TTake a look at Sam Altman. Seriously, check Google Images, and you’ll notice an abundance of photos featuring the endearing Lost Puppy from Silicon Valley, showcasing the OpenAI chief sporting a clever grin. Yet, I suggest hiding the lower half of his face in these images. Suddenly, Sam’s expression takes on the haunting gaze of the boyfriend of a missing woman, pleading for her return: “Please come home, Sheila. We’re worried about you, and we just want you back.”

Don’t be alarmed if the humor feels misplaced, crude, or somewhat manipulative. I rely on OpenAI’s guiding principle: reciprocity. Content creators must formalize and painstakingly select subjects for use in generated content. outside to be utilized in any manner users see fit. I haven’t received any word from Sam, leading me to believe I know precisely where he is because I placed Sheila there. After all, he seems to fit the archetype that often accompanies the term “visibly.”

For Sam, the past fortnight has revolved around the debut of the AI video generator Sora 2 (a remarkable enhancement from the Sora of just ten months prior) and his entanglement in issues surrounding copyrighted content. Additionally, there were announcements about further interconnected transactions involving OpenAI and chip manufacturers like: Nvidia and AMD. This has led to the OpenAI frenzy, with total transaction volume surpassing $1 trillion just this year. While you can enjoy videos showcasing meticulously designed characters manipulated into digital puppets by uncreative, bigoted individuals, it also means that with OpenAI, you could lose your home in a disastrous financial collapse if the bubble bursts.

I don’t wish to offend the creators of Sora. I’ve strolled through art galleries and realized that if I were to deface an artwork with a ridiculous doodle, it would surprisingly add value; hence, if I didn’t want it, I wouldn’t have exposed it to the public. Moreover, none of the tech giants seem to lead a civilized life, so they probably cannot fathom any creative value worth preserving from being tarnished for profit. If you’ve followed Sam’s frequent reading lists, you’ll see it’s akin to the “Business Philosophy” section of a mediocre airport bookstore. This week, they mainly wanted to convey that Sora 2 is about being cool and fun. “Seeing your feed filled with memes about yourself isn’t as bizarre as you might think,” Sam assured us. So all is well! Though, I think it’s beneficial to note that while you’re inundated with simulated revenge content in a modern-day version of Byzantium, you’re also one of the most influential individuals globally profiting immensely from it. confuse “guardrail.”

I’ve heard people propose that OpenAI’s motto should be “It’s better to ask for forgiveness than permission,” but that misplaces the priority. Its real motto appears to be, “We do what we wish, and you simply deal with it.” Consider Altman’s recent political trajectory. “For those familiar with German history in the 1930s” Sam forewarned back in 2016, reflecting on Trump’s actions. It seems he has reconciled this concern in time to join. Donald Trump’s second inauguration. Perhaps, to extend his well-crafted analogy, it’s due to him being among the entrepreneurs welcomed into the Prime Minister’s office to claim their portion of the gains. “Thank you for being such a pro-business, pro-innovation president,” Sam effused to Trump at a recent White House dinner for tech executives. “It’s a refreshing change.” Unsurprisingly, the Trump administration has chosen to evade AI regulation entirely.

On the flip side, recall what Sam and his skeptical comrades stated earlier this year when it was suggested that the Chinese AI chatbot DeepSeek might have leveraged some of OpenAI’s work. His organization issued a concerned statement, asserting, “We are aware of and investigating indications that DeepSeek may have improperly extracted our models. We will provide further details as we learn more.” “We are taking proactive and assertive measures to safeguard our technology.” Interestingly, OpenAI appears to be the only entity on earth with the ability to combat AI theft.

This week, Hollywood talent agencies took the initiative to coax some form of temporary silence from Altman. I posted flannel—if not in riches, then certainly in striving to establish a “new kind of engagement” with those he has openly referred to as “rights holders.” Many of us remember just a short while ago, when rights holders held all the power. Those who possess rights. In other words, the hint lies within the terminology. However, Sam embodies the post-light era. The question arises: if he is bestowing creative rights, can we genuinely believe he’s not also conferring other types of rights?

OpenAI desires what all nurturing platforms ultimately aim for: users to remain within their realm indefinitely. It is clearly poised to become the new default homepage of the internet, much like Meta once was. Are childhood privacy catastrophes, election manipulation controversies, and child exploitation crises not far off?

Because, incredibly, we have already traversed this life cycle. But I suppose we must revisit it, right? Or more accurately, since Sam’s company is advancing at an unprecedented pace, we have already done it again. Initially, we admire the enigmatic engineer Pied Piper as a brilliant and unconventional altruist, only to later uncover that he is not as he appears and that his technology poses greater risks than we comprehended, leading to our failure to regulate it, rendering us the victims. In many ways, this mirrors a poor AI reinterpretation of a film we’ve already witnessed. If Altman’s model can learn, why can’t we?

• Marina Hyde is a columnist for the Guardian

• A year at Westminster: John Crace, Marina Hyde, and Pippa Crellard
  On Tuesday, December 2nd, Crace, Hyde, and Crellard will reflect on this remarkable year alongside special guests. It will be streamed live from the Barbican in London and available worldwide. Reserve your ticket here or on guardian live

• Do you have thoughts on the subjects discussed in this article? Click here if you would like to send an email response of up to 300 words for publication in our email section.