patterns of light.
“For example, we can directly stimulate the area that corresponds to a dog's brain,” Güçül says, “and in that way create a richer visual experience that is closer to that of a sighted person.”
TThe state of the art in AI just got a little bit further along: On Friday, Anthropic, an AI lab founded by a team of disgruntled OpenAI staffers, released the latest version of its Claude LLM. From Bloomberg:
The company announced on Thursday that a new model of the technology behind its popular chatbot, “Claude,” is twice as fast as its most powerful predecessor. In its evaluation, Anthropik said the model outperformed leading competitors such as OpenAI in several key intelligence capabilities, including coding and text-based reasoning.
Anthropik just released the previous version of Claude, 3.0, in March. This latest model is called 3.5, and it’s currently only available on the company’s mid-range model, “Sonnet.” The company says a faster, cheaper, less powerful “Haiku” version is coming soon, as well as a slower, more expensive, but most powerful “Opus.”
But even before Opus arrived, Anthropic claimed to have the best AI on the market. In a series of head-to-head comparisons posted on the company’s blog, 3.5 Sonnet outperformed OpenAI’s latest model, GPT-4o, in tasks like math quizzes, text comprehension, and undergraduate-level knowledge. It wasn’t a clean sweep, with GPT maintaining the lead in several benchmarks, but it was enough to justify the company’s claim that it’s on the cutting edge of what’s possible.
From a more qualitative perspective, AI seems to be a step forward. Anthropic states:
They have a significantly improved ability to understand nuance, humor, and complex instructions, and they excel at writing high-quality content in a natural, relatable tone.
They’re grading their own homework, and their explanation matches the changes I’ve noticed: No matter where the technical benchmarks are, I find talking to the latest version of Claude more enjoyable than any AI system I’ve used before.
But the company isn’t just selling power updates. Instead, in a way favored by smaller competitors around the world, Anthropic is focusing as much on cost as it is on features. The company claims that Claude 3.5 is not only smarter than its predecessor, but also cheaper.
AI is a concern for Apple as it consumes a lot of power.
During its global developers conference, Apple unveiled its strategy to integrate AI into daily life, primarily focusing on the latest iPhone users.
Apple’s latest AI models are compatible with the iPhone 15 Pro and Pro Max, the only devices featuring the A17 processor. Additionally, Macs up to three years old with M1, 2, or 3 chips, as well as iPad Pros with similar internal hardware, can benefit from the upgrade.
The more affordable iPhone 15 models come with the A16 Bionic chip introduced in 2022 and 6GB of memory, compared to 8GB in the pricier Pro models. This difference is crucial because the M1 chip powering Macs is equivalent to the A14 processor in 2020 iPhones.
Numerous model numbers highlight that advanced AI features won’t function on just any phone, as many require high-performance devices. If Apple aims to deliver AI technology, it must do so through its data centers—an endeavor that poses challenges, as stated by Kari Paul:
At the core of Apple’s AI privacy measures is its new private cloud computing technology, where most of the computing is done in-house for Apple Intelligence features on devices. However, for tasks exceeding device capabilities, processing is outsourced to the cloud while safeguarding user data.
To uphold privacy, Apple only exports necessary data for each request, implements additional security measures at endpoints, avoids indefinite data storage, and offers tools and software related to its private cloud for third-party validation.
When it comes to AI queries, complete privacy—offered by online backup or messaging services—remains challenging due to server requirements for accurate responses. Apple has long stressed its commitment to privacy, setting itself apart from competitors like Facebook and Google with its “what happens on iPhone stays on iPhone” pledge.
Apple CEO Tim Cook attending an event in Cupertino, California in September 2023.
Photo: Bloomberg/Getty Images
Apple’s solution involves running user data-free data centers designed in-house to validate the integrity of the software. Security researchers are provided with tools to verify the software’s authenticity running on Apple’s servers.
Yet, the question remains: Can Apple be trusted? Huawei’s similar efforts failed to prove its independence from the Chinese government. Trust in Apple’s commitment to privacy is growing, but accommodating AI’s rise forces Apple to compromise its foundational principles.
While Apple emphasizes privacy, the implementation of AI features like Apple Intelligence may necessitate data transfer to ensure functionality, blurring the lines of privacy assurances.
Considering a transition from smartphone to a light phone?
The Light Phone III, a device enticing those seeking freedom from distractions.
Photo: LightPhone
Exploring products outside the conventional smartphone market reveals devices like Humane and Rabbit, showcasing the expanding realm of hardware addressing users’ varying needs.
Anti-phones, exemplified by devices like the Light Phone III, cater to individuals desiring a balance between digital detox and modern conveniences, offering customizable tools optimized for an unobtrusive experience.
The Light Phone III provides a range of optional tools tailored for LightOS, including alarms, calculators, calendars, directories, and more, designed for a thoughtful user experience.
The device’s intentional limitations, such as omitting a web browser, restrict access to streaming services and encrypted messaging platforms, aligning with the anti-distraction philosophy.
Navigating the transition to an anti-phone involves weighing the desire for reduced digital demands against the practicalities of work and personal life, posing a contemplative dilemma.
Exploring the broader technological landscape
A captivating portrayal of “AI” by Miles Astley.
Photo: Miles Astley
G7 leaders have raised concerns about China’s support for Russia in the Ukraine war and the production of cheap goods causing “harmful overcapacity,” despite German apprehensions.
During the annual summit held in Puglia under the Italian presidency, a 36-page report by the U.S. condemned Chinese subsidies on products like solar panels and electric vehicles, attributing them to global distortions, market disruptions, and overcapacity that threaten worker resilience, industry security, and economic stability.
Specifically, U.S. officials pointed out China as a major supplier of materials used by Russia against Ukraine, expressing concern over the long-term security implications. Despite Ukrainian President Zelensky’s assurance that Chinese leaders vowed not to provide weapons to Moscow, U.S. President Joe Biden highlighted China’s arms supply to Russia.
The U.S., Japan, and the EU, along with an informal eighth partner at the G7 summit, have expressed worries over Beijing’s heavy subsidies in green energy and technology sectors flooding global markets with unfairly priced products, creating stiff competition for Western companies, especially in the green technology space.
The National Security Council spokesman, John Kirby, mentioned plans to address China’s non-market policies having detrimental global effects. China’s top official, First Vice Premier Ding Xuexiang, is set to visit Brussels to discuss EU plans for increased tariffs on Chinese-made electric cars.
In a bid for diversity at the summit, global leaders from countries like India, Turkey, UAE, Brazil, and Mauritania were invited to participate. The G7 emphasized the importance of cooperation to address collective challenges, as expressed by Italian Prime Minister and G7 President Giorgia Meloni.
Russia faced stiff consequences at the summit, including wider sanctions, loss of control over state assets, and a new 10-year US-Ukraine security pact. A proposed $50 billion loan to Ukraine funded by interest profits from Russian state assets marked the beginning of economic pressure on Russia.
The final statement from the summit demanded Russia to cease its illegal aggression in Ukraine, pay reparations for inflicted damage, and explore legal options to enforce compliance. Russia dismissed the security pact as a nominal agreement and criticized the appropriation of frozen asset proceeds.
While Zelensky addressed Indian Prime Minister Modi on revising India’s reliance on Russian oil, discussions focused on the rising oil prices set by Russia and India’s growing purchases of Russian offshore crude.
The U.S. and EU impose price restrictions on Russian crude sales to prevent Western involvement unless sold below a certain cap. India, currently the largest buyer of Russian offshore crude, has not joined this ban, prompting calls for stricter price caps and actions against transportation exceeding the imposed price limit.
Microsoft, OpenAI and Nvidia are under increased scrutiny for their involvement in the artificial intelligence industry as U.S. regulators have reportedly agreed to investigate these companies.
The New York Times reported that the US Department of Justice and the Federal Trade Commission (FTC) have reached an agreement to investigate key players in the AI market, with the investigation expected to be completed within the next few days.
The Justice Department will lead an investigation into whether Nvidia, a leading chip maker for AI systems, has violated antitrust laws aimed at promoting fair competition and preventing monopolies, according to Wednesday’s NYT.
Meanwhile, the FTC will scrutinize OpenAI, the developer of the ChatGPT chatbot, and Microsoft, the largest investor in OpenAI and supporter of other AI companies.
