Tag Archives: DeepMind

How Google’s DeepMind Tool Accelerates Hurricane Behavior Predictions

Posted on by
Reply

As Tropical Storm Melissa wreaked havoc south of Haiti, meteorologist Philippe Papin from the National Hurricane Center (NHC) firmly believed it was on the verge of evolving into a formidable hurricane.

In his capacity as the lead forecaster, he forecasted that within a mere 24 hours, the storm would escalate to a Category 4 hurricane and shift its course toward Jamaica’s coastline. Up to that point, no NHC forecaster had made such an announcement. What a daring prediction for quick validation.

However, Mr. Papin had an ace up his sleeve: artificial intelligence, specifically Google’s newly released DeepMind hurricane model from June. As expected, Melissa transformed into an unbelievably strong storm that devastated Jamaica.

NHC forecasters are increasingly depending on Google DeepMind. On the morning of October 25th, Mr. Papin elaborated on this in a public forum. He also shared on social media that Google’s model was central to his confidence: “Approximately 40 out of 50 members of the Google DeepMind ensemble predict Melissa will reach Category 5. While we are cautious about predicting its intensity due to track uncertainty, it remains a strong possibility.”


“Rapid intensification is likely as the storm traverses very warm ocean waters, characterized by the highest ocean heat content in the entire Atlantic Basin.”

Google DeepMind’s first AI model specifically designed for hurricanes has now surpassed traditional weather forecasters at their own game. It has accurately predicted all 13 Atlantic storms so far this year, outperforming human forecasters in course prediction.

Ultimately, Melissa made landfall in Jamaica as a Category 5 hurricane, marking one of the most powerful landfalls recorded in nearly two centuries across the Atlantic. Mr. Papin’s audacious forecasts could provide Jamaicans with critical time to brace for disasters, potentially safeguarding lives and property.

Google DeepMind is revolutionizing weather forecasts in recent years, and the parent forecasting system that the new hurricane model is based on has also excelled in identifying last year’s large-scale weather patterns.

Google’s models function by discovering patterns that traditional, slower, physics-based weather models may overlook.

“They operate much faster than their physics-based counterparts, with increased computational efficiency that saves both time and resources,” remarked former NHC forecaster Michael Rowley.

“This hurricane season has demonstrated that emerging AI weather models can be competitive, and in some instances, more accurate than the slower, traditional physics-based models that have long been our standard,” Rowley noted.

It’s important to note that Google DeepMind exemplifies machine learning—not generative AI like ChatGPT. Machine learning processes large data sets to identify patterns, allowing models to generate answers in minutes using standard computing resources. This stands in stark contrast to the flagship models employed by governments for decades, which take hours to compute using some of the world’s largest supercomputers.

Nevertheless, the fact that Google’s model has quickly surpassed traditional models is nothing short of remarkable for a meteorologist devoted to forecasting the planet’s most powerful storms.

Skip past newsletter promotions

Former NHC forecaster James Franklin expressed his admiration: “The sample size is now significant enough to conclude this isn’t merely beginner’s luck.”

Franklin indicated that Google DeepMind has eclipsed all other models in tracking hurricane paths globally this year. As with many AI models, high-end intensity predictions can sometimes miss the mark. Earlier this year, Hurricane Erin rapidly intensified to Category 5 in the northern Caribbean, while Typhoon Karmaegi struck the Philippines on a recent Monday.

Looking ahead, Franklin mentioned his intention to engage with Google during the upcoming offseason to enhance DeepMind’s output by providing additional internal data for better assessment of its predictions.

“What concerns me is that while these predictions appear very accurate, the model’s output operates like a black box,” Franklin remarked.

No private or commercial entity has ever developed a leading weather model that allows researchers to scrutinize its methods. Unlike the majority of models built and maintained by the government, which are available to the public at no cost, Google has established high-level resources for DeepMind; published in real-time on a dedicated website, though its methodologies largely remain concealed.

Google is not alone in harnessing AI for challenging weather forecasting issues. Governments in the US and Europe are also working on their own AI weather models, demonstrating enhanced capabilities compared to previous non-AI versions.

