OpenAI's Sam Altman Wants AI in the Hands of the People - and Universal Basic Compute? (youtube.com) 79
OpenAI CEO Sam Altman gave an hour-long interview to the "All-In" podcast (hosted by Chamath Palihapitiya, Jason Calacanis, David Sacks and David Friedberg).
And when asked about this summer's launch of the next version of ChatGPT, Altman said they hoped to "be thoughtful about how we do it, like we may release it in a different way than we've released previous models...
Altman: One of the things that we really want to do is figure out how to make more advanced technology available to free users too. I think that's a super-important part of our mission, and this idea that we build AI tools and make them super-widely available — free or, you know, not-that-expensive, whatever that is — so that people can use them to go kind of invent the future, rather than the magic AGI in the sky inventing the future, and showering it down upon us. That seems like a much better path. It seems like a more inspiring path.
I also think it's where things are actually heading. So it makes me sad that we have not figured out how to make GPT4-level technology available to free users. It's something we really want to do...
Q: It's just very expensive, I take it?
Altman: It's very expensive.
But Altman said later he's confident they'll be able to reduce cost. Altman: I don't know, like, when we get to intelligence too cheap to meter, and so fast that it feels instantaneous to us, and everything else, but I do believe we can get there for, you know, a pretty high level of intelligence. It's important to us, it's clearly important to users, and it'll unlock a lot of stuff.
Altman also thinks there's "great roles for both" open-source and closed-source models, saying "We've open-sourced some stuff, we'll open-source more stuff in the future.
"But really, our mission is to build toward AGI, and to figure out how to broadly distribute its benefits... " Altman even said later that "A huge part of what we try to do is put the technology in the hands of people..." Altman: The fact that we have so many people using a free version of ChatGPT that we don't — you know, we don't run ads on, we don't try to make money on it, we just put it out there because we want people to have these tools — I think has done a lot to provide a lot of value... But also to get the world really thoughtful about what's happening here. It feels to me like we just stumbled on a new fact of nature or science or whatever you want to call it... I am sure, like any other industry, I would expect there to be multiple approaches and different peoiple like different ones.
Later Altman said he was "super-excited" about the possibility of an AI tutor that could reinvent how people learn, and "doing faster and better scientific discovery... that will be a triumph."
But at some point the discussion led him to where the power of AI intersects with the concept of a universal basic income: Altman: Giving people money is not going to go solve all the problems. It is certainly not going to make people happy. But it might solve some problems, and it might give people a better horizon with which to help themselves.
Now that we see some of the ways that AI is developing, I wonder if there's better things to do than the traditional conceptualization of UBI. Like, I wonder — I wonder if the future looks something more like Universal Basic Compute than Universal Basic Income, and everybody gets like a slice of GPT-7's compute, and they can use it, they can re-sell it, they can donate it to somebody to use for cancer research. But what you get is not dollars but this like slice — you own part of the the productivity.
Altman was also asked about the "ouster" period where he was briefly fired from OpenAI — to which he gave a careful response: Altman: I think there's always been culture clashes at — look, obviously not all of those board members are my favorite people in the world. But I have serious respect for the gravity with which they treat AGI and the importance of getting AI safety right. And even if I stringently disagree with their decision-making and actions, which I do, I have never once doubted their integrity or commitment to the sort of shared mission of safe and beneficial AGI...
I think a lot of the world is, understandably, very afraid of AGI, or very afraid of even current AI, and very excited about it — and even more afraid, and even more excited about where it's going. And we wrestle with that, but I think it is unavoidable that this is going to happen. I also think it's going to be tremendously beneficial. But we do have to navigate how to get there in a reasonable way. And, like a lot of stuff is going to change. And change is pretty uncomfortable for people. So there's a lot of pieces that we've got to get right...
I really care about AGI and think this is like the most interesting work in the world.
And when asked about this summer's launch of the next version of ChatGPT, Altman said they hoped to "be thoughtful about how we do it, like we may release it in a different way than we've released previous models...
Altman: One of the things that we really want to do is figure out how to make more advanced technology available to free users too. I think that's a super-important part of our mission, and this idea that we build AI tools and make them super-widely available — free or, you know, not-that-expensive, whatever that is — so that people can use them to go kind of invent the future, rather than the magic AGI in the sky inventing the future, and showering it down upon us. That seems like a much better path. It seems like a more inspiring path.
I also think it's where things are actually heading. So it makes me sad that we have not figured out how to make GPT4-level technology available to free users. It's something we really want to do...
Q: It's just very expensive, I take it?
Altman: It's very expensive.
But Altman said later he's confident they'll be able to reduce cost. Altman: I don't know, like, when we get to intelligence too cheap to meter, and so fast that it feels instantaneous to us, and everything else, but I do believe we can get there for, you know, a pretty high level of intelligence. It's important to us, it's clearly important to users, and it'll unlock a lot of stuff.
