This feels like a narrative being pushed. Tech oligarchs these days are flexing by showing off how much they are able to bully government, and Altman wants to get in on the game by spinning things this way. I’m not convinced they really bent the rules for OpenAI any more than usual given they only employ a few hundred people.
I doubt OpenAI has to try very hard to convince state politicians, especially to the point of bullying. It's usually the complete opposite where they throw money at you to stay.
Imagine if California managed to scare away the hottest company in the world and all the tax revenue it brings...
It was a comment revealing his attitude towards what would soon become his customers in a globally-impactful business over which he has sole control. It’s more relevant than ever today.
In most cases those comments weren’t about the service they still running at 40. His comment was also him self-reporting that he shouldn’t be trusted with user data, when his whole business revolves around user data. If it was some off color joke in poor taste, I wouldn’t care so much.
He's an entirely different person now. There is no one under the age of 45 on the planet who would say "you know, yeah, I'm fundamentally the same person I was 22 years ago."
I'm not saying he's a better person. Just different. Judge him by what he says and does now, which is no better.
I feel that I fundamentally am the same person. More experienced, of course. Less naive and idealistic. But my sense of right and wrong? Pretty much the same.
You can be an asshole at 22 and still be an asshole at 45. You might be an asshole in different ways, but an asshole is still an asshole. As I'm often reminded of myself
The things Donald Trump says are sufficiently untethered from reality that whether a given statement is truth or lie could be used as a pseudorandom number generator, so I don't know if that statement counts for or against my claim.
I don't disagree with your point, but nothing is absolute. In this case, he's essentially done a tremendously good job of showing us since then that, no, we can't trust him. He's more than lived up to his words.
Wow. The 2019 novel “The Last Astronaut” hypothesized about a fictional interstellar object coming into the solar system, called “2I” in the novel for short, but back here in real life, we’re already up to 3I.
Agree, as someone who has spent way too much time studying the way motion picture film looks up close, this isn’t very realistic looking. It’s really just a form of dithering.
Capitalists always hate capitalism when it comes to employees getting paid what they are worth. If the market will bear it, he should embrace it and stop whining.
> I’ve been feeling pretty good about my benchmark! It should stay useful for a long time... provided none of the big AI labs catch on.
> And then I saw this in the Google I/O keynote a few weeks ago, in a blink and you’ll miss it moment! There’s a pelican riding a bicycle! They’re on to me. I’m going to have to switch to something else.
Yeah this touches on an issue that makes it very difficult to have a discussion in public about AI capabilities. Any specific test you talk about, no matter how small … if the big companies get wind of it, it will be RLHF’d away, sometimes to the point of absurdity. Just refer to the old “count the ‘r’s in strawberry” canard for one example.
Honestly, if my stupid pelican riding a bicycle benchmark becomes influential enough that AI labs waste their time optimizing for it and produce really beautiful pelican illustrations I will consider that a huge personal win.
strawberry -> DeepSeek, GeminiPro and ChatGPT4o all correctly said three
strawberrry -> DeepSeek, GeminiPro and ChatGPT4o all correctly said four
stawberrry -> DeepSeek, GeminiPro all correctly said three
ChatGPT4o even in a new Chat, incorrectly said the word "stawberrry" contains 4 letter "r" characters. Even provided this useful breakdown to let me know :-)
Breakdown:
stawberrry → s, t, a, w, b, e, r, r, r, y → 4 r's
And then asked if I meant "strawberry" instead and said because that one has 2 r's....
This doesn’t make ARC a bad benchmark. Tech giants will have a significant advantage in any benchmark they are interested in, _especially_ if the benchmark correlates with true general intelligence.
This is a fantastic comment and I couldn’t agree more. I don’t what they’re going to come up with as a result of this partnership but I expect that it will be utterly lacking in the qualities you describe. You really put your finger on it.
I don’t know why you’re being downvoted. It is absolutely in the cards for the ownership class to send out swarms of drones to kill us all.
The best we can hope for is to all end up in permanent slums picking through garbage to survive.
One peep of protest a little too loud and suddenly we’ll all be living the same scenario that’s visible all over the “combat footage” portions of the web: you are walking along outdoors, you hear a buzz, and then suddenly you look up to realize there’s a drone coming at you. You panic and try to run but by the time you hear it, it’s too late. It catches up to you easily and drops a bomb on you. Or maybe it’s a smaller more targeted (and disposable) one that zips straight at your forehead and explodes on contact, destroying your brain.
People may say I’m a doomsayer but just look back at the history of labor in the US. They sent out privatized security to shoot anyone who didn’t cooperate. And that was when they still needed the workers! Just imagine what they’ll do when they no longer need anyone!
There's a lot of truth to that however the supply chain for everything today has never been longer or had more intermediaries. It's like a house of cards built ever higher and no one has budgeted anything for resilience. JIT, outsourcing, etc, everything is cheaper at the cost of resilience.
Oh yes I’ve thought of that, but I’m sure they have too. If AI advances to the point where everyone is out of work, then it will have also likely advanced to the point where it can aid in getting around the current supply chain fragilities.
That angle unfortunately just doesn’t give me much hope.
Really curious to know what the test shots from Boss Films looked like!
Richard Edlund’s team’s work on various films of the 80s was impeccable (including Ghostbusters, Die Hard, Big Trouble in Little China, and more). They were the inheritor of the crown of high quality 65mm VFX after Douglas Trumbull quit the business (the pinnacle of his work being films like Blade Runner and Brainstorm).
I have to wonder what was wrong that caused the last minute switch to ILM!
