in
I Was Wrong About AI
Brian shares his current views on AI. The episode title is technically correct.
Show Notes (click or tap to expand)
To atone for my clickbait title, here are 10 of current thoughts on AI/LLMs: 1. It's very impressive software, unbelievable how well it works 2. It's not sentient and never will be (not improving exponentially) 3. It's not even autonomous. There will always be a human needed at the start (prompts, direction) and end (correcting mistakes) 4. The only way 2 and 3 are wrong is if this software summons demons. There is an above zero chance it's demons 5. Either way, I'm avoiding it and so should you 6. AI is not inevitable, you do not have to use it 7. It's a disaster for humanity if AI works well 8. It's a disaster if AI doesn't work well but people continue to pretend that it does 9. The past three years of failed predictions by alarmists have numbed people to the actual detrimental effects 10. AI doesn't create new problems, it exacerbates existing problems Not to mention the power costs (my main problem with cryptocurrencies) and content theftFeedback
Email tto@koser.us