Altman also thinks there's "great roles for both" open-source and closed-source models, saying "We've open-sourced some stuff, we'll open-source more stuff in the future.
"But really, our mission is to build toward AGI, and to figure out how to broadly distribute its benefits... " Altman even said later that "A huge part of what we try to do is put the technology in the hands of people..." Altman: The fact that we have so many people using a free version of ChatGPT that we don't — you know, we don't run ads on, we don't try to make money on it, we just put it out there because we want people to have these tools — I think has done a lot to provide a lot of value... But also to get the world really thoughtful about what's happening here. It feels to me like we just stumbled on a new fact of nature or science or whatever you want to call it... I am sure, like any other industry, I would expect there to be multiple approaches and different peoiple like different ones.
Later Altman said he was "super-excited" about the possibility of an AI tutor that could reinvent how people learn, and "doing faster and better scientific discovery... that will be a triumph."
But at some point the discussion led him to where the power of AI intersects with the concept of a universal basic income: Altman: Giving people money is not going to go solve all the problems. It is certainly not going to make people happy. But it might solve some problems, and it might give people a better horizon with which to help themselves.
Now that we see some of the ways that AI is developing, I wonder if there's better things to do than the traditional conceptualization of UBI. Like, I wonder — I wonder if the future looks something more like Universal Basic Compute than Universal Basic Income, and everybody gets like a slice of GPT-7's compute, and they can use it, they can re-sell it, they can donate it to somebody to use for cancer research. But what you get is not dollars but this like slice — you own part of the the productivity.
Altman was also asked about the "ouster" period where he was briefly fired from OpenAI — to which he gave a careful response: Altman: I think there's always been culture clashes at — look, obviously not all of those board members are my favorite people in the world. But I have serious respect for the gravity with which they treat AGI and the importance of getting AI safety right. And even if I stringently disagree with their decision-making and actions, which I do, I have never once doubted their integrity or commitment to the sort of shared mission of safe and beneficial AGI...
I think a lot of the world is, understandably, very afraid of AGI, or very afraid of even current AI, and very excited about it — and even more afraid, and even more excited about where it's going. And we wrestle with that, but I think it is unavoidable that this is going to happen. I also think it's going to be tremendously beneficial. But we do have to navigate how to get there in a reasonable way. And, like a lot of stuff is going to change. And change is pretty uncomfortable for people. So there's a lot of pieces that we've got to get right...
I really care about AGI and think this is like the most interesting work in the world.
non-AI summary: (Score:5, Insightful)
Re: (Score:2)
Altman argues for OpenAI to become an essentiall utility, like water and electricity companies, except also obtain a de facto monopoly via dumping (technical economic term meaning you give away products free or below cost so as to make it more difficult for competitors to enter, or stay, in the market).
When your competitors are Google, Amazon and Microsoft, dumping won't work.
Re: (Score:3, Interesting)
TBF Microsoft isn't an OpenAI competitor, OpenAI is closer to being a Microsoft subsidiary at this point.
Re: (Score:2)
I dunno, plumbing sounds like the perfect example. Sure most people could figure out how to replace a bit here or there, but what if they wanted step by step instructions for installing a garbage disposal, in their existing sink. Maybe they could cobble together a half dozen YouTube videos to get something that matches their exact setup, or maybe they could upload a good picture to an AI assistant with a link to the manual for their disposal and it will tell them exactly what to buy and how to install it
Re: (Score:2)
I can't believe I still need to post this: AI is not capable of understanding, reasoning, or even calculating. It simply can not create the kinds of tailored experiences you're after. That's just not how these things work.
it will tell them exactly what to buy and how to install it and will properly note things like relative drain heights and slope and how to tie in with their weird 1980s dishwasher.
No, it won't. That's a silly science fiction.. An AI is simply not capable of producing output like that with any accuracy. It can't meaningfully estimate heights, slope, or be able to identify, let alone adapt to, any unusual equipment. Again, that's just not how these things work.
J
Re: (Score:3)
Crypto scammer bro says what? (Score:5, Insightful)
No one gves a fuck, Mr Worldcoin.
Con artist,
Re: (Score:2)
Attention is all he needs.
Re: (Score:2)
Attention is all he needs.
Oh, I think a generous sprinkling of gullibility tends to help.
Re: (Score:2)
It's an AI joke. Probably the most famous foundational paper of modern AI is called "Attention Is All You Need" [neurips.cc], which introduced the Transformers architecture based on an advanced attention mechanism.
But the title certainly applies to Altman ;)
give it a rest (Score:5, Insightful)
this is, like, the third post about this interview today. who cares about public statements of some suit-and-tie, whose job is to lie to the public?
