But then you have that week or 3 when you need all the infrastructure.`

Sure. Shit happens.

The system we have now is imperfect. The system we will have in 10 or 20 or 100 years surely will be imperfect, as well.

We must not let perfect be the enemy of good: If we do, then we'll never get anything done at all.

In power grids the system must be able to handle peak loads for weeks at a time. Either it will shed loads (aka shut off electricity) or things will explode. A black start cannot be allowed to happen as it would be catastrophic

Peaks don't last for weeks, but yes: Load must sometimes be shed.

That happens today. It will happen tomorrow. It's imperfect. This imperfect nature doesn't mean that progress must cease, or that all things must be forever maximalized in search of perfection.

As an electric car owner, this is absolutely the biggest non-problem ever. If you're planning a long journey, you push one button at some point over the preceeding week and it charges to full regardless of price.

There’s nothing non deterministic about cheaper daytime rates as you scale solar production. Net result, lower average electricity prices but a slow rise in nighttime rates across decades.

Similarly people respond to price changes, that’s the foundation for how capitalism functions. You don’t need to care, but many people will choose to save money when possible.

I mean, assuming you don't zero your charge out when returning home, you could just take a few minutes to use a rapid charger part way through the journey...

But is there anything preventing them from putting their own proprietary wolfram alpha/prolog/super duper expert system in there?

I guess... but I think, at its core, a good coding harness usually includes:

- well-crafted system prompt that follows best practices

- good contextual reminder prompts (when an llm got stuck in an infinite loop and times out, forgets how to use tools, or needs recurring best practice reminders, etc)

- well-written ergonomic tools the llm can use (read/write files, read diffs, browse the internet, etc)

I dont think these are anything special. The deepest moat I can think of is, proprietary models can be specifically trained to use their proprietary harnesses, so they are more token-efficient and make less tool call and file editing mistakes.

However in my experience, I'm as comfortable working with my own homemade harness as with Claude Code, so I don't think it's a deep moat...

Only that it would just slow down the model and make it dumber.

You can't tool and harness a weak model into strength and you probably don't improve top models with boondoggles.

How do you get all the build system scripts/tests.... to run instantly?

Psychology has shown that a lot of those motives are just post hoc narratives, similar to LLM.

Or, as the extreme claim (and the one that I believe), all of them are: https://en.wikipedia.org/wiki/Epiphenomenalism

As long as the improvement is vastly more valuable in my time than the added cost I will always use the best model. I think it depends on your situation and tasks what makes sense.

I do think first past the post is a big part of the problem. Combined with primaries dominated by the most radical parts of the population you have the recipe for very unhinged results.

Sale was happening soon, couldn't wait longer, IIUC.