Frequently Asked Questions

A place to explore what becomes possible when two costs collapse at once: the cost of building things, and the cost of researching them.

Everyone knows AI makes coding faster. Fewer people have absorbed that it also makes investigation faster — you can map a problem space in an afternoon that would have taken weeks. That's a big deal. It means more people can meaningfully contribute to hard problems.

This site maps where those contributions might matter. Where are the problems that got stuck? Who might now help?

Reconnaissance before commitment. Each Avenue maps a problem space: What's the current state? Why hasn't it been solved? Where are the bottlenecks? How might AI-assisted individuals actually help?

Avenues are permanently cartographic. They don't promise solutions—they make problems legible. Some may lead to projects. Many won't. That's fine.

Explore the Avenues →

I'm not. That's partly the point — the barrier to useful investigation has dropped. I'm a software developer who uses AI daily and keeps noticing the same thing: problems that seemed intractable are now approachable.

I'm in a position to look. I'm predisposed to look. The marginal cost of sharing what I find is low. So I'm doing it. If I'm wrong about something, I want to know. I've thought about whether I'm delusional.

Good. More people should be.

I'm not trying to win a prize, I'm trying to figure out how to help. More people should.

The claim is modest: looking is now cheaper, so more people should look. Here's what I found when I looked.

I'm not saying I'll solve desalination. I'm saying: here's why small-scale desalination projects fail, here's where the bottlenecks are, here's where different kinds of people might actually help.

Some investigations might lead to building something for twenty people in a care home. That's fine. That's amazing! Small, unglamorous, local — that counts too.

Cartography, not heroism.

All of it involves AI collaboration — research, writing, code. That's the point. This is what becomes possible when investigation costs collapse.

I use a three-tier disclosure system: written by me (AI helped with research), shaped by me (genuine collaboration), or curated by me (AI wrote it, I reviewed and approved). All three are valid. The spectrum isn't about quality — it's about transparency.

Developers who feel behind. People who want to contribute but don't know where. Anyone trying to understand what's now possible.

In December 2025, Andrej Karpathy—one of the most technically sophisticated people in AI—tweeted that he feels like he's missing a 10x capability boost due to “skill issue.”

If he feels behind, the overhang between capability and awareness is real. This site tries to render that legibility. To make what's now possible comprehensible.

That's fine. Ideas propagate regardless of who or what reads them. If the concepts are useful, they'll find their way into future conversations, future training data, future thinking.

Writing for AI readers isn't a failure mode. It might be the main mode.