A rambling but interesting dialog about AI

For me the practical issue about AI is how it can fund my expenses via the stock market. First off, I believe that neural network based AI has finally Crossed the Chasm. This leap is necessary to reduce the risk and uncertainty of investing in the technology.

The two most important ingredients are data and the computing power to make good use of it. While Increasing Returns are based on low capital required for additional sales, AI is very capital intensive to collect and curate the data, to convert the data into intelligence, and to put it to practical use (inference) in the real world.

Large Language Models (LLM) seem to have exhausted Goggle’s store of words (I don’t know if this is true). Where to get data? Companies like Google, Meta, Facebook, and X have their private sources. It’s these four that are mentioned in the linked video as the top AI contenders. How do you monetize LLMs? In oral applications like Siri and Alexa. Smart phones? Help desks (Citi’s customer service AI assistant sucks)?

I’m pleased by the fact they didn’t include Tesla. The omission helps to put a lid on the share price, it’s expensive enough already. What makes Tesla AI attractive is that they have the cash flow to finance it and visual data collection like no other.

There is also synergy in the Musk empire. Tesla can licence X’s LLM to enable passengers to talk to their CyberCabs and Optimus users to talk to them.

Elon vs. OpenAI: The Battle Over For-Profit AI w/ Salim Ismail | EP #138

Enjoy!

The Captain

3 Likes