The Wall Street Journal also reported on Thursday that the FTC is investigating whether Microsoft structured a recent deal with startup Inflection AI in a way to avoid antitrust scrutiny.
In March, Microsoft hired Mustafa Suleiman, CEO and co-founder of Inflexion, to lead its new AI division and agreed to pay the company $650 million to license its AI software.
The FTC has shown interest in the AI market before, ordering OpenAI, Microsoft, Google parent Alphabet, Amazon, and Anthropic to provide information on recent investments and partnerships involving generative AI companies and cloud service providers.
An investigation into OpenAI was launched last year based on allegations of consumer protection law violations related to personal data and reputations being at risk.
Jonathan Cantor, head of the Justice Department’s antitrust division, stated that the department will “urgently” investigate the AI sector to examine monopoly issues and the competitive landscape in technology.
The Alameda, California City Council decided not to permit ongoing controversial experiments related to technology that aims to brighten clouds. The project involves spraying saltwater onto a former aircraft carrier’s deck at the city’s pier to test devices that can create and measure aerosol plumes. This research could lead to marine cloud brightening, a form of climate intervention aimed at making clouds more reflective to send heat back into space and help mitigate global warming. Despite the potential long-term benefits, the council unanimously voted against continuing the experiments.
This decision placed city officials in the midst of a national debate about geoengineering and whether testing such technology should be allowed. The council’s vote does not reject the science or the concept of geoengineering but rather criticizes the lack of transparency, safety vetting, and misguided approach of the researchers. The researchers from the University of Washington had already begun the experiment on the USS Hornet in Alameda without proper public disclosure. The experiments aimed to study the effects of increasing water droplets in clouds to make them more reflective to sunlight, which could help reduce global warming.
The council’s concerns focused on potential health risks to the community surrounding the experiment site, lack of regulatory measures, and insufficient transparency from the project leaders. Despite claims of safety from the researchers, city leaders stopped the experiments, prompting a discussion on the risks and benefits of geoengineering.
While project organizers expressed disappointment with the council’s decision, outside environmental groups warned about the broader implications of geoengineering, fearing unintended consequences that could impact global climate efforts. The debate highlights the challenges of advancing geoengineering research in the face of skepticism and concerns about the potential risks associated with altering natural systems.
yes
Have you ever heard the story about a drunk man looking for his keys under a streetlight? After pacing back and forth for a while and rummaging through the floor, his friend asks him where he thinks he dropped his keys. He points to a dark spot across the street. “Then why don’t you look there?” his friend asks. He shrugs. “Because that’s where the light is.” It’s a good joke. Everyone laughs.
Let’s talk about online political advertising.
“Microtargeting” doesn’t exist anymore, explains The Guardian’s Jim Waterson.
Don’t expect to see Cambridge Analytica-style micro-targeting of political ads using personal information in this general election. The tactic is now seen by many as ineffective “bait” and is increasingly being blocked by social media platforms. Digital strategist Tom Edmonds said Facebook has banned political campaigns from using many of the tactics used in past elections. “If you run a campaign to 500 people, you’re not getting a ton of revenue, you’re just getting a ton of harassment,” he said.
Microtargeting was feared because of its potential negative effects on democracy — if you can target 1,000 different messages to 1,000 different demographics, the very notion of a single national conversation begins to break down — but in reality, microtargeting never really worked.
After all, the biggest competitor for a company like Cambridge Analytica was Facebook itself: the social network’s advertising tools make it less worthwhile to spend billions creating profiles and microtargeting individual voters when you can leave all the targeting decisions to Facebook itself. The social network allows advertisers to set “performance objectives.” [like sales, clicks, or signups]You set a spending limit and then you just sit back and wait for the company to do whatever it takes to maximize your profits. The company will also choose the best combination of words and images to increase your chances of success.
But Facebook can only help you so much. For example, if you’re creating ads for a particular candidate, who should you focus your time and money on? Those who are likely to win, or those who are sure to lose? If you answered the latter, you’d be better off working for the Conservative Party. From our article:
The strategy, known within the party as the “80/20” approach, involves concentrating all of its spending on the 80 seats it is most likely to lose in 2019 and the 20 seats it is most likely to gain.
Facebook’s ad spending reports show the party is pouring money into exactly these constituencies: Since January, more than half of the party’s spending on the social network has been directed to the 80 closest constituencies or those not held by the party at all.
A conference staff member speaks in front of Facebook’s demo booth at F8, Facebook’s annual developer conference in San Jose, California. Photo: Noah Berger/AP
We began monitoring meta ad spending to see if the reported “80/20 strategy” held up. It’s one thing to propose it two years before an election, but quite another to follow through with it just a month away.
But we also started monitoring Meta’s ad spending because we could. The company keeps a library of all political ads, publishes total spending, and requires residency verification before launching a new ad. This library has received a lot of criticism over the years, but at least it exists. Not only that, but the library has a powerful toolset that allows you to write your own software to query and answer questions more serious than “are there any interesting ads that someone paid for recently?”
But like a drunk person searching for his keys, it’s unlikely that this topic is actually on Facebook. Across large swaths of the country, conversations that once took place on public social networks have migrated to private channels, led by Meta’s WhatsApp. What’s left of Facebook itself is smothered in AI-generated rubbish and disconnected from reality by algorithmic tweaks that highlight “friends and family” content. That trend is doubly pronounced on Meta’s Twitter clone, Threads, which actively and openly downgrades any kind of political content.
Although conversations are growing on TikTok, the platform is difficult to cover: Observer research into digital campaigns has had to focus on the official TikTok feeds of political parties.
TikTok is free; paid advertising by politicians or political parties isn’t allowed. But it won’t be easy. Social media teams will have to work harder to convince the app’s notoriously opaque algorithms to let their content flow organically to users’ phones. The more people who like, share, comment and repost a video, the better the chances. For smaller, more agile parties with smaller budgets, TikTok can feel like it has everything to win: views, engagement and people finally finding out they exist. Creators who know how to do it think Labour is off to a good start.
Election conversations are happening on TikTok — and there’s a lot of it, as the platform’s tightly curated algorithmic feed allows people of all ages to have their own discussions — but it’s nearly impossible to observe from the outside without using brute force techniques like tallying up views of videos tagged “snack.”
Of course, WhatsApp conversations are even worse: with end-to-end encryption and sparse public “channels,” doing data journalism tracking election chats is a dead end.
And then there’s AI. Suspicions remain that the rise of AI systems will have some impact on this election, but here, too, we have to look at where the light is. It’s very clear (and we’ve never really seen it before) that deepfake videos are circulating on Twitter, the platform now known as X. What’s invisible to us is that wavering voters are conversing with ChatGPT to try to decide where to type X, if that’s even happening in the first place.
In the UK, these questions feel largely academic; apart from personality-driven local elections, the final outcome feels more foregone than at any time in my life. But in the US, where voters go to the polls in five months’ time, the same questions will be asked. And the answers may hold the key to which side the coin falls.
So let’s go find them right away.
A wider texscape
Footage from the fake documentary “Olympics Has Fallen” produced by Russian influencer actor Storm-1679. Photo: Storm-1679/Microsoft Threat Analysis Center
by the way Deepfake According to Microsoft, the fake Tom Cruise video (pictured above) was used to spread disinformation about the Olympics.
Is the Internet Bad?. That is certainly Marbonot everything has been smooth sailing during the first nine months online.
Internal Google Database Tracking Privacy and Security Breaches Leaked to 404 MediaOne of the biggest threats is that YouTube employees could secretly check upcoming big video uploads to get information ahead of time.
Voters support raising the minimum age Social media The number of people using the app in the UK has risen to 16, according to a poll by The Guardian.
Microsoft’s “RecallA “clone of the Mac app Rewind” has been built into the OS. It has been described as a security “catastrophe”.The AI service stores everything a user has ever seen on a computer in a database to help answer questions for law students, which critics say makes it an attractive target for hackers.
According to the CEO of the AI hardware company, the next industrial revolution has already begun. The announcement was made at a crowded stadium in Taipei on Sunday by Nvidia, as they revealed new products and plans to drive advancements in artificial intelligence.