The next frontier in AI weather forecasting seems to be for startups to address sub-seasonal forecasts and challenges that have so far proven difficult. To enhance advance warning of tornado outbreaks and flash floods—a goal supported by US government funding. Additionally, a company named WindBorne Systems is launching weather balloons to bridge gaps in the U.S. weather observation network, recently diminished by the Trump administration.

Source: www.theguardian.com

DeepMind and OpenAI Achieve Victory in the International Mathematics Olympiad

Posted on by
Reply

AIs are improving at solving mathematics challenges

Andresr/ Getty Images

AI models developed by Google DeepMind and OpenAI have achieved exceptional performance at the International Mathematics Olympiad (IMO).

While companies herald this as a significant advancement for AIs that might one day tackle complex scientific or mathematical challenges, mathematicians urge caution, as the specifics of the models and their methodologies remain confidential.

The IMO is one of the most respected contests for young mathematicians, often viewed by AI researchers as a critical test of mathematical reasoning, an area where AI traditionally struggles.

Following last year’s competition in Bath, UK, Google investigated how its AI systems, Alpha Proof and Alpha Jometry, achieved silver-level performance, though their submissions were not evaluated by the official competition judges.

Various companies, including Google, Huawei, and TikTok’s parent company, approached the IMO organizers requesting formal evaluation of their AI models during this year’s contest, as stated by Gregor Drinner, the President of IMO. The IMO consented, stipulating that results should be revealed only after the full closing ceremony on July 28th.

OpenAI also expressed interest in participating in the competition but did not respond or register upon being informed of the official procedures, according to Dolinar.

On July 19th, OpenAI announced the development of a new AI that achieved a gold medal score alongside three former IMO medalists, separately from the official competition. OpenAI stated the AI correctly answered five out of six questions within the same 4.5-hour time limit as human competitors.

Two days later, Google DeepMind revealed that its AI system, Gemini Deep Think, had also achieved gold-level performance within the same constraints. Dolinar confirmed that this result was validated by the official IMO judges.

Unlike Google’s Alpha Proof and Alpha Jometry, which were designed for competition, Gemini Deep Think was specifically crafted to tackle questions posed in a programming language used by both Google and OpenAI.

Utilizing LEAN, the AI was capable of quickly verifying correctness, although the output is challenging for non-experts to interpret. Thang Luong from Google indicated that a natural language approach can yield more comprehensible results while remaining applicable to broadly useful AI frameworks.

Luong noted that advancements in reinforcement learning—a training technique designed to guide AI through success and failure—have enabled large language models to validate solutions efficiently, a method essential to Google’s earlier achievements with gameplay AIs, such as AlphaZero.

Google’s model employs a technique known as parallel thinking, considering multiple solutions simultaneously. The training data comprises mathematical problems particularly relevant to the IMO.

OpenAI has disclosed few specifics regarding their system, only mentioning that it incorporates augmented learning and “experimental research methods.”

“While progress appears promising, it lacks rigorous scientific validation, making it difficult to assess at this point,” remarked Terence Tao from UCLA. “We anticipate that the participating companies will publish papers featuring more comprehensive data, allowing others to access the model and replicate its findings. However, for now, we must rely on the companies’ claims regarding their results.”

Geordy Williamson from the University of Sydney shared this sentiment, stating, “It’s remarkable to see advancements in this area, yet it’s frustrating how little in-depth information is available from inside these companies.”

Natural language systems might be beneficial for individuals without a mathematical background, but they also risk presenting complications if models produce lengthy proofs that are hard to verify, warned Joseph Myers, a co-organizer of this year’s IMO. “If AIs generate solutions to significant unsolved questions that seem plausible yet contain subtle, critical errors, we must be cautious before putting confidence in lengthy AI outputs.”

The companies plan to initially provide these systems for testing by mathematicians in the forthcoming months before making broader public releases. The models claim they could potentially offer rapid solutions for challenging problems in scientific research, as stated by June Hyuk Jeong from Google, who contributed to Gemini Deep Think. “There are numerous unresolved challenges within reach,” he noted.

Topics:

Source: www.newscientist.com

AI from DeepMind outperforms current weather predictions in accuracy

Posted on by
Reply

Weather forecasting today relies on simulations that require large amounts of computing power.

