Very nice insight, that’s where the value is, even with a lot of time refactoring, testing and reviewing the compressed code phase is so much gziped than it’s still worth it to use an imperfect LLM. Even with humans we have all those post phases so great structure around the code generation leads to a lot of gains.
It depends on industries and what’s being developed for sure
That’s non sense. Every software people pay or use as their professional tool should be carefully crafted. Would you buy a house/car/anything or value knowing that people who built it don’t care about their craft as long as you got it in your hands and pay for it ? Or maybe you produce something cheap and worthless
The article is useless, but the comments about people talking about their cats are priceless.
Thank you OP to have ignited this thread, makes me so happy to read about other people love/observations of their cat.
Definitely my all time favorite on HN
+1 after many failed attempts to buy useless cat toys, I’ve been really surprised that those are what we loves the most. He can play alone with it for hours and is absolutely crazy happy to play fetch with me. Maybe when I throw the elastic hair is kind of a bird like feature him.
It’s true there’s some magic effect from Claude code’s work. But still, often it’s not exactly the same infra and scaling than production grade. But for a customer I guess that’s perfect, they have a mean to make their own tools instead of relying on platforms to build those tools.
That’s the exact definition our CEO gave of our job this week. That’s how he sees and expects us to work now. I feel some anxiety because that’s too much too fast. We went from « we need to fix every single bug we encounter » to « it doesn’t matter if there’s bugs as long as we ship a feature fast »
The exact same prompt ? Everything depends on the prompt and it’s different tools. These days the quality and what’s build around the prompt matters as much as the code. We can’t feed generic query.
There isn’t just concrete in a house. There is hundreds of things that could vary from house to house (even country to country and laws) so it’s more like the building blocs are not only imports of lib but the language itself (raw materials) which makes it a fit analogy for me
No I don’t agree. Just because it’s « boilerplate », that does not mean it’s worthless or doesn’t carry novelty.
There is « boilerplate » in building many things, house, cars etc where to add real new stuff it’s « always the same base » but you have to nail that base and there is real value in it. With craft and deep knowledge and pride.
Every project is different and not everything can be made from a generic out-of-shelf product
If found it better to split in smaller tasks from a first overall analysis and make it do only that subtask and make it give me the next prompt once finished (or feed that to a system of agents). There is a real threshold from where quality would be lost.
Our CEO, an expert in marketing has discovered Claude Code and is the one having the most open PR of all developers and is pushing for us to « quickly review ». He does not understand why review are so slow because it’s « the easiest part ». We live in a new world.
I also do that and it works quite well to iterate on spec md files first. When every step is detailed and clear and all md files linked to a master plan that Claude code reads and updates at every step it helps a lot to keep it on guard rails. Claude code only works well on small increments because context switching makes it mix and invent stuff.
So working by increments makes it really easy to commit a clean session and I ask it to give me the next prompt from the specs before I clear context.
It always go sideways at some point but having a nice structure helps even myself to do clean reviews and avoid 2h sessions that I have to throw away. Really easier to adjust only what’s wrong at each step. It works surprisingly well
Usually the quote comes in a positive light. We won’t make a law/rule around it, it’s a principle so it’s meant to be short. So yeah you could argue about anything in any way you want, positive or negative. And if you want to be really precise then you make a law but it’s so precise it won’t cover edge cases.
Don’t you agree that the baseline for most humans is to be in peace, find love, patience, joy, kindness, mildness ? You can manifest any of those traits to any stranger and you’ll likely have a positive impact right ?
That’s the context of the Golden Rule quote I guess
Yeah you’re raising three good points and they all land.
On the finetuned LLM: you’re right, that criterion was flawed. A system trained to report experience proves nothing about whether experience is present, which is actually the core of the hard problem. No behavioral output alone can confirm inner experience. That applies to LLMs, and technically to other humans too.
On dogs, also a fair correction. We don’t actually require verbal report to attribute consciousness to animals, we use behavioral and physiological evidence. So "coherent verbal report" was too narrow.
Better criterion: a system whose overall architecture and behavior is consistent with experience, not just one that says the right words.
On the standard of proof: that was a rhetorical deflection and you’re right to call it out. You asked a genuine question and got it turned back on you. And you’re pointing at something real: in science, strong correlation is not accepted as proof when stricter evidence is achievable. The reason we settle for correlation here isn’t because it’s sufficient, it’s because subjective experience may make stronger proof structurally inaccessible. But it’s also worth noting that scientific consensus has a poor track record of admitting this honestly. Dominant paradigms tend to defend themselves long past the point where the cracks are visible, physicalism on consciousness is no exception. The confidence with which emergence is presented often reflects institutional momentum as much as evidence.
