Sam Altman is the modern version of a snake oil salesman
It's Time to Stop Taking Sam Altman at His Word
241–250 of 599 posts
Re: It's Time to Stop Taking Sam Altman at His Word
#242The field is extremely research oriented. You can't stay on top with good engineering and incremental development and refining.
Google just paid over $2.4 billion to get Noam Shazeer back in the company to work with Gemini AI. Google has the deepest pool of AI researchers. Microsoft and Facebook are not far behind.
OpenAI is losing researchers, they have maybe 1-2 years until they become Microsoft subsidiary.
Re: It's Time to Stop Taking Sam Altman at His Word
#243I can't imagine taking The Atlantic seriously on anything. My word. You aren't actually supposed to read the endless ragebait. Contrary to the Atlantic's almost always intentionally misleading framing, the "dot com boom" did in fact go on to print trillions later and it is still printing them. After what was an ultimately marginal if account clearing dip for many. I say that as someone who would be deemed to be an Ai…
> I can't imagine taking The Atlantic seriously on anything.
Agreed - I stopped taking The Atlantic seriously after their 2009 cover story, "Did Christianity Cause the Crash?"[1] To ignore CDOs, the Glass-Steagal repeal, the co-option of the ratings agencies and the dissolution of lending standards, and instead blame the Great Recession on a few obnoxious megapastors is to completely discard the magazine's credibility.
[1] https://www.theatlantic.com/magazine/archive/2009/12/did-chr...
Re: It's Time to Stop Taking Sam Altman at His Word
#244There are legit criticisms of Sam Altman that can be levied but none of them are in this article. This is just reductive nonsense. The arguments are essentially: 1. The technology has plateaued, not in reality, but in the perception of the average layperson over the last two years. 2. Sam _only_ has a record as a deal maker, not a physicist. 3. AI can sometimes do bad things & utilizes a lot of energy. I normally rea…
I think LLM technology, not necessarily all of CNN, has plateaued. We've used up all the human discourse, so there's nothing to train it on. It's like fossil fuels. They took billions of years to create and centuries to consume. We can't just create more. Another problem is that the data sets are becoming contaminated, creating a reinforcement cycle that makes LLMs trained on more recent data worse. My thoughts are t…
What data are you using to back up this belief?
Progress on benchmarks continues to improve (see GPT-o1).
The claim that there is nothing left to train on is objectively false. The big guys are building synthetic training sets, moving to multimodal, and are not worried about running out of data.
o1 shows that you can also throw more inference compute at problems to improve performance, so it gives another dimension to scale models on.
Re: It's Time to Stop Taking Sam Altman at His Word
#245Earlier quoted context omitted.
barber demands paper cash only...I have to stand in line at USPS and DMV Surely this is just a case of the future not being evenly distributed. All of these 'problems' are already solved and the solution is implemented somewhere, just not where you happen to be.
You're probably right. I'll wait until it gets better distributed. It's just that personally, it's been hard to reconcile the grandiose talk with what's actually around me, that's all.
The weird distribution was something I experienced when I last visited California in 2018 and couldn't use my card in most places, despite having been to Kenya a few years before that and seeing M-Pesa all over the place.
Re: It's Time to Stop Taking Sam Altman at His Word
#246Earlier quoted context omitted.
That's a really good comment and insight, but understandably I think it is aimed at this forum and a technical audience. It landed well for me in terms of the near team impact of LLMs and other models. But outside this forum, I think our field is in a crisis from being very substantially oversold and undersold at the same time. We have a very limited ability to define human intelligence, so it is almost impossible to…
LLMs aren't really simulating intelligence so much as echoing intelligence. The problem is when the hype machine causes the echoes to replace the original intelligence that spawned the echoes, and eventually those echoes fade into background noise and we have to rebuild the original human intelligence again.
I appreciate this, that is why I said, "LLMs and other models". Knowing the probability relations between words, tokens, or concepts/though vectors is important, and can be supplemented by smaller embedded special purpose models/inference engines and domain knowledge in those areas.
As I said, it is overhyped in some areas and underhyped in others.
Re: It's Time to Stop Taking Sam Altman at His Word
#247To recap OpenAI's decisions over the past year: * They burned up the hype for GPT-5 on 4o and o1, which are great step changes but nothing the competition can't quickly replicate. * They dissolved the safety team. * They switched to for profit and are poised to give Altman equity. * All while hyping AGI more than ever. All of this suggests to me that Altman is in short-term exit preparation mode, not planning for AGI…
Let's go with the no next gen models, then we'd be looking at an operationalized service. I might be up for paying $20/month for one. Certainly more value than I get from Netflix. All I would want from the service is a simple low-friction UI and continued improvements to keep up with or surpass competitors. They could manage that.
The long-term problem may be access to quality/human-created training data. Especially if the ones that control that data have AI plans of their own. Even then I could see OpenAI providing service to many of them rather than each of them creating their own models.
Re: It's Time to Stop Taking Sam Altman at His Word
#248To recap OpenAI's decisions over the past year: * They burned up the hype for GPT-5 on 4o and o1, which are great step changes but nothing the competition can't quickly replicate. * They dissolved the safety team. * They switched to for profit and are poised to give Altman equity. * All while hyping AGI more than ever. All of this suggests to me that Altman is in short-term exit preparation mode, not planning for AGI…
It does look like an exit. Employees were given the chance to cash in some of their shares at $86 billion valuation. Altman is getting shares. New "investors" are Microsoft and Nvidia. Nvidia will get the money back as revenue and fuel the hype for other customers. Microsoft will probably pay in Azure credits. If OpenAI does not make profit within two years, the "investment" will turn into a loan, which probably mean…
> If OpenAI does not make profit within two years, the "investment" will turn into a loan, which probably means bankruptcy.
I don’t believe this is accurate. I think this is what you’re referring to?:
Under the terms of the new investment round, OpenAI has two years to transform into a for-profit business or its funding will convert into debt, according to documents reviewed by The Times.
That just means investors want the business to be converted from a nonprofit entity into a regular for-profit entity. Not that they need to make a profit in 2 years, which is not typically an expectation for a company still trying to grow and capture market share.
Source: https://www.nytimes.com/2024/10/02/technology/openai-valuati...
Re: It's Time to Stop Taking Sam Altman at His Word
#249Earlier quoted context omitted.
Dissolve the safety team. You just made everyone stop reading the rest of your post by falsely claiming that
I don’t even know how this fake news started
Maybe the fact that they actually did dissolve the safety team formerly led by Ilya Sutskever in the aftermath of Altman's coup [0]? I'm genuinely unsure what part of this you're questioning.
[0] https://www.bloomberg.com/news/articles/2024-05-17/openai-di...
Re: It's Time to Stop Taking Sam Altman at His Word
#250tl;dr author complains that Sam's predictions of the future of AI are inflated (but doesn't offer any of his own), and complains that AI tools that surprised us last year look mundane now. The article is written to appeal to people who want to feel clever casually slagging off and dismissing tech. > it appears to have plateaued. GPT-4 now looks less like the precursor to a superintelligence and more like … well, any…
> it appears to have plateaued. GPT-4 now looks less like the precursor to a superintelligence and more like … well, any other chatbot. This is such a ridiculous sentence. GPT-4 now looks like any other chatbot because the technology advanced so the other chatbots are smarter now as well. Somehow the author is trying to twist this as a bad thing.
If everyone can catch up so easily, OAI has no moat and SamA is further full of shit asserting that they do.