Petrovich9/Getty Images/iStockphoto

Google DeepMind claims its latest weather forecasting AI can predict faster and more accurately than existing physics-based simulations.

GenCast is the latest in DeepMind's ongoing research project to improve weather forecasts using artificial intelligence. The model was trained on 40 years of historical data from the European Center for Medium-Range Weather Forecasts (ECMWF). ERA5 ArchiveThis includes regular measurements of temperature, wind speed, and barometric pressure at various altitudes around the world.

Data up to 2018 was used to train the model, and then 2019 data was used to test predictions against known weather conditions. The company found that it outperformed ECMWF's industry standard ENS forecasts 97.4% of the time, and 99.8% of the time when forecasting more than 36 hours ahead.

Last year, DeepMind collaborated with ECMWF to create an AI that outperformed the “gold standard” high-resolution HRES 10-day forecast by more than 90%. Previously, he developed a “nowcasting” model that used five minutes of radar data to predict the probability of rain over a given one square kilometer area from five to 90 minutes in advance. Google is also working on ways to use AI to replace small parts of deterministic models to speed up calculations while maintaining accuracy.

Existing weather forecasts are based on physical simulations run on powerful supercomputers to deterministically model and estimate weather patterns as accurately as possible. Forecasters typically run dozens of simulations with slightly different inputs in groups called ensembles to better capture the variety of possible outcomes. These increasingly complex and large numbers of simulations are computationally intensive and require ever more powerful and energy-consuming machines to operate.

AI has the potential to provide lower-cost solutions. For example, GenCast uses an ensemble of 50 possible futures to create predictions. Using custom-built, AI-focused Google Cloud TPU v5 chips, each prediction takes just 8 minutes.

GenCast operates at a cell resolution of approximately 28 square kilometers near the equator. Since the data used in this study were collected, ECMWF's ENS has been upgraded to a resolution of just 9 kilometers.

Yilan price DeepMind says AI doesn't have to follow, and could provide a way forward without collecting more detailed data or performing more intensive calculations. “If you have a traditional physics-based model, that's a necessary requirement to solve the physical equations more accurately, and therefore to get more accurate predictions,” Price says. “[With] machine learning, [it] It is not always necessary to go to higher resolution to get more accurate simulations and predictions from your model. ”

david schultz Researchers at the University of Manchester in the UK say AI models offer an opportunity to make weather forecasts more efficient, but they are often over-hyped and rely heavily on training data from traditional physically-based models. states that it is important to remember that

“is that so [GenCast] Will it revolutionize numerical weather forecasting? No, because in order to train a model, you first have to run a numerical weather prediction model,” says Schulz. “These AI tools wouldn't exist if ECMWF didn't exist in the first place and without creating the ERA5 reanalysis and all the investment that went into it. It's like, 'I can beat Garry Kasparov at chess. But only after studying every move he's ever played.''

Sergey Frolov Researchers at the National Oceanic and Atmospheric Administration (NOAA) believe that further advances in AI will require higher-resolution training data. “What we're basically seeing is that all of these approaches are being thwarted.” [from advancing] “It depends on the fidelity of the training data,” he says. “And the training data comes from operational centers like ECMWF and NOAA. To move this field forward, we need to generate more training data using higher-fidelity physically-based models. .”

But for now, GenCast offers a faster way to perform predictions at lower computational costs. kieran hunt A professor at the University of Reading in the UK believes ensembles can improve the accuracy of AI predictions, just as a collection of physics-based predictions can produce better results than a single prediction. states.

Mr Hunt points to the UK's record temperature of 40C (104C) in 2022 as an example. A week or two ago, there was only one member of the ensemble who was predicting it, and they were considered an anomaly. Then, as the heat wave approached, the predictions became more accurate, providing early warning that something unusual was about to happen.

“You can get away with it a little bit if you have one member who shows something really extreme. That might happen, but it probably won't happen,” Hunt says. “I don’t think it’s necessarily a step change; it’s a combination of new AI approaches with tools we’ve been using in weather forecasting for a while to ensure the quality of AI weather forecasts. There is no doubt that this will yield better results than the first wave of AI weather forecasting.”

topic:

Source: www.newscientist.com

DeepMind AI achieves second place at International Mathematical Olympiad

Posted on by
Reply

DeepMind’s AlphaProof AI can tackle a wide range of math problems

Google DeepMind

Google DeepMind’s AI won a silver medal at this year’s International Mathematical Olympiad (IMO), the first time an AI has made it onto the podium.

