Any semi-competent AI Agent can do that with a plan you've written in 5 minutes.

I may be wrong, but it seems when people are talking about easy glue code, they’re talking about web services API, not OS API, not graphics or sound API, not file formats libraries,…

I know it's a piece of shit API done in the worst possible way on purpose (they don't want openness, but had to fulfill a law that mandates "openness") because I had previously tried to do it manually - twice. I ran out of whisky before I got anything done.

Sonnet _3.5_ almost one-shotted it with just the API "documentation" they had and access to Python and curl.

People have also hooked stuff into proprietary APIs on "smart" devices with zero documentation, just by having an Agent tirelessly run through thousands of permutations to figure it out.

Grandparent said:

> The single best feature of any programming language is that it is precise.

C overtook a more precise language family because it has features other than precision that people cared about. Perhaps a better tradeoff of expressiveness and readability with precision.

Grandparent could be correct, and precision is the best feature of C, despite being less precise than ASM. And its better expressiveness nets out to a better overall programmer experience. I just wanted to point out that precision is something we do trade away for other things we want.

To me, the notion of precision isn't in any way related to whether any given statement is sound. It's about the behavior of the language for sound programs.

Your suggestion that because languages have specified undefined behaviour, they are somehow not precise, makes little sense.

In fact, voting with your feet and leaving is far more effective at fixing political issues than the democratic voting process.

Yes. I've done so myself. The fact that people do this all the time doesn't mean it's the best thing to do when your country has problems.

People also move houses all the time. It's a big undertaking. Not the default solution whenever your kitchen needs renovations.

> In fact, voting with your feet and leaving is far more effective at fixing political issues than the democratic voting process.

Citation needed. Sounds very defeatist.

I don't know enough about the domain to be objectively helpful, so it's all wishy-washy feelings on my part. I keep reaching for orchestrating things with threads in Rust where most people would probably reach for async these days. The only language where I've felt fine embracing the blessed async system is Haskell and its green threads (which I understand come with their own host of problems).

Goddamn was I wrong! Their CPU coolers are the most well-designed, thoughtfully planned, amazingly performing consumer product I've seen in a while. 10/10, highly recommend! I'll use them for all PC cooling needs going forward.

Parse, don't validate was written around Haskell!

That's like saying "there's only so many ways to greet your neighbor, so any text that simply greets your neighbor isn't copyrightable – and therefore no text is copyrightable".

I'd like to hear more nuance with regards to this line of reasoning. Can you conceive of a model that contains highly non-trivial representations of IP owned by others than yourself? Can you conceive that you might "order" the model to "produce" that IP? What happens then?

Try this both for "open source code" as the IP, and "the novel I wrote", and "latest Hollywood movie". The model does not have to be a real model currently available. It's just a thought experiment.

Try also to elaborate on the sliding scale between "an AI model" and "a compression system".

You are presumably human. We have granted humans specific exemptions in copyright law. We have not granted that to LLMs. Why are we so eager to?

Are you telling me that I can use the thing, but I can't use it if I process it through an LLM? It get slippery, fast.

If I write a story, I can put it online. That doesn't mean it's ok to take that story and publish it in an anthology.

There's also a TON of irony here. What an about face it is, for the community at large* to switch from "information wants to be free, we support copyleft and FOSS" to leaning so heavily on an incredibly conservative reading of IP law.

If we take the point of view that LLMs are tools (I agree), then people need to be absolutely certain that these tools don't contain (compressed) representations of copyrighted works.

People seem not to want to do that. And they argue that the LLMs have "learned" or "been inspired" by the copyrighted works, which is OK for humans.

This is the problem. People can't even agree on which of two mutually exclusive defenses to appeal to! Are LLMs tools which we have to ensure aren't used to reproduce copyrighted work without permission, or are they entities that can be granted exemptions like humans can? It can't be both!

> There's also a TON of irony here. What an about face it is, for the community at large* to switch from "information wants to be free, we support copyleft and FOSS" to leaning so heavily on an incredibly conservative reading of IP law.

True. While IP-owning companies like Microsoft now say "it's online, so we can use it".

It's bizarre.

I'll tell you what: I'll drop my conservative stance in defense if FOSS when Windows and the latest Hollywood movie are "fair use" for consumption by whatever LLM I cook up.

I've pointed out elsewhere in this thread that this is the opposite of how the real world works.

In actual fact, people who need software built hire a tool (e.g., a software developer like me) to build it for them. That tool - me or you - has inside it a tremendous library of copyrighted works represented. I've worked on enough different projects over the decades that the next CRUD function, or rule-driven data-entry tool, or whatever, that I build is going to draw very significantly from the last ones I built. And those last ones were copyrighted, with those rights held by my employer at the time, and maybe even protected by NDA or defense-style classifications.

Is your position that this is OK so long as it's stuff that I can keep in my squishy brain, but the moment that mechanism moves to silicon, it somehow becomes fundamentally different?

The other major argument I see in this thread is that for LLMs it's different because there's a third party who is aggregating the data, and selling me (or my employer) use of that tool. But this doesn't change the overall picture at all. It just adds one more layer of dereferencing into it. The addition of that middleman hasn't altered the moral landscape: how is hiring me, along with what's in my memory, different from hiring the combination of me plus a helper to supplement my memory? There's an aspect of scale, I suppose. With that helper I can achieve greater quantities, but it's not changing the story in a qualitative way.

Humans are distinct from tools, both ethically (to most people) and legally. You may not see it this way, but it is the majority opinion and the stance of the law in most jurisdictions. The rest of your paragraph falls apart without considering humans as tools.

(Incidentally: you can own tools. I don't think you want to open that door…)

> Is your position that this is OK so long as it's stuff that I can keep in my squishy brain, but the moment that mechanism moves to silicon, it somehow becomes fundamentally different?

Yes. We, humans, structured our laws because we consider ourselves and our squishy brains special.

This is, for example, why you don't get charged with murder for terminating a computer program. We, the humans, have decided that the right not to be terminated only applies to humans (and other animals, but then because we grant them that protection).

It doesn't need to. Laws are for humans.

Laws don't give rights to chainsaws. Or lawnmowers. Or kitchen knives, hammers, screwdrivers, and spades.

You can't use any of those to commit a crime and then claim that the law specifically did not exclude those tools.

Why are you seemingly in favour of carving out an exemption for LLMs?

Laws are for humans.

Arguing that the law did not specifically address "intentionally killing a person by tickling them till they died" means that you found a loophole which can be used to kill people is...

well, it's in the "not even wrong" category...

We gave certain temporary monopoly on certain uses to humans under rules little understood by laymen even if their livelihood depends on it.

And I'm worried that once that has been sufficiently normalized, laws and interpretations of them will adapt to whatever best suits those users. Which will mean copyrightwashing of FOSS. My only hope then is that surely if free software can be copyright-washed by the big guys, then so can the little guy copyright-wash the big guys' blockbuster movies or whatever, which might lead to some sort of reckoning.