wilbur_whateley

Then why do I feel like me and not this larger structure?

I think humans and other life forms have this ability on a long enough timeline through evolution.

To me, AI is about the ability to solve novel problems. People who think supervised learning can do that usually believe that new knowledge comes from (1) combining old ideas, and (2) trial and error. I don't think that's true. If you think about the state space of all possible ideas, it's clear that humans can explore a very small subset of that space and supervised learning just imitates humans (with great speed and memory). RL also doesn't work, but it at least attempts to learn independently of human knowledge.

How did he punish them?

My own experience. I'm working on something complex that's not in the datasets these models were trained on. There I see V4 flash breaking down and hallucinating much more often than GPT/Claude. For normal, common tasks, I also don't see much of a difference.

V4 flash is much worse than any Claude model. If you're doing something simple, it can be a good way to save money though.

Games from that time had so much soul. I wish I were better at the art side of things.

You're right, I'll add that. You remember them after playing for some time, but it is useful for a new player.

Yes! Special abilities (Queen-Maenad and Pirate-Pirate) deal 20 damage to neighboring pieces. The only difference between 10 HP pieces and 30 HP pieces is that 30 HP pieces can survive that.

Claude with Sonnet medium effort just used 100% of my session limit, some extra dollars, thought for 53 minutes, and said:

API Error: Claude's response exceeded the 32000 output token maximum. To configure this behavior, set the CLAUDE_CODE_MAX_OUTPUT_TOKENS environment variable.