The IMO is considered the world’s most prestigious competition for young mathematicians, and answering the exam questions correctly requires mathematical ability that AI systems typically lack.

In January, Google DeepMind showed off AlphaGeometry, an AI system that could answer IMO geometry problems as well as humans could, but it wasn’t in a real competition and couldn’t answer questions in other areas of math, such as number theory, algebra, or combinatorics, that are needed to win an IMO medal.

Google DeepMind has now released a new AI called AlphaProof that can solve a wider range of math problems, and an improved version of AlphaGeometry that can solve more geometry problems.

When the team tested both systems together on this year’s IMO problems, they got four out of six questions right, earning them 28 points out of 42 possible points – good enough for a silver medal, just one point short of this year’s gold medal threshold.

At the competition held in Bath, England, last week, 58 athletes won gold medals and 123 won silver medals.

“We all know that AI will eventually be better than humans at solving most mathematical problems, but the rate at which AI is improving is astounding,” he said. Gregor Doliner“It’s incredible to have missed out on gold at IMO 2024 by just one point just a few days ago,” said IMO Chairman Jonathan McClellan.

At a press conference, Timothy Gowers A University of Cambridge researcher who helped grade AlphaProof’s solutions said the AI’s performance was surprising, and that it seemed to have found the “magic keys” to solve the problems in a way that was similar to humans. “We thought that these magic keys would probably be a bit beyond the capabilities of an AI, so we were quite surprised in one or two cases where the program actually found them,” Gowers said.

AlphaProof works similarly to Google DeepMind’s previous AIs that can beat the best humans at chess and Go. All of these AIs rely on a trial-and-error approach called reinforcement learning, in which the system finds its own way of solving a problem by trying it again and again. However, this method requires a large number of problems written in a language that the AI can understand and verify, and IMO most such problems are written in English.

To avoid this, Thomas Hubert Using Google’s Gemini AI, a language model like the one that powers ChatGPT, the DeepMind researchers and his colleagues transformed these problems into a programming language called Lean, allowing the AI to learn how to solve them.

“You’ll start by solving maybe the simplest problems, and then you’ll be able to learn from solving those simple problems and then tackle the harder problems,” Hubert said at the press conference, and the answers will be generated in a lean language so they can be immediately verified for correctness.

Despite AlphaProof’s impressive performance, it was slow, taking three days to find a solution. That’s compared to 4.5 hours for the contestants, but AlphaProof failed to solve either of the two problems. The problems were about combinatorics, the study of counting and arranging numbers. “We’re still working on figuring out why that is, and if we can do that, that will help us improve the system,” AlphaProof says. Alex Davis At Google DeepMind.

It’s also not clear how AlphaProof arrives at its answers, or whether it uses the same mathematical intuition as humans, Gowers said. But he said Lean’s ability to translate proofs into English makes it easy to check whether they’re correct.

“The results are impressive and a significant milestone,” Jordy Williamson “There have been many attempts to apply reinforcement learning based on formal proofs, but none have been very successful,” say researchers at the University of Sydney in Australia.

Systems like AlphaProof may help working mathematicians develop proofs, but they obviously don’t help them identify the problems they need to solve and tackle, which takes up the majority of researchers’ time, he says. He Yanghui At the London Mathematical Institute.

Hubert said the team hopes that by reducing false responses, AlphaProof can help improve Google’s large-scale language models like Gemini.

Trading firm XTX Markets is offering a $5 million prize to any AI that can win a gold medal at the IMO (dubbed the AI Mathematics Olympiad), but AlphaProof is ineligible because it is not publicly available. “We hope that DeepMind’s progress will encourage more teams to apply for the AIMO prize, and of course we would welcome a public submission from DeepMind itself,” said Alex Gerko of XTX Markets.

topic:

Source: www.newscientist.com

AI Industry Faces Risks, Employees from OpenAI and Google DeepMind Sound Alarm

Posted on by
Reply

A group of current and former employees from prominent artificial intelligence companies has published an open letter. The committee warned of inadequate safety oversight within the industry and called for better protection for whistleblowers.