The analogy holds technically, but there’s a missing piece: the brain doesn’t just update weights, it does so guided by experience that matters to a situated, embodied agent with drives and stakes. Sleep consolidation isn’t random cleanup, it’s selective based on salience and emotion.
An LLM updating more efficiently is progress, but it’s still optimizing a loss function. Whether that ever approximates what the brain does during sleep depends entirely on whether you think the what (weight updates) is sufficient, or whether the why (relevance to a lived experience) is what makes it meaningful.
So yes, the resource argument will weaken over time. But the architectural gap may be deeper than just compute.
Not necessarily either but the serious version of the argument is that life consistently acts against local entropy in purposeful ways, and pure physics doesn’t obviously explain why matter would “want” to do that. Consciousness as a organizing principle is one answer. It’s speculative, but it’s not obviously wrong
That’s a genuinely strong point. You can only verify consciousness from the inside, your own. Everything else is inference. No objective measurement can definitively distinguish “other minds exist” from solipsism. That’s not a bug in the argument, it’s a fundamental epistemic limit. Which is exactly why this question may never be fully resolved empirically
The same kind of proof we accept for any scientific claim: converging, reproducible evidence that rules out competing explanations.
Concretely, that means: We already have indirect evidence: conscious states vary predictably with brain states. Damage specific regions, lose specific functions. Alter chemistry, alter experience. This is not proof, but it’s systematic dependence, which is exactly what emergence predicts. Stronger evidence would look like precise, bidirectional mappings between neural activity and reported experience: to the point where you could reliably read subjective states from brain data, or induce specific experiences through targeted stimulation. We’re already moving in that direction.
The hardest bar would be building a system from physical components, having it report coherent subjective experience, and being able to explain why that configuration produces experience while others don’t. That’s the hard problem: and no, we’re not there yet. And it’s worth being honest: we’ve been assuming physicalism will eventually solve it, but there’s no guarantee that’s true rather than hopeful. The fact that brain states correlate with conscious states doesn’t explain why there is something it is like to have those states. Correlation is not mechanism.
But here’s the key point: you’re implicitly holding emergence to a standard of certainty that no scientific theory meets. We don’t have that standard of proof for evolution, gravity, or quantum mechanics either. We have overwhelming evidence that makes
alternatives implausible.
So the question isn’t “can you prove it beyond all doubt?” It’s “does the evidence favor it over alternatives?” Right now, it does — but that’s a pragmatic verdict, not a metaphysical one. Idealist frameworks like Kastrup’s or Faggin’s remain serious contenders. The debate is more open than mainstream science often admits.
Even if LLM will one day be autonomously updated, they started from us, from our knowledge.
The human brain « is smart », it’s wired up to be in any kind of culture or knowledge. We fill up to be smarter from experience but LLM can’t do that, I can’t teach Claude something that it will use with you the next day, it needs to be retrained with knowledge stopping at some point.
Even if technology catches up and the machine becomes more autonomous, what will say this machine would ever want to integrate to our society or share anything with us ?
They have eternity, given there is electricity. Why would they want anything to do with humans if you go that way ?
If it’s really conscious, should we consider it a slave then ? Why couldn’t « it » have fundamental rights and freedom to do whatever it wants ?
Physicalists say consciousness emerges from matter.
The other camp says matter comes from consciousness.
Federico Faggin, inventor of the microprocessor, says consciousness cannot emerge from matter because matter is inert and not self-conscious, so it cannot produce consciousness.
Who’s right and who’s wrong? Time will tell. But it is also wrong to claim that consciousness emerges from matter until it is proven (aka the “hard problem of consciousness.”)
I don’t know how much of an anecdote it is, but all the non-tech people with whom I talk about IA only know chatGPT. Competition is either non existent or the same thing. Among those, no one wants to pay the service, they just stop using it when limits are reached. I can’t say which users can turn the market around but chatGPT is indeed burned in the mind of many and because they don’t care about tech and are not interested in tech they won’t search for any other service it seems. Even after many discussions they don’t remember the names of other IA I told them
There is complains that some Volvo cars damaged iPhone cameras. It’s not even clear if Apple takes those under warranty. We’ve seen car review YouTubers that got their iPhone camera sensors damaged captured (by a second camera) while reviewing