Re:give it a rest (Score:5, Insightful)
You don't understand: we had more than 3 non-AI stories in a row. No news outlet worth its salt lets more than 2 hours pass without talking about AI in 2024!
So Slashdot used whatever material they had on hand to correct that intolerable slip. Expect this interview to be milked again in 2 hours.
Re: (Score:2)
I repent and will embrace this scourge and rejoice!
Re: (Score:2)
You don't understand: we had more than 3 non-AI stories in a row. No news outlet worth its salt lets more than 2 hours pass without talking about AI in 2024!
Wonder what will be “worth its salt” when humans are no longer writing the stories about AI.
If you thought humans were self-centered narcissistic attention whores, just wait until you see what AI is being trained to do.
Re: (Score:2)
this is, like, the third post about this interview today. who cares about public statements of some suit-and-tie, whose job is to lie to the public?
(Clicky McBait) ”Hey! Thats not very nice. I only got one job you know..”
Re: (Score:3)
Re:give it a rest (Score:5, Insightful)
this is, like, the third post about this interview today. who cares about public statements of some suit-and-tie, whose job is to lie to the public?
While it doesn't warrant 3 stories, you absolutely *should* care about this. AI is slated to become one of the world's biggest drivers in energy consumption. It is going to encroach every part of your life, whether you're a tech-bro early adopter, or just some normal guy who gets this shit shoved down your throat by the industry looking to make a dollar.
Also this is Slashdot, so emerging tech trends discussed by the leader of a company at the forefront of the latest tech development is one of the core pieces of interest on this site.
No one likes dupes, but if you're not interested in the story at all then I really have to wonder what site you think this is.
Re: (Score:2)
"AI" isn't "the latest tech development", it is the latest investor fad.
Re: (Score:3)
"AI" isn't "the latest tech development", it is the latest investor fad.
False. You can see that yourself in both investments in equipment, changes and effects on products, and even something as simple as changes in open source hobbyists committing on git-hub. You not liking it doesn't change the fact that until now we haven't had the hardware to do what we are doing (latest tech development) nor the actual interest both corporate and non-corporate for it.
Re: (Score:2)
Yes, I can see tons of hype.
You eating it up wholesale doesn't change the fact that we're seeing the dot-com booms all over again.
Re: (Score:2)
Re: (Score:2)
Hehe
How exceptionally convenient... (Score:5, Insightful)
One can certainly see why a supplier of chatbot would be enthusiastic about a new guaranteed market for it(in much the same way that the agricultural lobby is very interested indeed in welfare programs as a means of getting money spent on their products); but that's quite different from it being a credible or respectable view.
Re: (Score:3, Insightful)
There's a huge difference.
UBI is tax payer distributions of cash to everyone.
Whereas what he wants is tax payer distributions to him to subsidize his expenses to gain more users and lock out competition.
Re: (Score:2)
It's a very low standard. I'm incredibly stupid but still hit that mark every time.
Replace "OpenAI" with "MicrosoftAI" (Score:5, Insightful)
Whenever you see "OpenAI", replace the name with "MicrosoftAI", and then see if that changes how you view what was said. If so, you now know you have been tricked by the word "open".
Re: Replace "OpenAI" with "MicrosoftAI" (Score:2)
OpenAI is open, but only to Microsoft.
Re: (Score:2)
Just tried it, turns out I'm more interested in the tech, the product, and its specific Terms of Use than your culture war. Do yourself a favour and actually look up how Microsoft and OpenAI relationship works, and realise there's a reason Microsoft has their own product if that company name triggers you so much.
Yeah, I feel all nice and fuzzy now (Score:4, Insightful)
When billionaire tech bros say they want empower the people, I feel a slight wave of nausea welling up.
Re: (Score:2)
Shut up and buy your World Coin, peasant.
Altman is just a trojan horse industry lobbyist (Score:3)
He'll throw off his sheep's clothes in just a few months again when we're all enslaved to the Mighty OpenAI.
Re:Altman is just a trojan horse industry lobbyist (Score:4, Funny)
Holy mixed metaphors, Batman! Is he a Greek warrior or a wolf? Is he a Greek wolf-in-sheep's-clothing hiding in a Trojan horse? It's so confusing!
Musk 2.0 (Score:2)
Re:Musk 2.0 (Score:5, Insightful)
This guys sounds worse than Elon Musk. Whoever falls for his shit, highly deserves whatever will happen to them. Will people never learn not to listen to snake oil sellers?
Judging by how Trump is doing in the presidential election polls that does not seem likely.
Re: (Score:2)
Yep, most people who use the term TDS are super cowards. You just happen to be a homophobe as well.
Nice in Theory (Score:3)
Re: (Score:1)
antisocial != asocial
Re: (Score:2)
"Too cheap to meter" (Score:3)
The same was said about nuclear power. I'm not waiting for that to happen, either.