The letter, advocating for a “right to warn about artificial intelligence,” is a rare public statement about the risks of AI from employees in a usually secretive industry. It was signed by 11 current and former employees of OpenAI and two current and former Google DeepMind employees, one of whom previously worked at Anthropic.

“AI companies have valuable non-public information about their systems’ capabilities, limitations, safeguards, and risk of harm. However, they have minimal obligations to share this information with governments and none with the public. We cannot rely on companies to share this information voluntarily,” the letter stated.

OpenAI defended its practices, stating that they have hotlines and mechanisms for issue reporting, and they do not release new technology without proper safeguards. Google did not respond immediately to a comment request.

Concerns about the potential dangers of artificial intelligence have been around for years, but the recent AI boom has heightened these concerns, leading regulators to struggle to keep up with technological advancements. While AI companies claim to be developing their technology safely, researchers and employees warn about a lack of oversight to prevent AI tools from exacerbating existing societal harms or creating new ones.

The letter also mentions a bill seeking to enhance protections for AI company employees who raise safety concerns. The bill calls for transparency and accountability principles, including not forcing employees to sign agreements that prevent them from discussing risk-related AI issues publicly.

In a recent report, it was revealed that companies like OpenAI have tactics to discourage employees from freely discussing their work, with consequences for those who speak out. OpenAI CEO Sam Altman apologized for these practices and promised changes to exit procedures.

The open letter echoes concerns raised by former top OpenAI employees about the company’s lack of transparency in its operations. It comes after recent resignations of key OpenAI employees over disagreements about the company’s safety culture.

Source: www.theguardian.com

Microsoft Appoints DeepMind Co-founder to Lead Newly Formed AI Division

Posted on by
Reply

Microsoft has named the co-founder of the British artificial intelligence research institute DeepMind as the head of its new AI division. Mustafa Suleiman, now 39 years old, co-founded DeepMind with Demis Hassabis and Shane Legg back in 2010. The company was later acquired by Google in 2014 for £400m and has since become the core of Google’s AI efforts. Suleiman left DeepMind in 2019 and will now lead Microsoft AI, a new organization focusing on the US company’s consumer products and research. Several members from Suleiman’s Inflection AI startup will also join the division.

Microsoft has made a multibillion-dollar investment in OpenAI, the developer of the ChatGPT chatbot, to develop generative AI technology. Satya Nadella, Microsoft’s CEO, praised Suleiman as a visionary product maker and a team leader with a bold mission. The new division will integrate Microsoft’s consumer AI efforts, such as the Copilot chatbot and the Bing browser, which utilizes ChatGPT technology. Copilot is a key element in Microsoft’s AI monetization efforts, enabling users to easily compose emails, summarize documents, create presentations, and more.

Suleiman’s colleague Karen Simonyan, also a co-founder of Inflection AI, will join the new division as a principal investigator. Meanwhile, Bloomberg News reported that Apple is in talks to incorporate Google’s Gemini AI product into the iPhone. Inflection AI, backed by Microsoft and Nvidia, has become one of the leading companies in the generative AI race.

Suleiman, who has roots in both Syria and the UK, recently published a book on AI titled “The Coming Wave.” In it, he discusses both the potential benefits and risks of AI technology, calling for an increase in research on AI safety. In an interview last year, he described the book as a “provocation,” noting the importance of predicting future trends and taking action to mitigate potential risks.

Source: www.theguardian.com

Liverpool FC and DeepMind collaborate to create artificial intelligence for soccer strategy consultation

Posted on by
Reply

Corner kicks like this one taken by Liverpool's Trent Alexander-Arnold can lead to goal-scoring opportunities.

Robbie Jay Barratt/AMA/Getty

Artificial intelligence models predict the outcome of corner kicks in soccer matches and help coaches design tactics that increase or decrease the probability of a player taking a shot on goal.

petar veličković Google's DeepMind and colleagues have developed a tool called TacticAI as part of a three-year research collaboration with Liverpool Football Club.

A corner kick is awarded when the ball crosses the goal line and goes out of play, creating a good scoring opportunity for the attacking team. For this reason, football coaches make detailed plans for different scenarios, which players study before the game.

