I'm starting back into the job hunt, and I'm realizing that alot of potential employers want to know about my approach and thinking with regards to generative AI tools, so I feel compelled to summarize that here.

The short answer, is that I use them. Specifically, I use the open source harness pi, and I alternate between small models like mistral and GLM depending on the task.

And also, I still in fact write code by hand. I also read code. I also talk about code, and expect my coworkers to be able to reason over code.

The reason for that is simple: I enjoy it. I don't actually think you need any other reason.

The reality is that I do not ship code I do not understand, and also, I do not restrict my understanding to the first order. Good engineers learn to read high volumes of code NOT because they form first order understanding of every line, but because they're building internal story / representation that likely just follows the continuum of sparseness between arbitrary literal constructs and metaphorical relations between them.

And that's fun! It feels good to construct internal representations that increase what one learns while being consistent with one's basis.

In an ideal scenario I wouldn't have to have a public moral or philosophical stance. I'd just do good work and enjoy it, around others who feel the same way. But we live in odd times, and it becomes increasingly impossible to avoid these questions even to find a job, I guess I have no choice but to be clear and stake this out simply.

I don't care about your theory of consciousness and I find the only useful interpretation is that it is ambiguous. Ambiguous. If your marketting or business theory requires me to carry your theory of consciousness, I'm going to have a boring time ducking your bullshit.

That said, I have no problem with these fields that have long histories of practical application: Cybernetics, Robotics, Higher-Order constructs of all kinds, Ecology, and Thermodynamics.

I also just enjoy simple things, saying I don't know, not trying to save or fix the world, but merely being curious in it. I think we can still be productive as Infinitesimals, and I hope to do just that.

So yeah, I'll use your LLM and I'll write agents or whatever to solve problems. But I will still hold one part of the problem, for me to enjoy and be curious about, and whatever level of detail I decide. I will not compromise that.