Jensen Huang, the CEO of Nvidia, is currently attending Computex, Taiwan’s largest technology exhibition, alongside CEOs from major semiconductor companies such as AMD, Intel, and Qualcomm. The focus of the event is on establishing AI as a mainstream technology in the industry.
Having been born in Taiwan, Huang is a prominent figure on the island, and his presence has generated significant interest from the media and the public. Nvidia is recognized as the leader in specialized chips and hardware essential for the development and operation of cutting-edge AI systems.
During his address at the National Taiwan University’s Sports Center, Huang mentioned the collaboration between companies and countries with Nvidia to revamp their traditional data centers into high-speed computing facilities, focusing on creating an AI factory for the mass production of artificial intelligence.
He introduced the Nvidia ACE generation AI, which can generate lifelike human avatars for industries like customer support. Several top technology companies, including Foxconn and Siemens, are leveraging Nvidia’s platform to develop AI-driven autonomous robots.
Nvidia recently unveiled its Blackwell platform, and Huang disclosed plans to launch an “Ultra” version in 2025. He also provided a glimpse of their upcoming graphics processing unit architecture, codenamed Rubin. Huang emphasized Nvidia’s commitment to accelerating the release of new GPU products annually.
In his forward-looking speech, Huang predicted that generative AI would play a significant role in almost every interaction with the internet or computers in the future. He concluded by praising Taiwan’s advanced semiconductor industry, which plays a crucial role in manufacturing essential components for various technologies.
Keynote addresses at Computex are also expected from AMD’s Lisa Su and Qualcomm’s Cristiano Amon, outlining their companies’ plans in AI. Other speakers include Intel’s CEO Pat Gelsinger and Arm’s president Rene Haas, highlighting the significance of AI-accelerated technologies.
Taiwanese manufacturers are pivotal to technology companies’ AI strategies as they produce advanced semiconductors required for powerful AI applications. Foxconn, known for electronics production, has shifted towards AI hardware, with their CEO projecting substantial growth in the AI server market.
Despite Taiwan’s crucial role in the global supply chain, there are concerns over China’s territorial claims and potential use of force. Tensions between Beijing and Taipei have escalated, with China conducting military exercises near Taiwan, including simulated blockades.
I I know I’m real. And you, dear reader, know that you are the real deal. But have you ever wondered if there’s something strange about other people on the internet? Feeling like the spaces you used to frequent are a little dead? You’re not alone. The “Dead Internet Theory” first appeared on the web nearly three years ago and was catapulted into the mainstream by: Atlantic Essay by Caitlin Tiffany:
The dead internet theory suggests that the internet has been almost completely taken over by artificial intelligence. Like many other online conspiracy theories, this one’s audience has grown thanks to discussions by a mix of true believers, cynical trolls, and bored and curious chatterboxes… But unlike many other online conspiracy theories, this conspiracy theory has no morsel of truth to it. Person or Bot: Does it really matter?
At the time of writing, the deadest part of the internet was the moribund pre-Mask Twitter. The site’s active curation provides the same “relevant content” to hundreds of thousands of users, who can post things like “I hate texting, so come over here and give me a hug” on Twitter. Adjusted and reposted. The distinction between humans and bots has also been blurred by recommendation algorithms that make humans behave like bots.
Beyond that central idea, the 2021 version of the conspiracy theory has taken a strange turn. One supporter, Tiffany, suggests that “the internet died in 2016 or early 2017 and is now not just ’empty and empty’ but ‘totally barren.’ …As evidence, the Illuminati pirates say, ‘I’ve seen it.'”
This theory was not wrong. It was just too early. Talk about the internet that died in the summer in front ChatGPT’s release echoes my colleagues at the Guardian who confidently declared in the summer of 2016 that: The next few years will be quiet.”
In 2021, the internet felt like death. This is because aggressive algorithmic curation has made people behave like robots. In 2024, the opposite will happen. Robots will now post just like humans. Here are some examples:
on Twitter itself, Musk rescues the site from the frying pan, throws it into a volcano, and then a poorly thought out monetization scheme buys a blue checkmark, attaches it to a large language model, and spins it out of control in response to viral content. I was able to make a profit by doing so. This social media network is currently paying verified users a portion of the ad revenue they receive from their comment threads, turning the most viral posts on the site into low-stakes Allbots battle royales. .
Death pervades Google. Being at the top of search results is a valuable position, so valuable that companies competing for it can’t afford to actually write about it. No problem. ChatGPT can create anything in an instant. Of course, this is only worth it if the resulting visitors are people who can make you money. Bad news, because…
…all over the web, bots account for about half Percentage of all internet traffic, according to a study by cybersecurity firm Imperva. Almost a third of all traffic is what the company calls “malicious bots,” carrying out everything from ad fraud to brute force hacking attacks. But even the “good bots” struggle to fall into this category. Google’s “crawlers” were welcome when updating search entries, but less so when they just trained an AI to repeat what users wrote, without submitting users. did.
And then there’s Crab Jesus. An unholy combination of Facebook content farms, AI-generated images, and automated testing to determine what goes most viral. led to weeks of viral content It features a combination of Jesus, a crustacean, and a female flight attendant. One such image depicted Jesus wearing a jacket made of shrimp and eating shellfish. Adding to the confusion was the sight of a kind of crab centaur savior walking arm in arm with what appeared to be the entire crew of the long-distance flight on the beach. It was at least interestingly bizarre and a step up from the previously viral 122-year-old female friend who posed in front of a homemade birthday cake.
As much as I’d like to offer a ray of hope, a little tip to reinvigorate the internet, I can’t. It really feels like the consumer internet is in the late stages of a zombie apocalypse. The good news is that there is a safe haven. While “private socials” like WhatsApp and Discord servers can hide from the onslaught in secrecy, smaller communities like Bluesky and Mastodon are hidden and safe for now.
In the medium term, I expect to see large platforms returning to the wilds of their services and trying to bring some humanity back to their services through a combination of account authentication and AI detection. But whether it will be too late by then is an open question.
Musk still needs a Twitter sitter
Elon Musk in Beijing in 2023. Photo: Wang Teishu/Reuters
At least there’s still one person on the internet. It’s Elon Musk. He spent $44 billion getting obsessed with posting and being called idiots on the platforms he owns. So his latest legal defeat will hit a sore spot after the U.S. Supreme Court declined to accept his plea to be released from his court-appointed posting babysitter. . From our story:
The Supreme Court on Monday rejected Elon Musk’s appeal over a settlement with securities regulators that required him to get prior approval for some tweets related to his electric car company Tesla.
The justices did not comment on leaving the lower court’s ruling against Musk in place, but Musk complained that the requirement violated the First Amendment and constituted a “prior restraint” on his speech. . The ruling came a day after he made an unannounced visit to China to secure a deal to deploy Tesla’s driver-assistance features locally.
For those who don’t have an encyclopedic memory of Elon, Musk tweeted in 2018 that he had “secured funding” to take Tesla private. The company was never taken private, and subsequent lawsuits revealed that he had only discussed it a few times at most. To end the bill, Musk resigned as Tesla chairman, paid $20 million and agreed to have in-house lawyers pre-approve all social media posts about the electric car maker.
He has since regretted it and is fighting to overturn that part of the contract (which he entered into voluntarily to avoid an adverse trial). “The preclearance clause at issue continues to cast an unconstitutional chill on Mr. Musk’s speech whenever he considers making it publicly,” his lawyers argued.
Well, the Supreme Court of the United States doesn’t care. The government did not take up his case, tacitly deciding that no real constitutional issue was at issue.
What’s strange is that the company’s in-house lawyers already seem to be taking a very hands-off approach to Musk’s posts. On Friday, he responded to early Facebook employee Dustin Moskowitz’s claim that Tesla is “the next Enron” by posting a photo of a dog putting its testicles in another dog’s face. (Please click at your own risk.) If that’s Mr. Musk’s tweet with “unconstitutional chills,” I don’t want to know what he would send if he felt truly free.
Wider TechScape
Artwork for Everyone Knows That. Illustration: Getty; Guardian Design
T
A few weeks ago, it was quietly announced that the Future of Humanity Institute, a famous interdisciplinary research center in Oxford, no longer has a future. It closed without warning on April 16th. Initially, its website contained only a short statement that it had been closed and that research could continue elsewhere within or outside the university.
