Top
Best
New

Posted by meetpateltech 10/28/2025

The next chapter of the Microsoft–OpenAI partnership(openai.com)
366 points | 480 commentspage 3
falcor84 10/28/2025|
I just want to say how nice it was to read those clear bullet points in this press release. I know that bulleted lists have been getting a lot of flack because of AIs overusing them, but it's really nice sometimes to not having to go treasure hunting in annoying marketing prose.
rvz 10/28/2025||
> Once AGI is declared by OpenAI, that declaration will now be verified by an independent expert panel.

By the time we get 30% global unemployment and another financial crash along the way in the next decade, only then OpenAI would have already declared "AGI".

Likely with in the 2030 - 2035 timeframe.

anonymous908213 10/28/2025||
> AGI AGI AGI AGI AGI AGI AGI AGI AGI

Spare me. Sam has been talking about ChatGPT already being AGI for ages, meanwhile still peddling this duplicitous talk about how AGI is coming despite it apparently already being here. Can we act like grownups and treat this like a normal tool? No, no we cannot, for Sam is a hype merchant.

respondo2134 10/28/2025||
it's notable that there is no talk about defining what exactly AGI is - or even spelling out the three letter acronym - because that doesn't serve his narative. He wants the general public to equate human intelligence with current OpenAI, not ask what does this mean or how would we know. He's selling another type of hammer that's proving useful in some situations but presenting it as the last universal tool anyone will ever need.
cogman10 10/28/2025|||
And because it's become apparent that LLMs aren't converging on what's traditionally been understood as AGI.

The promise of AGI is that you could prompt the LLM "Prove that the Riemann Hypothesis is either true or false" and the LLM would generate a valid mathematical proof. However, if you throw it into ChatGPT what you actually get is "Nobody else has solved this proof yet and I can't either."

And that's the issue. These LLMs aren't capable of reason, only regurgitation. And they aren't moving towards reason.

dagss 10/28/2025||
When I ask Claude to debug something it goes through more or less the same steps I would have done to fine the bug. Add some logging, run tests, try an hypothesis...

Until LLMs got popular, we would have called that reasoning skills. Not surpassing humans but better than many humans within a small context.

I don't mean that I have a higher opinion about LLM intelligence than you do, but perhaps I have a lower opinion on what human intelligence is. How many do much more than regurgitate, tweak? Science has taken hundreds of years to develop.

The real question is: When do knowledge workers loose their jobs. That is close enough for "AGI" in its consequences for society, Riemann hypothesis or not.

joomla199 10/28/2025|||
Did you read the whole thread and all of your own comment each time you had to type another half-word? If not, I’m afraid your first statement doesn’t hold.
parliament32 10/28/2025|||
AGI is pretty clearly defined here: https://openai.com/charter/

> OpenAI’s mission is to ensure that artificial general intelligence (AGI)—by which we mean highly autonomous systems that outperform humans at most economically valuable work—benefits all of humanity.

So, can you (and everyone you know) be replaced at work by a subscription yet? If not, it's not AGI I guess.

interactivecode 10/28/2025|||
Sam is doing the same playbook Elon used Tesla's full self-driving dreams
respondo2134 10/28/2025|||
but with 10x or 100x the chutzpah
vultour 10/28/2025||
This entire house of cards is built on the expectation that "AGI" is just around the corner. The moment Altman relents in his grift is the moment the bubble pops and we're in for a wild ride.
Oras 10/28/2025||
> OpenAI has contracted to purchase an incremental $250B of Azure services, and Microsoft will no longer have a right of first refusal to be OpenAI’s compute provider.

So OpenAI could be on Google (GCP) and AWS, and possibly Claude and Gemini on Azure? that could be a good thing.

I use OpenRouter in multiple applications, the practicality of having one provider to host all possible LLMs is such a win to try and iterate without having to switch the cloud (big for enterprise who are stuck with one cloud provider)

Shank 10/28/2025||
The API offerings are still only on Azure. It just means OpenAI doesn’t have to buy compute exclusively from Microsoft.
easton 10/28/2025||
They also say "Non-API products may be served on any cloud provider.". I wonder what products they are thinking about. If I sell you a EC2 image with GPT-5 on it, is that a API?

My assumption is that they mean PaaS model hosting (so azure's ai service, bedrock, vertex), but I don't know what other product OpenAI is thinking about selling via a cloud provider unless it's training tooling or something.

nostrebored 10/28/2025||
Open source models are the current product line that fits the bill.
lysecret 10/28/2025||
Yea I use gcp vertex for this it already has Claude hope for OpenAI’s models eventually too.
reducesuffering 10/28/2025||
I loathe that creating a non-profit organization supposedly guided by a charter to "ensure that artificial general intelligence (AGI) benefits all of humanity." is actually about being a $500b company corporate capital ownership with million $ pay packages. I mean, it looks like you really can do whatever you want if you have the most $ for the best lawyers and the gumption to not give af.
spumpydump 10/28/2025||
A tangent, but it feels more and more like the AGI maximalists of 2025 are by and large the NFT maximalists from 2022 (who in turn were the NoCode maximalists of 2020) that are looking for the next metaphorical penny stock to sell.

That logical fallacy of, “I spent a week teaching myself this topic and now I’m ready to talk about it like an expert.”

nalinidash 10/28/2025||
Shows how much they valued 'AGI' wrt how we valued it in the textbook. https://techcrunch.com/2024/12/26/microsoft-and-openai-have-...
porridgeraisin 10/29/2025||
Aside from the investor fodder about AGI which was quite funny to read, the part where OpenAI is able to independently pursue consumer hardware is interesting. It points to OpenAI potentially entering wearables space soon - remember they hired Jony Ive?
bongodongobob 10/28/2025|
I don't understand what MS is doing. The only AI available at work is M365 Copilot. It's absolutely terrible. Tiny context window, super guardrailed, can barely handle a 100 line PowerShell script. It's so so bad. I don't get it.
More comments...