TacticAI was trained on data from 7176 corner kicks from England's 2020-2021 Premier League season. This includes each player's position over time as well as their height and weight. You learned to predict which player will touch the ball first after a corner kick has been taken. In testing, Ball's receiver ranked him among TacticAI's top three candidates 78% of the time.

Coaches can use AI to generate tactics for attacking or defending corners that maximize or minimize the chances of a particular player receiving the ball or a team getting a shot on goal. This is done by mining real-life examples of corner kicks with similar patterns and providing suggestions on how to change tactics to achieve the desired result.

Liverpool FC's soccer experts were unable to distinguish between AI-generated tactics and human-designed tactics in a blind test, favoring AI-generated tactics 90% of the time.

But despite its capabilities, Veličković says TacticAI was never intended to put human coaches out of work. “We are strong supporters of AI systems, not systems that replace AI, but augment human capabilities and allow people to spend more time on the creative parts of their jobs,” he says.

Velicković said the research has a wide range of applications beyond sports. “If you can model a football game, you can better model some aspects of human psychology,” he says. “As AI becomes more capable, it needs to understand the world better, especially under uncertainty. Our systems can make decisions and make recommendations even under uncertainty. It’s a good testing ground because it’s a skill that we believe can be applied to future AI systems.”

topic:

Source: www.newscientist.com

DeepMind AI integrates fact checker to make groundbreaking mathematical findings

Posted on by
Reply

DeepMind’s FunSearch AI can tackle mathematical problems

Arengo/Getty Images

Google DeepMind claims to have made the first ever scientific discovery in an AI chatbot by building a fact checker that filters out useless output and leaves behind only reliable solutions to mathematical or computing problems. Masu.

DeepMind’s previous achievements, such as using AI to predict the weather or the shape of proteins, rely on models created specifically for the task at hand and trained on accurate, specific data. I did. Large-scale language models (LLMs), such as GPT-4 and Google’s Gemini, are instead trained on vast amounts of disparate data, yielding a wide range of capabilities. However, this approach is also susceptible to “hallucinations,” which refers to researchers producing erroneous output.

Gemini, released earlier this month, has already shown hallucination tendencies and even gained simple facts such as: This year’s Oscar winners were wrong. Google’s previous AI-powered search engine even had errors in its self-launched advertising materials.

One common fix for this phenomenon is to add a layer on top of the AI ​​that validates the accuracy of the output before passing it on to the user. However, given the wide range of topics that chatbots may be asked about, creating a comprehensive safety net is a very difficult task.

Al-Hussein Fawzi Google’s DeepMind and his colleagues created a general-purpose LLM called FunSearch based on Google’s PaLM2 model with a fact-checking layer they call an “evaluator.” Although this model is constrained by providing computer code that solves problems in mathematics and computer science, DeepMind says this work is important because these new ideas and solutions are inherently quickly verifiable. is a much more manageable task.

The underlying AI may still hallucinate and provide inaccurate or misleading results, but the evaluator filters out erroneous outputs, leaving only reliable and potentially useful concepts. .

“We believe that probably 90% of what LLM outputs is useless,” Fawzi says. “If you have a potential solution, it’s very easy to tell whether this is actually the correct solution and evaluate that solution, but it’s very difficult to actually come up with a solution. So , mathematics and computer science are a particularly good fit.”

DeepMind claims the model can generate new scientific knowledge and ideas, something no LLM has ever done before.

First, FunSearch is given a problem and a very basic solution in the source code as input, and then generates a database of new solutions that are checked for accuracy by evaluators. The best reliable solutions are returned as input to the LLM with prompts to improve the idea. According to DeepMind, the system generates millions of potential solutions and eventually converges on an efficient result, sometimes even exceeding the best known solution.

For mathematical problems, a model creates a computer program that can find a solution, rather than trying to solve the problem directly.

Fawzi and his colleagues challenged FunSearch to find a solution to the cap set problem. This involves determining the pattern of points where three points do not form a straight line. As the number of points increases, the computational complexity of the problem increases rapidly. The AI ​​discovered a solution consisting of 512 points in eight dimensions, larger than previously known.