The institute, dedicated to the study of humanity’s existential risks, was founded in 2005 by Swedish-born philosopher Nick Bostrom and quickly made a name for itself beyond academia. Many high-tech billionaires praised the institute, especially in Silicon Valley, and provided financial support.
Mr. Bostrom is perhaps best known for his 2014 best-selling book. super intelligence, which warned of the existential dangers of artificial intelligence, but also became widely known for his 2003 academic paper “Are You Living in a Computer Simulation?” The paper argues that over time, humans are likely to develop the ability to create simulations that are indistinguishable from reality, and if this is the case, it has already happened and we may be the simulation. insisted.
I interviewed Bostrom more than a decade ago, and he had one of those elusive and rather abstract personalities that perhaps lends credence to simulation theory. He was pale, had a reputation for working all night, and seemed like the type of person who didn’t go out much. The Institute appears to be aware of this social shortcoming. final reporta long inscription written by Fuji Heavy Industries researcher Anders Sandberg states:
“We have not invested enough in the politics and socialization of the university to build long-term, stable relationships with faculty…When epistemology and communication practices become too disconnected, misunderstandings flourish.”
Nick Bostrom: “Proudly provocative on paper, cautious and defensive in person.” Photo: Washington Post/Getty Images
Like Sandberg, Bostrom is an advocate of transhumanism, the belief in using advanced technology to improve longevity and cognitive abilities, and is said …
If ByteDance exhausts all legal avenues to fight a bill banning TikTok from U.S. app stores, the Chinese company would prefer to shut down the app rather than sell it, according to four sources.
The core algorithms that TikTok relies on are crucial to ByteDance’s overall operations, making it highly unlikely that the app powered by these algorithms would be sold, sources close to the parent company said.
Despite TikTok’s popularity with over 1 billion users, it still operates at a loss and represents a small portion of ByteDance’s total revenue and daily active users. In a worst-case scenario, the parent company would choose to shut down TikTok in the US rather than sell it to an American buyer.
Shutting down TikTok would have minimal impact on ByteDance’s business, allowing the company to maintain its core algorithms, as per unidentified sources not authorized to speak to the media.
ByteDance confirmed in a statement on Toutiao that it has no plans to sell TikTok, responding to speculations suggesting otherwise. The company is not considering selling TikTok’s US business without its recommendation algorithm.
TikTok CEO Shou Zi Chew expressed optimism about blocking the bill signed by President Joe Biden, banning the app used by 170 million Americans. The company believes it will win the appeal against the ban.
The bill, passed by the US Senate, raises concerns about potential access to Americans’ data by China or surveillance through apps. The deadline for TikTok’s sale is set for January 19, but it could be extended if ByteDance shows progress.
ByteDance’s revenue increased from $80 billion in 2022 to nearly $120 billion in 2023, with TikTok’s US daily active users accounting for 5% of the company’s total DAUs worldwide, sources familiar with the matter revealed.
The core algorithm shared by TikTok and ByteDance’s domestic apps like Douyin is a valuable asset that ByteDance is not willing to sell to a competitor. Separating the algorithms from TikTok’s US assets would be complex, making it an unlikely option, sources added.
Former US Treasury Secretary Steven Mnuchin’s interest in forming an investor group to acquire TikTok may not attract buyers for ByteDance’s US assets, excluding the algorithm, sources noted.
Investors valued ByteDance at $268 billion when they offered to repurchase about $5 billion worth of stock in December.
Artist's impression of the Brightline West high-speed rail line
brightline west
Today, construction begins on America's first full-scale high-speed rail line, connecting the suburbs of Los Angeles to the bright city of Las Vegas, Nevada. The project could not only allow people in the United States to finally experience high-speed passenger trains of European and Asian standards, but also provide a commercial model for building high-speed rail elsewhere in the United States. be.
A groundbreaking ceremony in Las Vegas today, attended by U.S. Secretary of Transportation Pete Buttigieg, along with officials from Nevada and California, marked the official start of construction on the Brightline West project. Brightline West, which aims to be completed within four years in time for the 2028 Summer Olympics in Los Angeles, will bypass lines of stranded cars and cross the median of Interstate 15 at speeds of 320 km/h. It is expected to fly passengers at speed. In the weekend traffic jam.
The $12 billion project is a bold gamble for Brightline and its owner, Fortress Investment Group, even considering a $3 billion federal grant announced by President Joe Biden in December 2023. It may seem like. But there are several reasons why Brightline West will be more successful than the rest of the United States. High-speed rail projects are delayed.
According to sources, Brightline is focused on connecting major markets that are approximately 400 to 550 kilometers apart. report By infrastructure consultancy AECOM. This represents a sweet spot where high-speed rail is highly competitive with driving and flying. His 350-kilometre trip on Brightline West from Las Vegas to the Los Angeles suburbs is expected to take him just over two hours, making it an attractive alternative to his four-hour drive, which 50 million people travel between cities each year. This is a great alternative.
“High-speed rail has proven to be a very efficient way to move large numbers of passengers within median distances,” he says. Jiao Junfeng At the University of Texas at Austin. “There are many success stories out there in European countries and Asian countries, and there are markets where high-speed rail operations have proven profitable.”
Another factor in Brightline's favor is that it leased access from Nevada and California to build Brightline West through the existing Interstate 15 corridor. This avoids the costs and delays typically associated with obtaining rights of way and acquiring land.
Reducing the risk of delays can also reduce overall project costs in the long term. California's own high-speed rail project was first approved by voters in 2008 to connect San Francisco and Los Angeles; project cost Soared from $33 billion to $128 billion. Other high-speed rail projects are currently being considered in Texas and the Pacific Northwest.
“When you're talking about preparing for construction or progressing construction, time is not on your side. [because of] say “inflation” Jean Whittington at the University of Washington in Seattle. “These projects are so large that it's like implementing multiple megaprojects that all depend on each other to complete successfully.”
One lesson U.S. National Railroad officials can learn from Brightline is to avoid lengthy planning stages and “focus on the costs of delays and indecision,” he said. Russell Jackson, Global Transportation Director at AECOM. He also suggested that while Brightline's approach focuses only on the most profitable routes, government funding could fill in the gaps in other cases.
“Public funds can be used for projects that are still needed to connect pairs of cities that are a little too close to travel by plane and too far to drive,” Jackson said.
I spend countless hours sifting through plastic pieces in my kitchen to determine if they can be recycled. If you have them, put them in a bag along with glass, cans, cardboard, and paper. If not, or if you’re not confident, put it in a plastic bag (not recyclable) and shove it in the cupboard under the stairs. I plan to drop it off in a non-recyclable plastic bin at my local supermarket. But the road to the landfill is paved with good intentions. Sometimes I get frustrated and throw it away.
I don’t know if my relentless culling will actually make any difference. We hope that what is recyclable will eventually be recycled. As for the others, which make up about half of my plastic waste, I don’t know their fate. I think there’s a reason it’s called “non-recyclable.”
We hope that you won’t have to waste your precious time on this kind of waste prioritization any time soon. A series of “advanced recycling” technologies are gradually being rolled out that promise to convert all kinds of used plastics into something very useful: plastic. The goal is to create a circular economy for this material by eliminating the need to make virgin plastic from crude oil and simply endlessly recycling what we already have. Plastic, once demonized as the scourge of modern society, could once again be great.
There are many things you can use. Since the 1950s, we have produced over 10 billion tons of her.
Child sexual exploitation is increasing online, with artificial intelligence generating new forms such as images and videos related to child sexual abuse.
Reports of online child abuse to NCMEC increased by more than 12% from the previous year to over 36.2 million in 2023, as announced in the organization’s annual CyberTipline report. Most reports were related to the distribution of child sexual abuse material (CSAM), including photos and videos. Online criminals are also enticing children to send nude images and videos for financial gain, with increased reports of blackmail and extortion.
NCMEC has reported instances where children and families have been targeted for financial gain through blackmail using AI-generated CSAM.
The center has received 4,700 reports of child sexual exploitation images and videos created by generative AI, although tracking in this category only began in 2023, according to a spokesperson.
NCMEC is alarmed by the growing trend of malicious actors using artificial intelligence to produce deepfaked sexually explicit images and videos based on real children’s photos, stating that it is devastating for the victims and their families.
