Not even close.

I stopped halfway through this paragraph and just googled Peter Putnam.

If I want an article to take hours to get to the point I'll go read a recipe blog.

I do wonder about Putnam's research though. Has it been looked into by experts in the field more recently? The article doesn't really give an answer to this.

That’s when it hit him: The goal is to repeat. Repetition isn’t a goal that has to be programmed in from the outside; it’s baked into the very nature of things—to exist from one moment to the next is to repeat your existence. “This goal function,” Putnam wrote, “appears pre-encoded in the nature of being itself.”

So, here’s the game. The system starts out in a random mix of “on” and “off” states. Its goal is to repeat that state—to stay the same. But in each turn, a perturbation from the environment moves through the system, flipping states, and the system has to emit the right sequence of moves (by forming the right self-reinforcing loops) to alter the environment in such a way that it will perturb the system back to its original state."

I'm a big fan of this line of thinking. I've been arguing for years that RL should be based in homeostasis and this seems right along those lines. I wish I could have talked with him!

And according to my friendly neighbourhood LLM, they're not immediately related.

---

* if it's simplification in the article is correct and I understand that simplification correctly

But this story is also a great reminder on the importance of readability. I can't name Putnam a genius if his works were so incomprehensible, that even best of the best had trouble deciphering them. Just like you wouldn't call programmer a genius if he wrote his best work in brainfuck.

I guess he never intended his works to be read. So I won't.

What seems clear is that he was in the right moment and with the right people to produce a great theory of the mind, he was thinking about defining the mind as a creator of heuristic for generating inductive rules, that is a very interesting idea, also that seems related to cellular automata and homeostasis, fixed point theory, error recovery and many other ideas that we, with our ability to look back in time, can see are located in their neighborhood. Recently we are using LLMs to produce heuristic or rules to generate python programs to explore proofs or new models, so the gems is still alive.

The field he was interested in it is full of little diamonds waiting for someone to find them, but perhaps he just put his foot in the door without entering (or creating) the field he was trying to envision. It could be that his paper could reveal some intuition that could guide us in a new search for meaning but from this post there is no any hint about that beyond such a desire.