Something to think about (Score:2)
Could we have distributed compute cause an upheaval here like bittorrent (protocol, not client) did to file distribution?
Where suddenly individual weak clients can join a swarm that has a massive total capacity, that would generate better results for everyone than it could individual results for any individual client within the swarm on its own.
Re: (Score:2)
Maybe:
https://dataconomy.com/2024/04... [dataconomy.com]
Re: (Score:2)
Almost all of these are "blockchain and crypto". I'm talking about an actual solution to the problem of distributed compute that would be actually functional for all without any of the nonsense that follows those two. That's why I listed bittorrent as a good example. Cohen made no money on inventing it. It was just a problem that he saw existing in real world, and he built a protocol for a distributed solution and let other people do whatever they wanted with it.
And now, even windows updates are distributed
Re: (Score:2)
Hah. Okay. Good luck with that. Meanwhile blockchain AI lurches forward, with or without you.
Re: (Score:2)
Lurches is unironically a good word to use here. It's a zombie, getting those bursts of energetic charges towards what it sees as free money to defr.... ahem, towards anything with brains.
And then slowly shambles around pointlessly for a long time after that small burst of energy is done.
Re: (Score:2)
Altman v Marx (Score:3)
"But what you get is not dollars but this like slice — you own part of the the productivity."
So the workers shall...control the means of production?
Hands of the People (Score:3)
Six or seven fingers?
Nothing ever becomes too cheap to meter (Score:2)
That was the promise of nuclear energy. Look how well that worked out.
It was the promise of phone service. Look how well that worked out.
Now we have someone saying it about AI. Guess how that will work out?
Re: Nothing ever becomes too cheap to meter (Score:2)
Except for ultra long distance calls, my phone service is unmetered. Has been for a while now.
Re: (Score:2)
It was the promise of phone service. Look how well that worked out.
The only reason any metering exists anymore is telcos milking older customers who are still on things like landlines. Most cell service is unmetered in the US for domestic and even some international calling. Ditto for most VOIP. Not to mention the many voice/video apps that are fully unmetered.
growing the future from the bottom up (Score:1, Interesting)
The word "compute" (Score:2)
Has the word "compute" changed meaning? I'm probably just old, but I hear it thrown around a lot in ways I didn't 10-20 years ago.
Re: (Score:2)
Here's an example, using it as a noun: "We need to upgrade our compute resources to handle the increased demand from new users."
Re: (Score:2)
That just sounds like you forgot an R. "We need to upgrade our computer's resources"
Compute is a verb, not a noun.
Re: (Score:2)
I agree, but lot's of people are using it as a noun these days. I was just wondering if anyone else found it odd :)
Re: (Score:2)
That just sounds like you forgot an R. "We need to upgrade our computer's resources"
What computer? Your "computer" is a VM floating around on a sea of hypervisor hosts. Its "disk" is file sitting on a storage array somewhere else in the datacenter. Its networking is just a driver in the OS connected to a virtual network stack.
Re: (Score:2)
ChatGPT credits instead of UBI? (Score:2)
This is pretty out of touch. "Don't bother trying to make sure everyone has enough money to buy food etc., why not just give them credits to use my company's generative AI product, and they can trade those!" What an asshole.
I'm still waiting for AI to do something useful. (Score:2)
They have a real problem they keep avoiding and never talk about. There isn't any I(Intelligence) in their AI.
Today's LLM's are just fancy automation and pattern matching on a grand scale.
Re: (Score:2)
Sounds like you don't know much about the technology. Yes a simplistic view is it's fancy automation & pattern matching, but conveniently skipping over how that pattern matching is "learning".
As far as AI doing something useful, the first thing that comes to mind is AI being used in medicine to detect tumors from MRI scans, x-rays, etc. You could triage healthcare in an emergency zone much faster with the help of these tools.
Re: (Score:2)
I can't say what he does or doesn't know from his comment. However, I can say with some authority that his assessment is essentially correct. There is nothing like understanding or intelligence here.
Also, what we call "learning" in AI is analogous but very much not the same as what we'd call "learning" in other contexts.
AI being used in medicine to detect tumors from MRI scans, x-rays, etc.
There's a story I can't find at the moment, about an AI that was able to identify something on xrays with perfect accuracy. It turned out to be picking xrays with a label in the corner. W
Re: (Score:2)
Re: I'm still waiting for AI to do something usefu (Score:2)
I agree. This would be great, but I don't think it is proven.
Imagine if, when the helicopter first flew, everyone went crazy about how this new antigravity technology would soon allow the colonization of distant galaxies. That is how I see the hype about the current crop of 'AI'.
"But what does it truly MEAN to Alta a Vista?" (Score:2)