The group emphasizes that AI-generated child abuse content hinders the identification of actual child victims and is illegal in the United States, where production of such material is a federal crime.
In 2023, CyberTipline received over 35.9 million reports of suspected CSAM incidents, with most uploads originating outside the US. There was also a significant rise in online solicitation reports and exploitation cases involving communication with children for sexual purposes or abduction.
Top platforms for cybertips included Facebook, Instagram, WhatsApp, Google, Snapchat, TikTok, and Twitter.
Out of 1,600 global companies registered for the CyberTip Reporting Program, 245 submitted reports to NCMEC, including US-based internet service providers required by law to report CSAM incidents to CyberTipline.
NCMEC highlights the importance of quality reports, as some automated reports may not be actionable without human involvement, potentially hindering law enforcement in detecting child abuse cases.
NCMEC’s report stresses the need for continued action by Congress and the tech community to address reporting issues.
We are witnessing the emergence of AI, a development that was unexpected by many. Let’s explore this further.
If you’ve interacted with an AI assistant extensively, you may have noticed a distinct quality in their responses. Unless deliberate efforts are made to deviate from the default settings, the text generated by AI can seem oddly produced, even if technically correct in terms of grammar and meaning.
Certain patterns are evident. Chatbots exhibit an exaggerated use of language models influenced by reinforcement learning with human feedback. The result? A willingness to please and a generally optimistic demeanor, desirable traits for an assistant role (or any role).
There are also areas where the AI hesitates to venture. Initiating a scenario like a romantic encounter between Mickey Mouse and Barack Obama can trigger various safety mechanisms, revealing the AI’s reluctance to engage in certain topics.
Other subtle indicators include a tendency to present both sides of an argument, an aversion to brief responses, and impeccable spelling and grammar, akin to “writing by a robot.”
Additionally, specific terms like “Learn more” recurrently used by ChatGPT raise flags on AI involvement, pointing towards the increasing integration of AI models in various fields for enhancing written content.
Dr. Jeremy Nguyen’s exploration suggests ChatGPT’s influence on certain PubMed articles. Photo: Jeremy Nguyen/X
A different dataset reveals that words like “Explore,” “Tapestry,” “Will,” and “Exploit” are more frequent in ChatGPT’s output compared to general internet usage.
Understanding the inner workings, GPT-4 serves as a vast language model processing extensive data to predict the next word in a sentence, a task requiring substantial data sets and sophisticated processing.
Transitioning this into a chatbot involves reinforcement learning with human feedback, engaging human testers to provide feedback for refining the AI’s responses, a resource-intensive process demonstrating the growing reliance on AI in content creation.
While AI advances in various sectors, the physical manifestation of AI technology is slowly emerging, as evidenced by Limitless’s innovative pendant for intelligent conversation tracking.
Looking ahead, AI hardware like Rabbit’s R1 proposes novel solutions for enhancing daily tasks through AI agents, emphasizing the evolving integration of AI in our lives.
As AI hardware gains traction, the future holds a vast landscape of possibilities, with ongoing advancements shaping the interconnected realms of AI technology and human interaction.
Numerous companies, including a national leisure center chain, are reassessing or discontinuing the use of facial recognition technology and fingerprint scanning for monitoring employee attendance in response to actions taken by Britain’s data authority.
The Information Commissioner’s Office (ICO) instructed a Serco subsidiary to halt the use of biometrics for tracking employee attendance at its leisure centers and prohibited the use of facial recognition and fingerprint scans. The ICO also issued stricter guidelines.
Following an investigation, the ICO found that more than 2,000 employees’ biometric data was unlawfully processed at 38 Serco-managed centers using facial recognition and, in two instances, fingerprint scanning to monitor attendance.
In response, Serco has been given a three-month deadline by the ICO to ensure compliance with regulations and has committed to achieving full compliance within that timeframe.
Other leisure center operators and businesses are also reevaluating or discontinuing the use of similar biometric technology for employee attendance monitoring in light of the ICO’s actions.
Virgin Active, a leisure club operator, announced the removal of biometric scanners from 32 properties and is exploring alternatives for staff monitoring.
Ian Hogg, CEO of Shopworks, a provider of biometric technology to Serco and other companies, highlighted the ICO’s role in assisting businesses in various industries to meet new standards for biometric authentication.
The new ICO standards emphasize exploring alternative options to biometrics for achieving statutory objectives, prompting companies to reconsider their use of such technology.
1Life, owned by Parkwood Leisure, is in the process of removing the Shopworks system from all sites, clarifying that it was not used for biometric purposes.
Continuing discussions with stakeholders, the ICO aims to guide appropriate use of facial recognition and biometric technology in compliance with regulations and best practices.
The widespread concerns raised by the ICO’s actions underscore the need for stronger regulations to protect employees from invasive surveillance technologies in the workplace.
The case of an Uber Eats driver facing issues with facial recognition checks highlights ongoing debates about the use of artificial intelligence in employment relationships and the need for transparent consultation processes.
Emphasizing the importance of respecting workers’ rights, the use of artificial intelligence in employment must be carefully regulated to prevent discriminatory practices and ensure fair treatment of employees.
○On April 6, Maryland passed the first “Kids Code” bill in the US. The bill is designed to protect children from predatory data collection and harmful design features by tech companies. Vermont’s final public hearing on the Kids Code bill took place on April 11th. This bill is part of a series of proposals to address the lack of federal regulations protecting minors online, making state legislatures a battleground. Some Silicon Valley tech companies are concerned that these restrictions could impact business and free speech.
These measures, known as the Age-Appropriate Design Code or Kids Code bill, require enhanced data protection for underage online users and a complete ban on social media for certain age groups. The bill unanimously passed both the Maryland House and Senate.
Nine states, including Maryland, Vermont, Minnesota, Hawaii, Illinois, South Carolina, New Mexico, and Nevada, have introduced bills to improve online safety for children. Minnesota’s bill advanced through a House committee in February.
During public hearings, lawmakers in various states accused tech company lobbyists of deception. Maryland’s bill faced opposition from tech companies who spent $250,000 lobbying against it without success.
Carl Szabo, from the tech industry group NetChoice, testified before the Maryland state Senate as a concerned parent. Lawmakers questioned his ties to the industry during the hearing.
Tech giants have been lobbying in multiple states to pass online safety laws. In Maryland, these companies spent over $243,000 in lobbying fees in 2023. Google, Amazon, and Apple were among the top spenders according to state disclosures.
The bill mandates tech companies to implement measures safeguarding children’s online experiences and assess the privacy implications of their data practices. Companies must also provide clear privacy settings and tools to help children and parents navigate online privacy rights and concerns.
Critics are concerned that the methods used by tech companies to determine children’s ages could lead to privacy violations.
Supporters argue that social media companies should not require identification uploads from users who already have their age information. NetChoice suggests digital literacy education and safety measures as alternatives.
During a discussion on child safety legislation, a NetChoice director emphasized parental control over regulation, citing low adoption rates of parental monitoring tools on platforms like Snapchat and Discord.
NetChoice has proposed bipartisan legislation to enhance child safety online, emphasizing police resources for combating child exploitation. Critics argue that tech companies should be more proactive in ensuring child safety instead of relying solely on parents and children.
Opposition from tech companies has been significant in all state bills, with representatives accused of hiding their affiliations during public hearings on child safety legislation.
State bills are being revised based on lessons learned from California, where similar legislation faced legal challenges and opposition from companies like NetChoice. While some tech companies emphasize parental control and education, critics argue for more accountability from these companies in ensuring child safety online.
Recent scrutiny of Meta products for their negative impact on children’s well-being has raised concerns about the company’s role in online safety. Some industry experts believe that tech companies like Meta should be more transparent and proactive in protecting children online.
I
In the early 2000s, Olivia Packenham came home from school, heard the familiar dial-up tone of her family’s computer connecting to the Internet, and was transported to the virtual game world of Neopets in her AOL browser.
Packenham, who started playing at the age of 8, played for many years before losing interest in high school. But in December 2023, after almost her 15-year hiatus, she logged back into her neopets.com. Then she realized that her childhood pets were waiting for her. Her favorite “Bruce” (the Neopets version of Penguin) is now over 21 years old.
Packenham says that when she opens Neopets today, it’s like going back to that time in her life. This time, my mother didn’t yell at me because the family’s phone line was clogged.