When tackling the problem of bin packing, where the goal is to efficiently place objects of different sizes into containers, FunSearch discovered a solution that outperformed commonly used algorithms. The result is a result that can be immediately applied to transportation and logistics companies. DeepMind says FunSearch could lead to improvements in more math and computing problems.

mark lee The next breakthrough in AI will not be in scaling up LLM to ever-larger sizes, but in adding a layer to ensure accuracy, as DeepMind has done with FunSearch, say researchers at the University of Birmingham, UK. It is said that it will come from.

“The strength of language models is their ability to imagine things, but the problem is their illusions,” Lee says. “And this study breaks that down, curbs that, and confirms the facts. It’s a nice idea.”

Lee says AI should not be criticized for producing large amounts of inaccurate or useless output. This is similar to how human mathematicians and scientists work: brainstorm ideas, test them, and follow up on the best while discarding the worst.

topic:

Source: www.newscientist.com

DeepMind AI outperforms top weather forecasts, with one caveat

Posted on by
Reply

Will the AI ​​tell me if I need an umbrella?

Sebastien Bozon/AFP via Getty Images

AI company Google DeepMind says AI can more accurately predict the weather 10 days out than current state-of-the-art simulations, but meteorologists are still trying to build weather models based on actual physical principles. It warns against abandoning and relying solely on patterns in the data. Expose the shortcomings of AI approaches.

Existing weather forecasts are based on mathematical models, which use physics and powerful supercomputers to definitively predict what will happen in the future. These models have gradually become more accurate by adding more details, but this requires more calculations, more powerful computers, and higher energy demands.

Rémi Lam and his colleagues at Google DeepMind took a different approach. The company’s GraphCast AI model is trained on his 40 years of historical weather data from satellites, radar, and ground-based measurements to identify patterns that even Google DeepMind can’t understand. “As with many machine learning AI models, it’s not that easy to interpret how the model works,” Lamb says.

To make the predictions, actual weather measurements taken at two points six hours apart from more than one million locations around the globe are used to predict the weather six hours into the future. These predictions can be used as input for another round that predicts another 6 hours into the future.

DeepMind researchers carried out this process using data Produces a 10-day forecast from the European Center for Medium-Range Weather Forecasts (ECMWF). They say it outperformed ECMWF’s “gold standard” High Resolution Forecasting (HRES) by providing more accurate forecasts on more than 90 per cent of the data points tested. At some altitudes, this accuracy increased to 99.7%.

Matthew Chantry He, who worked with Google DeepMind at ECMWF, said his organization had previously seen AI as a tool to complement existing mathematical models, but in the past 18 months it has seen a shift in the way AI can actually provide predictions on its own. He said that he has become able to do so.

“We at ECMWF believe this is a very exciting technology that has the potential not only to reduce energy costs when making forecasts, but also to improve them. Creating a reliable operational product “Probably more work is needed, but this is likely the beginning of a revolution in the way weather forecasts are made, and this is our assessment,” he says. According to Google DeepMind, using GraphCast to make a 10-day forecast takes him less than a minute on a high-end PC, but with HRES it can take several hours on a supercomputer.

But some meteorologists are wary of trusting weather forecasting to AI. Ian Renfrew According to researchers at the University of East Anglia in the UK, GraphCast currently lack the ability to marshal data into its own starting state, a process known as data assimilation. In traditional predictions, this data is carefully incorporated into simulations after thorough checks on physics and chemistry calculations to ensure accuracy and consistency. Currently, GraphCast must use a starting state prepared in the same way by ECMWF’s own tools.

“Google won’t be doing weather forecasts any time soon because they can’t assimilate the data,” Renfrew said. “And data assimilation is typically one-half to two-thirds of the computation time for these forecasting systems.”

He says there are also concerns about completely abandoning deterministic models based on chemistry and physics and relying solely on AI output.

“Even if you have the best predictive model in the world, what’s the point if the public doesn’t trust you and you don’t take action? We ordered the evacuation of 30 miles of Florida’s coastline. “If nothing happens, it will blow away the trust that has been built over decades,” he says. “The advantage of a deterministic model is that you can investigate it. If you get a bad prediction, you can investigate why that prediction is bad and target those aspects for improvement.”

topic:

Source: www.newscientist.com