AI's Investment Implications

Cathie insists she is right despite underperforming S&P on her favorite 5-year rolling period.

I keep seeing people comparing Nvidia with Cisco. But even if it’s right Cisco had a tremendous 10-year run during dot-com. We are only at Year 2 of the AI run. Isn’t it premature to wheel out the Cisco comparison at this moment in time?

Second, Nvidia actually has more software engineers on its payroll than hardware. Its crown jewel is not the GPU’s but CUDA. Meta is working to lessen its dependence on CUDA by making PyTorch portable but it’s not easy. CUDA is the accelerant in accelerated computing. It’s become industry standard, schools teach it, many thousands of professional programmers also knew it and a whole ecosystem has developed around CUDA. It’s like x86. Very difficult to dethrone.

1 Like

Saw many people said the energy in SF right now is a lot like back in the early 90s when the commercial internet was just getting started. It’s very interesting Hayes Valley turned out to be the world center of AI. It’s a very tiny residential neighborhood.

The energy is feeding on itself. Anyone who wants to build anything in AI will want to come to SF to tap into the network and energy, which makes it even more attractive to future AI founders. Flywheel will be spinning faster and faster.

1 Like


Huh? Thought is the last leg. Last leg is usually explosive… rocketing.

According to the YouTube, most ideas are about cars. Not interested. Also, not interested in looking at a display all day. I am interested in how to re-vitalize myself, become magnetic to women, has a sexy voice and ofc, immortality :slight_smile: Can AI helps?

AI startup activities really picked up after release of GPT3 api last year. Now every young founder wants to build AI. Before those api founders would need to build their own ai models. Most aren’t sophisticated enough to do that. They just know how to mix things.

That was like in the early 90s when every young founders wanted to build a website.

I don’t think we are at last leg, unless this AI wave is really tiny.

1 Like


Mosaic 1994. Dotcom busted Mar 2000. Since…

Hence we have 4 more good years of AI.

1994 -2000 were the best years.


Here you go! I guess this guy made “A-Balling” popular. I did not realize what it was until someone tried to do it to me at Mountain Charlie’s in Los Gatos. A coworker explained it to me. The guy was annoying AF. He finally gave up when I told him it’d take longer to walk outside than to deal with him.

As a young founder building with Ai you’re spot on lol


It is the last leg. Leg is not year… if similar to Dotcom, 4 more good years. Last leg is usually explosive, a blow-off top.

4 and 5 9s are the standard for business applications. Chat GPT will only guarantee 99.5% and that’s if you pay extra for dedicated hardware. I’m not sure the technology is ready for business applications that will scale.

Tag: AI
Sep 6, 2023.
Check again on Sep 5, 2028.

1 Like

Bullish for Microsoft and Nvidia.

That tool is very cool. I’m not sure about their monetization though. The free version is enough for most people. It makes it super fast to create a user guide.