Gary’s Q&A

Listener question & Gary’s response · Transcript passage

Derrin Brown asks: You’ve mentioned before that you’re unsure who the winners will be in AI. Now that we’re seeing reports of companies like Uber burning through their entire annual AI budget in just four months, and critics arguing that firms like OpenAI and Anthropic can’t actually afford their own compute costs, do you think the current ‘tokenomics’ model, where companies charge by usage, but the underlying infrastructure requires massive, ongoing debt, is fundamentally sustainable? Or are we looking at a structural bubble similar to what we saw with the dot-com crash?

I have no idea what the end business model for AI companies will be. I’m not even sure what the AI industry is going to look like.

Things are changing at a fast pace, and I don’t think even AI companies know what the future holds. Not only are AI models getting more powerful, but they are also getting cheaper and lighter. It is now possible to run some LLMs on desktop computers.

I do think we are in an AI bubble, and honestly, that is what is to be expected. Everyone can see the potential of artificial intelligence, just like how in the past everyone could see the economic potential of the internet and railroads.

Lots of money is going into AI, just like money flowed into railroads and internet companies before. During those bubbles, most of the railroads and internet companies failed, but the ones that survived became giants.

The valuations AI companies are getting are not justified by the revenue they are bringing in or their profits, which are nonexistent.

I’m not an investment expert, but it seems like there will be a correction at some point. I don’t know when or how big it will be, but it seems inevitable.