NVIDIA having quasi-infinite demand for chips & Tesla having quasi-infinite demand for bots is probably correct TBH. If Tesla can truly manufacture a humaoid that can be completely generalized in almost all use-cases, the number of chips we'll need for training and inference is completely and utterly ridiculous JUST for the Bot use case. It literally makes no sense. You can manufacture labor at will, and that labor is powered by chips. I think NVIDIA has caught the first wave of the AI Industrial Revolution that will probably last AT LEAST 10 years IMO. NFA, but I don't see how it can be anything but that. AI is way too useful and way too powerful. I had a completely revolutionary moment with AI this week for my own business. This thing is seriously no joke. It's transformational.

To give a very concrete example of how it has transformed my workflow: When I record a video or create a piece of content, the ideation process is by far the most brutal part. The ideation stems from ingesting untold amounts of data/info about the world at any given moment, and then synthesizing that information in my head into a combination of something that's useful/valuable, something that's interesting, and something that my audience will appreciate. I literally do this every day, close to 10 hours, full stop. I try my best to turn off my brain on the weekends but sometimes it's futile. That's a massive amount of energy that I use to generate just one idea. Now, I have a project in Claude's Sonnet 4.5 that has my last 90 days' worth of transcripts from every single one of my videos, a large sample of transcripts of my most natural videos (the Cybertruck videos), and a large reference knowledge base of everything that I've ingested recently from an info/data perspective. This project can generate ideas on the fly that I'm most likely to find interesting based on my own interests, which will automatically align with my audience, but instead of spending an entire day trying to puzzle all the pieces together, Claude can give me hundreds of them in a day. And when I inevitably come across one in 5 minutes that would've taken me 3 hours, I can then use that same exact project to generate a script for me on a video that is damn near 90% of what I would've said anyway. After nipping and tucking for ~10 minutes, I have an interesting video concept that will hit every important variable I look for when creating a video, going from birth to a recorded video in 30 minutes. This is absolutely and totally insane. That means that in theory, I should be able to record 16 high quality videos PER DAY - vs one. I'm going to give it a shot and see what happens. Will probably never hit 16 but maybe 5? We shall see.