Thoughts on thinking

I am somewhat in disbelief as my reaction to AI tools has been pretty much the exact opposite to the author’s. I had thousands of short notes, book ideas, etc. before AI, most of which were scattered in a notepad program.

Since AI, I’ve made genuine, real progress on many of them, mostly by discussing the concepts with ChatGPT, planning outlines, creating research reading lists, considering angles I hadn’t considered, and so on. It’s been an insanely productive process.

Like any technology, AI tools are in some sense a reflection of their users. If you find yourself wanting to offload all thinking to the machine, that’s on you, not the tool.

What we need is a type of block chain that only records new lines of thought, and ascribes a kind of novelty index for concepts. That way researchers can stay on their intended side of the boundary of human knowledge, either well-trodden, averaged prior art or exploratory and theoretical. Then ambitious thinkers can gamify searching for truly new tokens.

The publication process seems to already basically function this way, no? Whether through books, academic papers, or even just a blog post. If you publish something first online and it gets indexed, that is the best we can probably do - at least while human thoughts are still stuck in human heads.

Thinking more about distilling information down in a compressed form like the LLMs. Then the uniqueness of ideas, not actual language like copyright enforcement, could be measured. Established ideas are common and diluted by human and synthetic authors. While concepts out on the edge are measured as a novel combination and might be worth more exploration. Just basing this on the personal conviction of "ask not what the sum of all human knowledge can do for you, but what you can add to the sum of all human knowledge."