“It was like walking into a museum of the early 2000s Internet,” said Packenham, 32. And he is one of the thousands of players who have returned to Neopets over the past year.
Packenham’s return was no coincidence. His Neopets, which announced a bold rebrand in 2023, is consciously courting former fans with the promise of reuniting with old digital friends who have remained largely unchanged.
Returning Neopians (user community name) cites several reasons for returning. A woman who went by the username Solabee said she started playing the game at age 9 and began playing almost every day after her return. The resurgence of Neopets felt very special to millennials, she said. In fact, 40% of her users are between the ages of 25 and 34, and 26% of her users are between the ages of 18 and 24, the next largest demographic. Revisiting this site reminds me of a hopeful time in Internet history, when logging online was still new and exciting. Many of the Neopians interviewed for this article prefer to keep their online personas separate from their real lives and asked to be quoted by their usernames.
“We are the most nostalgic generation, so when major global events or crises occur, we all want to go back to the safest time in our lives, which is our childhood,” Soraby said.
Olivia Packenham's Neopets, the penguin-like "Bruce" variety. Photo: Courtesy of Olivia Packenham.
Neopets' strong start and long decline
Neopets started in 1999 as an early internet universe where users could care for a variety of virtual pets and play mini-games. A crude social network that predates Facebook, the site allowed users to add friends, send each other messages, exchange resources and virtual currency, and battle. It peaked in the mid-2000s with more than 25 million active users, but its popularity quickly declined as competing gaming and social sites exploded on the Internet.
The platform has changed hands multiple times since its founding, slumped amid acquisitions, and had just 100,000 users when it was acquired by Chinese company NetDragon in 2017. Further deepening the decline was the decline of Adobe's Flash Player (software). Powered most of the site – Browsers began to be phased out Neopets started around 2017 and was officially discontinued in 2020. Neopets received little attention and the interface was not updated.
That's because entrepreneur and investment consultant Dominic Lo, who joined NetDragon as new markets director in 2020, launched an internal campaign to revive the Neopets brand, calling it a "giant leap of faith." ” until he called it. Ms Lo, 36, recalled that after she immigrated to Hong Kong as a child, she used Neopets to keep in touch with friends in Canada. He realized that his Neopets was at risk because NetDragon had shut down several underperforming sites it had acquired.
“Despite its decline over the past decade, when lack of updates and flash outages made half the site unplayable, a quarter of the remaining users log on every day,” he said. “Seeing how close-knit the rest of the community is, and as a player myself, I resonated with the emotional attachment that drove these core users to support Neopets. Perhaps this is why We’ve found that we have probably the most sticky customer base.”
Law persuaded NetDragon's upper management to give Neopets a "final blow" to save it, spinning Neopets into an independently owned company with a management team of which he is now CEO. We finalized the acquisition agreement. The move was backed by undisclosed external investment. Under a group known as Neopets Team (TNT), the company began restoring the site's functionality, updating its design, and improving its most famous classic games. Next, TNT wants to enable more mobile capabilities. Although much of the site is still bogged down with some glitches, Roe said that its failure to change with the times has ironically been an asset.
“With no updates, we basically preserved early internet culture in its purest form,” he said. “When a user logs on, the game looks 99% the way he did, and his pet is there. It's like reuniting with an old friend. There aren't many experiences in life that allow you to relive your childhood. there is no.”
Difficulties in reviving Neopia
Efforts to revive Neopets have intensified, with monthly user numbers nearly tripling to 300,000 in the past six months, and the company on track to be profitable by the end of 2024, Lo said. .
“We believe this is the beginning of recovery. But there is still much work to be done,” Lo said.
In addition to prioritizing bringing lapsed users back to the site, Neopets is seeking to license the intellectual property for its trading card game and branded Monopoly board.
“We're already on track to be profitable, but we want to make sure it's sustainable and future-proof our intellectual property so we can survive for years to come.” Mr Law said.
As pressure continues to make Neopets financially viable in the real world, the company's new leadership will also have to contend with the digital financial component of the site, which operates on the Neopoints cryptocurrency. With few updates to the site over the years, a huge black market has emerged for specific pets, and the site's virtual economy has suffered from hyperinflation.
Neopets' original logo and creatures. Photo: Neopets
Neopets runs on a complex economy consisting of two main components: items and Neopoints. Throughout the years
“circleWait a minute, wait a minute. You haven’t heard anything yet.” So was the first line of dialogue heard in the 1927 feature film jazz singer. This was the first time that the mass media conveyed the sights and sounds of the scene together, and the audience was mesmerized.
Since then, black and white has given way to color, frame rates and resolutions have increased, and sound quality has improved, but the media we consume still remains overwhelmingly, if not exclusively, our eyes and ears. We are responding to
The average person now spends nearly seven hours a day watching screens, and with most of that time spent indoors, our overreliance on sight and sound is only increasing. But if a human considers that he is a five (or five) animal, probably even more) senses, aren’t we ignoring other abilities? And what is it doing to us?
Many psychologists classify our primary senses as either rational or emotional, and there is evidence to support this. “Odor [and taste are] Charles Spence, professor of experimental psychology at the University of Oxford, says, “Rational senses such as hearing and vision are directly connected to emotional processing areas of the brain.” In fact, Spence says more than half of the neocortex, and therefore more than half of the brain’s volume, is devoted to processing what we see.
There’s no denying that we are highly visual creatures, which is part of the reason why our media is primarily audiovisual. “I think this is largely due to the fact that much of the information we consider important today is conveyed through visual and auditory means,” said Meike Scherer, an assistant professor in the Department of Psychology at Durham University. “But what we think is important isn’t necessarily what we need.”
If you ask people which sense they can’t live without, most people will say sight, but evidence shows that what we really lack is smell. “The rates of suicide and suicidal ideation are much higher among people with anosmia, because anosmia is so tied to our emotions,” Scherer says.
So does ignoring some senses in favor of others affect our emotional lives? Our emotional health is tied to our social health, but… The answer is almost certainly yes. “Smell is a very important cue for social communication, but this is something that is not implemented in any of the technologies we use today,” Scherer says.
For example, it has been found that after shaking someone’s hand, we tend to subconsciously smell their palm. “It gives you hints about all sorts of things, from their health to their age and even their personality,” Spence says. “A lot is lost when we only interact digitally.”
Touch is equally important to our emotional lives, and the finger-focused haptics of digital devices are not enough. C-tactile afferents are a type of nerve receptor that is abundant in the hairy skin of the arms (but not on the pads of the fingers) and has been shown to produce positive emotions when stimulated. “These receptors like slow, warm, tactile strokes,” says Spence.
The cool and smooth touch screen of your smartphone cannot replace other human skin, which is soft, warm and imperceptibly smelly. For adults, this may mean less satisfaction with their social lives, but for a generation of children who are increasingly socialized through technology, the effects can be profound.
Scherer says children learn to interpret their own senses by referring to each other’s senses. We learn to associate subtle smells with the sound of someone yelling or the sight of a smile, and may learn to use these signals to navigate social situations in the future. “Children who grow up with less input basically have less training to be able to categorize what certain things smell like and what certain exposures mean,” Scherer said. To tell. “If you suddenly take away something that has evolved over millions of years, you’re not only removing one sense from her, but it’s affecting how all of her other senses work.”
Marianna Obrist, Professor of Multisensory Interfaces at University College London, said: Everything is multisensory.
For example, it’s easy to think that the experience of eating is primarily about taste, but the shape and color, smell and sizzle, temperature, texture and weight of food are influenced by our senses of sight, smell, hearing and touch. appeal to. “All these senses are already activated before you eat,” says Obrist. Then there’s mouthfeel, the physical sensation of spiciness and sourness, and of course, flavor.
Removing just one of those sensations can affect the entire experience. For example, if you eat ice cream in the dark, It is unlikely that you will enjoy it, or even be sure of what it tastes like. “Each time we receive multisensory stimulation, we are able to develop a better and richer representation of our surroundings,” Scherer says.
So What are we doing to make our technology more multisensory? sense x, an EU-funded project aimed at helping designers come up with new ways to integrate feel, smell and taste into products. The team’s efforts included spraying scents under subjects’ noses to highlight key moments in director Christopher Nolan’s film interstellar, irradiate ultrasound to simulate contact, Powerful acoustics to suspend food It can be attached to the tongue without the need for wires or tubes.
