My "try to understand" take: We subsidize corn, then use it yo make a less efficient fuel. The money involved in this process likely takes away from subsidies to other forms of energy. There are a great many activities we do not subsidize, but solar is one that if we did, would produce an outsized benefit to society. And the more we do, the better. Redirecting an ethanol subsidy to solar would be a far more beneficial long term strategy for energy independence and overall standard of living in the US. Going all in on Solar would be a transformative and likely relatively short investment period that would last and benefit a long time. We have done many large scale infrastructure projects in the US, and it is frustrating to see the resistance to this one, being both less disruptive and more "all around win" than any other i can think of.

Feel little like this is generated and not based on experience. Claude.md should be short. Typescript strict mode isnt a gotcha, itll figure that out on its own easily, imo omit things like that. People put far too much stuff in claude, just a few lines and links to docs is all it needs. You can also @Agents.md and put everything there instead. Dont skills supercede commands? Subagents are good esp if you specify model, forked memory, linked skills, etc. Always ask what you can optimize after you see claude thrashing, then figure out how to encode that (or refactor your scripts or code choices).

Always separate plan from implementation and clear context between, its the build up of context that makes it bad ime.

The intro paragraph sounds exactly like Claude’s phrasing. So much so that I couldn’t read the rest of the article because I assumed I could just ask Claude about the topic.

"The gap is larger than most people realize" is a a dead giveaway.

Exactly this. If there is some nuance in the article vs what Claude can tell you, then that's worthwhile. This article is just generated with a specific prompt on style but very little content editing. What's the point? It's like posting the results of a Google search. The prompt would have been more interesting.

It's not against the rules to post AI slop here, and I don't necessarily think it should be. But I do wonder how we value written content going forward. There's value to taste and style and editing and all the other human things... there's very little value in the actual words themselves. We'll figure it out.

> Dont skills supercede commands?

Don't skills sit in context while custom slash commands are only manually invoked?

The difference isn't clear to me, especially since, upon googling it right now, I see that skills can also be invoked with a /slash.

They’re the same thing now except that skills can have some frontmatter to allow the agent to execute them automatically.

i also worry but am also shocked how far a single $20 sub gets me on side project. i pay for 3 (cc, codex, gemini) but am almost never going beyond cc, even when im merging several prs a day.

This is really interesting; ive done very high level code maps but the entire project seems wild, it works?

So, small model figures out which files to use based on the code map, and then enriches with snippets, so big model ideally gets preloaded with relevant context / snippets up front?

Where does code map live? Is it one big file?

So, I have a pro@coder/.cache/code-map/context-code-map.json.

I also have a `.tmpl-code-map.jsonl` in the same folder so all of my tasks can add to it, and then it gets merged into context-code-map.json.

I keep mtime, but I also compute a blake3 hash, so if mtime does not match, but it is just a "git restore," I do not redo the code map for that file. So it is very incremental.

Then the trick is, when sending the code map to AI, I serialize it in a nice, simple markdown format.

- path/to/file.rs - summary: ... - when to use: ... - public types: .., .., .. - public functions: .., .., ..

- ...

So the AI does not have to interpret JSON, just clean, structured markdown.

Funny, I worked on this addition to my tool for a week, planning everything, but even today, I am surprised by how well it works.

I have zero sed/grep in my workflow. Just this.

My prompt is pro@coder/coder-prompt.md, the first part is YAML for the globs, and the second part is my prompt.

There is a TUI, but all input and output are files, and the TUI is just there to run it and see the status.

This isnt unique to top AI researchers. Top talent has a long history of being averse to authoritarian/despotism at least in part because, by near definition, it must suppress truth. You cant build the future effectively with that approach.

It does. But then, it's how i talk to myself. More generally, it's how i talk to people i trust the most. I swear curse and insult, it seems to shock people if they see me do it (to the llm). If i ask claude or chatgpt to summarize the tone and demeanor of my interactions, however, it replies "playful" which is how im actually using the "insults".

Politeness requires a level of cultural intuition to translate into effective action at best, and is passive aggressive at worst. I insult my llm, and myself, constantly while coding. It's direct, and fun. When the llm insults me back it is even more fun.

With my colleagues i (try to) go back to being polite and die a little inside. its more fun to be myself. maybe its also why i enjoy ai coding more than some of my peers seem to.

More likely im just getting old.

1. Best part of this (satirical) post is, the service they offer isn't really needed. LLM's can do this already for small projects, and soon likely will for large ones too. You don't need a company to do this, we all have the LLM tooling to do it. Critical we're all spending time thinking about what that means in a thoughtful way.

2. For the sake of argument assume 1 is completely true and feasible now and / or in the near term. If LLM generated code is also non copyrightable... but even if it is... if you can just make a copyleft version via the same manner... what will the licenses even mean any longer?

I believe the latter was key and democratization of the armed forced was at least one reason the dictators (Julius etc) were able to maintain loyalty of their troops as they eliminated the republic as such. Im not super well read here though.

Higher level, Fukuyamas political order series does a great deep dive into these kinds of topics, really blew my mind, and made many archaic seeming political structures make far more intuitive sense to me afterwards.

I’ll have to give those a read. For others, I’ve linked their wiki pages below:

https://en.wikipedia.org/wiki/The_Origins_of_Political_Order

https://en.wikipedia.org/wiki/Political_Order_and_Political_...

Yeah once a given senator commanded AND paid the local full time legions for a given region, their loyalty followed.

By that time too the men serving in the legions were increasingly locals and not Roman or even from Italy.

Right but LLM companies are building frontier models with frontier talent while trying to sock up demand with a loss leader strategy, on top of an historic infrastructure build out.

Being able to coat efficiently run frontier models is i think, not a high priced endeavor for an org (compared to an individual).

IMO the proposition is little fishy, but its not totally without merit and imo deserves investigation. If we are all worried about our jobs, even via building custom for sale software, there is likely something there that may obviate the need at least for end user applications. Again, im deeply skeptical, but it is interesting.

> Being able to coat efficiently run frontier models is i think, not a high priced endeavor for an org

Running proprietary model would make you subject to whatever ToS the LLM companies choose on a particular day, and what you can produce with them, which circles back to the raison d'etre for the GPL and GNU.

Until all software copyright is dead and buried, there is no need for copyleft to change tack. Otherwise there rising tide may rise high enough to drown GPL, but not proprietary software.

Open source is easier to counterfeit/license-launder/re-implement using LLMs because source code is much lower-hanging fruit, and is understood by more people than closed-source assembly.

I agree with you on non judgement but would push back - if you'll violate your principles for a cush job, they aren't really principles you have.

Even though I strongly agree with the other person about reasons why people wouldn't leave...

I agree even strongly with what you just said: "if you'll violate your principles for a cush job, they aren't really principles you have."

The reality is, I don't think people really understand what a deeply held principle is. It's often a non-negotiable.

And then sometimes you have to question your principles and perhaps let them go. This can happen, for example, when children grow up and become adults. Their parents _should_ do a lot of letting go.

Perhaps folks involved with electronic devices are too used to a black & white decision world. Computer says no or computer says yes, there is no maybe. The real world of principles, morals, emotions, humans etc is filled with maybes and that can become hard to navigate for computers.

Agreed. But in that case, it's no longer a principle because, as you said, it's been let go.

The point above was about prioritizing something above the principle one still keeps.

I work in marketing, nothing black and white about it.