191–200 of 298 posts
Let me translate this blog post out of corpspeak...: OpenAI has agreed to use Azure exclusively and for free, in return for Azure massively scaling up it's GPU capacity (which OpenAI was hitting before, hence threatening to move). In return, OpenAI gives an extensive license to Microsoft to use ChatGPT in Bing (and other products). Now why couldn't the blog post have just said that rather than hiding all the details?
Earlier quoted context omitted.
Because they didn't want to openly say "Microsoft gives us GPUs and allows us to use them for free, hence practically buys us. In return we almost exclusively license what we produce to them, for free." Because it looks bad. Microsoft wants to be seen in a different light, and OpenAI wants to be perceived as an open, academic, capped-profit for a noble reason collective. It's all smoke and mirrors.
> OpenAI wants to be perceived as an open, academic, capped-profit for a noble reason collective. Does anybody actually believe this? Their name has been a running joke for years already.
I'll list some pros/contras in no particular order:
CONTRA: I can't download ChatGPT and run it locally. Why not? Clearly they are hoarding it.
PRO: They genuinely overtrain toward responses that the model judges benefit humanity, to the point of failing at the task. I can't think of a specific example but everyone knows what I mean. It is a frustration to users and doesn't help their image. They seem to be overtraining this way out of the goodness of their hearts.
PRO: its non-profit or capped-profit structure would let it behave in the way it claims it is behaving. (By contrast, a for-profit public company in some sense has a fiduciary responsibility to maximize value for shareholders rather than benefit for humanity.)
PRO: ChatGPT is available for free and Dall-E gives some free credits.
CONTRA: Dall-E doesn't give enough free credits. It might have a longer wait time for free users but it shouldn't hard cap them at a certain number of free credits per month. For example perhaps free users could not be subject to any reasonable usage limits (a few hundred images per day should be fine) but could have to wait 1 extra minute for each one.
CON: its architecture isn't open, it doesn't share how it trained its models, it doesn't publish its datasets and parameters that would let other companies achieve the same results (reproduce their work and have their own version), its research is closed off and unavailable.
CON: the company is not directly applying AI to proactively solving social problems.
CON: (debatable) some users who are not sources of revenue have been banned. From experimentong with a chatbot (not any hard legal limits just things like exploring their sexuality with the chatbot. It's just a chatbot, it would make sense to exclude such conversations from further training data or feedback, keep male and female researchers at OpenAI from having to review unsavory conversations, or even warn or at most temporarily ban the user. An account closure is super extreme for natural forms of interaction that don't go near legal limits. It does not seem like an open company. Alternative viewpoint: on the other hand, we keep certain flamebait subjects out of HN. Perhaps having certain subjects banable is necessary for their open mission at a larger level - specifically, users will stop going there for sex since they associate it purely with productivity, similar to how we associate HN with intellectual curiosity, which is a result of very heavy moderation here and without which this place would not exist in its form?)
Overall based on the above considerations, I think that OpenAI is not "open" but if it started publishing downloadable models for anyone to run, I would feel differently.
Google and Facebook are the clear leaders in AI and they been spending a lot of money building out their AI infrastructure over the last decade. You haven't heard much about Microsoft's investments other than them buying rights to OpenAI models. Seems like they could have saved a lot of money if they started investing years ago.
Google bought DeepMind for 400m 8 years ago. You could argue DeepMind is on the same level or above OpenAI.
I asked ChatGPT to write a poem based on this post. Here's what it wrote: --- Two hearts, two minds, working side by side In pursuit of a mission to make AI safe and wide Microsoft’s investment allows us to stay true To our capped-profit company, our mission, and you From cloud computing to real-world use We collaborate to ensure our systems are not abused OpenAI API, Azure’s OpenAI Service, and Designer too We’re wo…
Priceless. Maybe companies should get AI to write PR releases. They say the same thing, except more honestly and funnier.
I laughed at independent research. Why do they even bother..
Earlier quoted context omitted.
This company will not reach AGI. Let's be real there for a moment. This company doesn't even have a decent shot at Google's lunch if Google comes to its senses soon, which it will.
_startup has no shot once incumbent comes to their senses_ is a claim that I think HackerNews of all places would be cautious in believing too fully. Is it likely Google or others with large Research wings can compete with OpenAI? Very probably so, but I’m assigning a non trivial risk that the proverbial emperor has no clothes and incumbents like Google cannot effectively respond to OpenAI given the unique constraint…
Earlier quoted context omitted.
Yes & No. GPT ChatGPT: probably not. It's not hard for other big players to enter this space. It's mostly egg-on-face for Google that they haven't given that Google basically invented the model that OpenAI uses and has big versions internally. There's nothing fundamental stopping Google Docs from adding ChatGPT to their UI and getting way more consumer training data than OpenAI can get without a similar play, or for…
> There's nothing fundamental stopping Google Docs from adding ChatGPT to their UI and getting way more consumer training data than OpenAI can get without a similar play. OpenAI could get exactly the same (or more, idk) data by integrating into Teams, considering the Microsoft partnership.
My point is chatgpt isn't a high-moat advantage for text/q&a for microsoft. Their top competitors here have similarly huge UI footprint. In contrast, program synthesis has a much higher data moat.
Earlier quoted context omitted.
If they reach AGI, or more simply replace a chunk of workers with AIs, this isn't far fetched to reach these numbers.
If they reach AGI, the AGI isn't necessarily going to be happy to work for free.
Earlier quoted context omitted.
The Patagonia thing was not legit at all according to Adam Hates Everything: https://www.youtube.com/watch?v=0Cu6EbELZ6I EDIT by "not legit" I mean "not authentic"
Adam's argument seems to be: 1. If instead of donating the company he had left it to his kids he would have paid a lot in taxes (2:58) 2. He donated the voting shares of the company to a 501c4 that will remain controlled by his family and is allowed to lobby the government (4:10) 3. Normally when you make a donation you're giving up influence over what happens after that (4:25). 4. Other billionaires do other things…
It not necessarily not legit, but he managed to keep a 3B business under family control and bypass paying around 700M in taxes in doing so. So the altruistic messaging that it was donated to save the world is mildly two-faced. There is nothing wrong with it, but the news stories did leave out a few of the details. I only bring this up because the message about OpenAI being structured in such a way doesn't pass the smell test knowing SV/VC and the key players involved. Again I admit Im cynical and can very well be wrong, it also doesn't affect my life so why do I care, but I bring it up for conversation on HN because I feel its fair to discuss it and the possibilities. [1]
[1] https://fortune.com/2022/09/16/patagonia-founder-legal-tax-l...
Earlier quoted context omitted.
100x is a great return even for YC standards, but the best returns that business angels, VCs and YC have had is in the order of magnitude of 10000x (yes, ten thousand). So capping at 100x still makes it attractive for investors, yet leaves a lot of potential capital for the non-profit. As one example, Sequoia invested in Airbnb at $0.01 per share, and Airbnb's current stock price is $102, almost exactly 10000x return…
Maybe that is a bit too much $0.01 per mean share would mean 6.5M USD valuation (current mkt cap is 65Bn). Accounting for dilution in investment rounds, let's say 4 x 20% dilution, that is around 52% penalty in valuation. Roughly, their entering price would be around 3-4M USD valuation. I am not saying in any way that this is a low return also, I may be wrong on my calculation, please, be free to correct me! ; )
Using share prices side steps dilution, which is a problem when one linearly scales valuation increases to wealth gains.