Sam Altman on What People Have Wrong About AI

OpenAI CEO Sam Altman shares his view on what might surprise people about AI in a conversation with Greylock partner Reid Hoffman.

#samaltman #openAI #chatgpt #gpt4 #greylock #AI #reidhoffman #venturecapital #shorts

[mikeypoop]

When AI begins to solve age-old mathematical problems, like the twin prime conjecture, and enlarge our circle of mathematical knowledge generally, I'll be absolutely floored. And I'm sure it will eventually..

[TheIgnoramus]

Clip cut is kinda bad, but he’s saying it has the potential to creat new knowledge. It doesn’t have to be sentient, just able to recognize holes in its knowledge and fill them.

[declup]

AlphaFold is a good example of the potential productivity of advanced machine-learning applications. It isn't a NLP model, but the technologies are very closely related. NLP models are likely to have similar upper limit of utility.

[Muller_Andr]

I’m opportuned to have guidance and supervision from some of Wall Street’s eyes and ears. This is an exciting time of year after a tough one all round. I made a significant profit with my MSFT holdings recently and have drawn almost 100k for the holidays.

[ultravidz]

No human has the capacity to connect dots across all scientific literature. Generating new insights from existing data is where AI thrives.

[huzle]

Well it already has, hasn't it? I read that AI discovered folds in protein that humans would've never known about. What this information means; I don't have a hecking clue.

But, that's enough to tell you the potential for AI. That's not even me adding the nano robots that'll help with complex surgery and help us see the human body in a new light.

[briancase]

He's right. These LLMs are already doing this, and they aren't the last word in AI.

[sharannagarajan]

I didn’t think Sam Altman also underestimated AI

[NBTJacklyn]

Love how accessible u make design
I’m not a designer but I love thinking about those things and seeing the BTS process

[revhthboma]

It can totally add to the total of human scientific knowledge! It can learn from itself.

[stuartadams]

The thing that turns out to be wrong is AI not killing us all thanks to OpenAI

[JobjobJob-fe]

What he is saying is that people underestimate AI
AI can in fact surpass human knowledge ❤

[mylittleheartscar]

Imagine Ai simping for Ai then we have to figure out how to undo all the lewds we made

[godmisfortunatechild]

I think what will turn out to be wrong is the expectation of a fair and equitable distribution of AI to all of humanity and not just the elite.

[galidornelkenmeer]

Yeah because when it actually starts doing those things, we are done as a species.

[Meditationmonk]

It can create new stories, art, music. sure it will eventually create new knowledge.

[senju]

To GPT-4: "I am going to buy this thing"
GPT-4 : Sounds good. But don't forget to buy this and this and
To Myself: WTF...this is way beyond trying to predict the next word.

[shayraecok]

Wait he days the people saying the second part are wrong?
So he thinks AI could cure cancer or not ?

[boringmanager]

We'll all be dead before we count 1-2

[kumaramit-qc]

He's expecting that someday AI will surpass God