I feel you on this one. I think that if AI continues at it’s current pace, climate change will sorta become like the Black Death sometime in the 2030s. That is, a big problem that is easily solvable with modern tech.

Some people just wanna believe Armageddon is coming. They may cite data, but at the end of the day, they actually just don’t want to let it go. I think it is a waste of time to try to reach these people. I just let them have it.

But then there’s others who just have bad data or no data on how tech is progressing. For them, citing cost declines and exponential trends and why they are happening can be very useful

For example, I’m not just blindly optimistic on AI progress through the next decade. Instead, I’m noticing that while the traditional Moore’s Law around transistor density seems to be slowing, ASICs and algorithmic improvement are more than making up for it. In terms of how much intelligence you can build for some set amount of money, the exponential actually seems to be SPEEDING UP. And now that large, somewhat general models have commercial viability, there are MASSIVE financial incentives to continue progress.