It’s hard to imagine I’ll be watching it any time soon. Colonel Kilgore’s speech by Robert Duvallapocalypse of hellThe most famous line, while the smell of eau na palm hits your nose from your laptop in the morning, the smell-taste interface may be just around the corner. Researchers are already using AI to try to find the primary odor that creates any odor, and Obrist hopes to create a digitally controlled system with applications in research, healthcare, and immersive reality experiences. I’m the chief scientific officer at OWidgets, a company that makes scent delivery systems.
Almost all the input we receive from electronic devices is visual or auditory, so it is processed by the cortex, the rational part of the brain. Photo: Alex Segre/Alamy
Companies like China’s Dexta Robotics are also bringing tactility to virtual reality with gloves called “gloves.” dexmo.
“Dexmo can provide haptic and force feedback simultaneously,” said Aler Gu, CEO of Dexta. “So when you scroll your finger over a virtual brick, you can feel the surface texture. When you grab a brick and move it from one point to another, you can feel its physical shape.”
Media that engage all of our senses will certainly enrich our daily interactions with technology, but it’s not hard to imagine more insidious uses emerging. In 1957, an American market researcher named James Vicary claimed to have created a movie by splicing together the scenes “Eating Popcorn” and “Drinking Coca-Cola.” He reported that sales of popcorn and Coca-Cola increased by 57.5% and 18.1% respectively, and the concept of subliminal advertising was born.
Vicary was later exposed as a fraudster, and the effectiveness of subliminal advertising has gained worldwide attention. discussion issues Since then, has technology that can deliver smells and tastes digitally become a gift to unscrupulous advertisers? Masu. [these senses]. They can be very powerful,” says Scherer. “We’re very emotional decision-makers, so there’s a lot of potential for that to influence our decisions.”
Research has shown that exposure to certain tastes and smells can influence our judgments of other people’s appearance and personality, and even change our behavior.For example, taste bitter foods can make us hostile,and 2005 patent application The scent of pink grapefruit suggests to men that it can make women appear younger than their actual age.
Obrist’s team discovered that: Sour taste makes people more willing to engage in risky behavior. “You might be doing electronic banking or shopping online and drinking a sour lemon drink. That may indirectly influence your decision-making,” she says. say. It’s not hard to imagine how e-commerce and gambling apps will be affected. Devices that can deliver tastes and smells can be exploited.
To some extent, this is already happening.Companies are known for pumping pleasant scents into their stores, and American chain Cinnabon Intentionally place the oven near the store entrancesometimes creating baking trays with just sugar and cinnamon to tempt passing shoppers.
From T-shirts with changing messages to carpets that can detect your position, the future of smart textiles seems to come straight out of a sci-fi novel.
Researchers now claim they have created a smart fiber that can achieve just that, without the need for a battery pack.
A team of Chinese researchers have developed textile-based electronics that utilize the human body as part of a circuit to harness electromagnetic energy from the environment.
This innovation could pave the way for a “body-bound” fiber electronics technology that functions without electronic chips or batteries and could be applied in various scenarios.
Co-author Chengyi Hou from Donghua University in Shanghai explained, “When electromagnetic energy passes through a fiber, it is converted into different forms of energy, including visible light or radio waves. Therefore, the fiber not only emits light but also produces an electrical signal when in contact with the human body.”
Hou highlighted that these radio signals are programmable by manipulating different aspects of the system, such as the fiber’s contact area with the body and its diameter.
The team stated that this method resolves a major challenge in integrating electronic systems into textiles, which is the necessity of rigid components.
Hou mentioned, “We have successfully achieved mass production of this new type of fiber electronics, which is as thin and soft as traditional fibers. The next step is to implement it.”
The team has created prototypes like a wearable cloth display with a cloth keyboard, intended for individuals with hearing impairments to aid in communication, as well as textile controllers for gaming.
Additionally, they developed a wireless tactile carpet that illuminates underfoot, providing emergency lighting at night and wirelessly transmitting signals to control household devices like lights.
Researchers have created a carpet that can glow underfoot and transmit signals that can be used to control switches in appliances such as lights. Photo: Yang Weifeng
Read more about the study here. The team assures that the fiber is constructed from three layers of inexpensive materials, making it durable, washable, and sweat-resistant.
An accompanying article suggests that this technology can also be utilized in robots, robotic prosthetics, and capturing haptic information to enhance human interactions and object recognition.
Dr. Luigi Occhipinti, a research director at the University of Cambridge specializing in smart electronics, biosystems, and AI, acknowledged the potential of this approach.
He stated, “By being constantly surrounded by various electromagnetic fields, we are developing innovative electronic textiles with skin sensors and unconventional electronics, powered uniquely through energy harvesting. This has the potential to unlock a new realm of self-powered wearable electronics for continuous health monitoring.”
Following a dry run of Taiwan’s presidential election this year, China is anticipated to disrupt elections in the United States, South Korea, and India with artificial intelligence-generated content, as warned by Microsoft.
The tech giant predicts that Chinese state-backed cyber groups will target high-profile elections in 2024, with North Korea also getting involved, according to a report released by the company’s threat intelligence team.
“As voters in India, South Korea, and the United States participate in elections, Chinese cyber and influence actors, along with North Korean cyber attack groups, are expected to influence these elections,” Microsoft mentioned.
Microsoft stated that China will create and distribute AI-generated content through social media to benefit positions in high-profile elections.
Although the immediate impact of AI-generated content seems low in swaying audiences, China is increasingly experimenting with enhancing memes, videos, and audio, potentially being effective in the future.
During Taiwan’s presidential election in January, China attempted an AI-powered disinformation campaign for the first time to influence a foreign election, Microsoft reported.
The Beijing-backed group Storm 1376, also known as Spamoflage or Dragonbridge, heavily influenced Taiwan’s elections with AI-generated content spreading false information about candidates.
Chinese groups are also engaged in influencing operations in the United States, with Chinese government-backed actors using social media to probe divisive issues among American voters.
In a blog post, Microsoft stated, “This may be to collect intelligence and obtain accurate information on key voting demographics ahead of the US presidential election.”
The report coincides with a White House board’s announcement of a Chinese cyber operator infiltrating US officials’ email accounts due to errors made by Microsoft, as well as accusations of Chinese-backed hackers conducting cyberattacks targeting various entities in the US and UK.
Please try to imagine. I remember being on the bus or walking in the park and having an important task slip out of my mind. Maybe you were planning to send an email, catch up on a meeting, or have lunch with a friend. Without missing a beat, just say out loud what you forgot, and a small device strapped to your chest or placed on the bridge of your nose will send you a message, summarize a meeting, or remind a friend to go to lunch. Send invitations. No need to poke at your smartphone screen and the job is done.
This is the kind of utopian convenience that a growing wave of technology companies are trying to achieve through artificial intelligence. Generative AI chatbots such as ChatGPT exploded in popularity last year as search engines like Google, messaging apps like Slack, and social media services like Snapchat raced to integrate the technology into their systems.
AI add-ons have become commonplace across apps and software, but as the first AI-powered consumer devices launch and compete for space with smartphones, the same generation technology is poised to enter the hardware realm.
One of the first people out of the gate was eye pin California startup Humane. It’s a wearable device that’s only slightly larger than a can of Vaseline and attaches to your shirt via a magnet. You can send texts, make calls, take photos, and play music. However, there is no app support and no screen. Instead, it uses a laser to project a simple interface onto your outstretched palm. The built-in AI chatbot can be instructed through voice commands to search the web or answer queries in much the same way you’d expect from ChatGPT.
“I plan to train Ai Pin to be my personal assistant to facilitate my writing and creative work,” said the Virginia-based company, which pre-ordered the device ahead of its initial U.S. launch in April, says Tiffany Jana, a consultant with Since she travels a lot, she thinks it would be nice to have a photographer and translator to accompany her. “I don’t have all the assistants and large teams that supported me in the past. I’ve always been a tech guy and enjoy ChatGPT.”
Meanwhile, Facebook’s parent company Meta has already Smart glasses equipped with AI Partnering with Ray-Ban and Chinese companies TCL and Oppo Companies followed suit with their own AI glasses. All of these have pretty much the same functionality as Ai Pin and are sold in a way that connects to an AI chatbot that responds to voice commands.
