AI's Investment Implications

Nvidia is fundamentally a Capex story. Companies in a rush to buy Nvidia hardware for their AI buildup. So it’s really curious, to me anyway, to see an allegedly AI company that didn’t spend anything on Capex.

PLTR is also a company talked about a lot among stock promotors and finance people. Don’t see much talk about it among the technical folks. Reminds me a lot of the Tesla fanboys talking nonsense about Tesla’s AI self driving systems or Dojo AI superchips. No technical content. Just cult worship.

:man_shrugging:

.

SMH.

.
Who are the companies that need AI chips? What are the chips for? When do they need chips? Where do they need chips? Why do they need chips?

Hint:

Hint2:
Who are the ultimate beneficiaries of the internet infrastructure?
Do you really need H100s? May be something cheaper and a great AI OS platform would be cheaper and better.

1 Like

Creative industry is not spare.

AI hallucinates and louses up. If humans have to 100% verify output what’s the productivity gain?
The creative arts don’t have that problem.

Movies are watched and edited many times before releasing.

Thought NVDA makes most $$$ from training chips, not inference chips? So why would an inference chip a threat?

Should we trust Chamath? His explanation of LPU is no difference from GPU.

Groq? Grok? Why not just use God?

There’s huge money spent on inference for now. It’ll end up on devices.

2 Likes

But AI can make the movie first. Saves a bundle on production and overpaid actors.

1 Like

2.5 years left of Beth’s prediction Nvidia will surpass Apple in valuation.

I have sold all AAPLs and bought NVDA :money_mouth_face: LFG.

Joking or true?

Joking. AAPL is my dividend stock. Need those passive income :slight_smile: I did sell some in 2020 to buy growth stocks… diworsification :face_with_symbols_over_mouth: Easy to look back, should have invested in TSLA, should have invested in NVDA, should have invested in SMCI, should have invested in crypto miners, … The truth is have I done so for potentials, my NW would go down the drain and I have to work 925.

1 Like

Interesting observation.

1 Like

It’s great OpenAI’s legal argument isn’t to deny doing it. It’s to argue that the NYT cheated to get the evidence. I don’t think that works in a civil lawsuit. NYT isn’t a law enforcement entity. OpenAI actually admits it’d be impossible to accurately train their models without unauthorized use of copyrighted material. I wonder how financially viable they’d be if they had to pay for all the data to train.

“Because copyright today covers virtually every sort of human expression—including blog posts, photographs, forum posts, scraps of software code, and government documents—it would be impossible to train today’s leading AI models without using copyrighted materials,” OpenAI wrote in a filing last month in the U.K., in response to an inquiry from the U.K. House of Lords.

“Limiting training data to public domain books and drawings created more than a century ago might yield an interesting experiment, but would not provide AI systems that meet the needs of today’s citizens,” OpenAI continued in the filing.

1 Like

https://twitter.com/elonmusk/status/1762265469195886878

1 Like

OpenAI chatbot doing the work of 700 human agents.

4 Likes