It’s a way to curb smartphone overuse by providing the same essential functionality without addictive apps.
If all of this sounds a lot like what your smartphone’s voice assistant or your living room’s Alexa already does, that’s because that’s essentially what it does. “Using AI in new devices is still the norm today,” says David Lindlbauer, an assistant professor at the Human-Computer Interaction Institute at Carnegie Mellon University in Pennsylvania. “Everyone uses Google Suggestions, Apple Siri to navigate their phones, or smart suggestions in apps on their phones.” The difference, he says, is that these new and upcoming devices will, which aims to embed AI capabilities in a “less obtrusive and more ubiquitous way.”
Its design intent is most evident in future products pendant From US startup Rewind and software developer Tab AI Avi Shiffman. These small devices hang around your neck and passively record everything you hear and say during the day, then transcribe the most important parts so you can read them back at your convenience later. Designed to summarize. These are essentially productivity tools that bundle the kinds of generative AI capabilities found elsewhere into standalone devices.
But why would you want a device that does more than what your smartphone already has? Partly to free yourself from the less-than-welcome elements. Humane is pitching Ai Pin as a way to curb smartphone overuse by offering the same important functionality without the addictive apps that make you scroll compulsively. “An alcoholic is not dependent on the bottle, but on the contents,” says Christian Montag, chair of molecular psychology at the University of Ulm in Germany, by analogy. He says social media platforms in particular are often interested in intentionally extending screen time in order to show more ads or collect personal data. say. Experiments show that when you use your smartphone in grayscale mode, Reduce user retentionremoving the screen completely can have even more severe effects.
While this may seem counterintuitive to the tech industry’s ever-increasing appetite for new features and gadgets, it’s probably not as alien as it first seems. “Many people wear headphones all day long,” says Lindlbauer. “Therefore, it is entirely possible to move away from the temptation of scrolling through doom and move towards technology that allows us to access the digital world constantly, but unobtrusively.”
However, discussions about their broader applications are beginning to take place. For some, the future of this technology lies not in how it can be integrated into existing platforms, but in whether it can fundamentally change the way platforms are accessed. “There will be no need to use different apps for different tasks,” former Microsoft CEO Bill Gates said in an article. Blog post outlining his vision. “Simply tell your device, in everyday language, what you want to do.” Then, leave it to your device to figure out what apps, platforms, and information it needs to complete the task you set.
This is an idea that will be put into preliminary practice in the next stage. R1. Developed by Rabbit, a Californian AI startup, the R1 is a handheld device that looks a bit like a portable gaming console and operates like a powerful voice assistant. However, it is designed to interact directly with an app on your phone on your behalf, rather than simply connecting to an AI chatbot that generates passive responses to your commands (like other wearable gadgets). Masu. The idea is that R1 acts as an all-in-one interface for your device, a kind of central app that can control everything else.
“We’re not building products for new use cases. We’re developing better, more intuitive ways to address existing use cases.” said Jesse Lyu, Chief Executive Officer of Rabbit. He describes the R1 as a “digital companion” that doesn’t replace your smartphone, but makes it easier to use.
The value of that approach will become clear when R1 launches later this year. However, similar experimental devices are expected to follow. Sam Altman, CEO of OpenAI, which developed ChatGPT, is reportedly already in talks with former Apple chief designer Jony Ive to explore hardware ideas. And a group of startups and Silicon Valley powerhouses are now racing to develop the chips and processors these new devices will need to power their AI models.
Whatever form these AI devices end up taking, they will be hard to compete with the globally connected, highly capable, and intuitively controlled glass rectangles that are in most of our pockets. You’re going to have to work. However, as ubiquitous as smartphones seem, they too have an expiration date. “Smartphones have only been with us for about 15 years,” says Lindlbauer. “I don’t want to believe that smartphones are the pinnacle of technology or that we’ll ever use them the way we do now. [another] 15 years.”
Officials in New York City revealed a pilot program on Thursday to implement handheld gun scanners in the subway system to enhance safety and reduce violence underground.
Mayor Eric Adams mentioned that the scanners will be set up at specific stations after a 90-day waiting period mandated by law.
“Ensuring the safety of New Yorkers in the subway system and preserving their trust in the system is crucial for keeping New York the safest metropolis in America,” Adams stated. The announcement also included plans to deploy extra outreach personnel to assist individuals with mental health issues living in the system.
Adams mentioned that authorities will seek companies with expertise in weapons detection technology, and eventually install the scanners in select subway stations to assess their effectiveness further.
The scanner, showcased by Mr. Adams and law enforcement officials at a news conference in Lower Manhattan, was developed by Evolv, a publicly traded company facing allegations of manipulating software test results to exaggerate the scanner’s effectiveness. The company is currently under investigation by U.S. trade regulators and financial regulators.
Evolv’s CEO, Peter George, described the AI-enabled scanner as utilizing “a secure ultra-low frequency electromagnetic field and advanced sensors for concealed weapons detection.”
Jerome Greco, overseeing attorney for the Legal Aid Society’s digital forensics division, cautioned that gun detection systems may trigger false alarms and cause unnecessary panic.
City officials have not disclosed the specific locations where the scanners will be deployed. A demonstration at the Fulton Street station showed the device beeping when an officer with a holstered gun passed, but not reacting to an officer with a cell phone or other electronic device. No false alarms were noted.
While violent incidents in the city’s subways are infrequent, recent high-profile shootings have highlighted safety concerns. The city recorded five murders in the subway system last year, a decrease from the previous year. The installation of the scanners follows a recent fatal accident at an East Harlem subway station, reinforcing the urgency of subway safety measures.
Modern conveniences like smart refrigerators, virtual assistants, security systems, and robot vacuums have made household chores easier since the invention of the washing machine in the 19th century.
Innovative products like the EZVIZ RE4 Plus robot vacuum and mop combo can help eliminate mundane daily tasks. Users can conveniently schedule cleaning routines, saving time and effort with just a few taps on the app.
Discover more about the cutting-edge technology behind robot vacuums and how they can streamline your daily life.
Smart technology for the home
Navigation
LiDAR (light detection and ranging) technology, originally developed in the 1960s for mapping and surveillance purposes, has evolved into a crucial component of autonomous vehicles and even some smartphones. LiDAR’s ability to create detailed spatial data is shaping the future of robotics technology.
The state-of-the-art EZVIZ RE4 Plus incorporates LDS LiDAR technology, enabling 360-degree scanning and mapping of your home’s layout with the capability to store multiple floor plans.
Smart Thinking
Premium robot vacuums like the EZVIZ RE4 Plus utilize advanced features to navigate and clean effectively based on environment data and user instructions.
App Remote Control and Map Customization
Through IoT connectivity, users can remotely schedule cleaning tasks and customize maps to suit their cleaning preferences. Smart robot vacuums like the EZVIZ RE4 Plus can adapt to different cleaning needs and scenarios.
Obstacle Avoidance and Fall Detection Technology
Premium robot vacuums are equipped with infrared obstacle avoidance sensors and fall detection technology to enhance safety and efficiency during cleaning.
Auto-Empty Feature
The EZVIZ RE4 Plus comes with an auto-empty base that disposes of dust and debris, reducing the need for frequent emptying. The dust bag typically requires emptying every 90 days for regular household use.
Why You Need a Robot Vacuum Cleaner
Modern smart devices like robot vacuums offer effortless, time-saving solutions for household cleaning tasks, making daily life more convenient and efficient for busy individuals and families.
Discover how the EZVIZ RE4 Plus robot vacuum can revolutionize your cleaning routine.
Effortless cleaning: compact, efficient, and obedient
Save time and effort: easy app control for quick cleaning
Convenience: automated cleaning even when you’re away
Flexibility: customizable cleaning schedules to fit your needs
Allergy-friendly: minimal bag changes every 3 months
Sustainability: reduce waste and costs with reusable materials
This website uses cookies so that we can provide you with the best user experience possible. Cookie information is stored in your browser and performs functions such as recognising you when you return to our website and helping our team to understand which sections of the website you find most interesting and useful.
Strictly Necessary Cookies
Strictly Necessary Cookie should be enabled at all times so that we can save your preferences for cookie settings.
