adastra22
adastra22 on Hacker News
Claude Code is one of the worst harnesses out there. The models just make up for it.
They said Claude. I'm pretty sure most normies have never heard of Claude. ChatGPT, yes. It's shortened to just "chat" now. But most people have no idea what Claude is.
Frankly I am surprised they were not disbarred.
I have never once had it work for me. Usually it feels like the model got a lobotomy. If I'm working with the model in a chat interface, I feel like we only really start to groove around 500k or so, at which point it knows my quirks and style, and is interpreting my intent correctly. That never, in my experience, survives compaction.
Maybe things are better now. It's been a while since I've compacted a conversation.
I assume that is only if compact is turned on? I never use compaction.
They don't live on top of the hill, all the time. The local chief, maybe. It is a defensive place to retreat to when under attack.
The latter, clearly? It is far more valuable to have a (badly) working system you can prod and probe than work from fantasy.
The Mythical Token-Minute.
Best against other AI output, still not even close to what a human would write.
That's a much more succinct way to put it, thank you.
In my case my marriage was falling apart, among other things, due to problems that ultimately stemmed from my ADHD. Therapy helped me fix my marriage, but only medication corrected the behaviors that led to that situation & kept me from repeating the same mistakes.
Do you not see the problems that arise here? Are we losing our accumulated engineering best practice knowledge? Standardized, common, highly reviewed subsystems provide working engineering know-how and accumulated lessons learned. Not even the best AI systems can keep that many considerations in mind when re-spinning code.
If this is your default assumption, I hope I never have to rely on code you generate.
GPT-5.6 Sol Ultrafast is running on Cerebras hardware.
How have I been inconsistent?
Verify against what? What I want verification against lives in the human's head, not the repo.
Are we now in a situation where all code is disposable, throwaway things with no long-lived projects?
We're talking about a programming language. Languages, and their compilers, evolve over time. Where is this one headed?
Legibility is not the concern. The issue is whether the point the docs are making are the actual point the author is concerned with.
Yeah that's a part of it for sure. Experience varies and replication is really difficult.
The reasons for my expectation that therapy doesn't address ADHD symptoms generally (I know it didn't for me) is more fundamental though: therapy is helpful in making conscious changes. The whole point of ADHD is that you're not consciously aware of the decisions you make. Any therapy that tries to address the root cause of ADHD issues boils down to "try to remember to not do that" when the f'ing point is that I will not remember, and I can't control that. Stimulants fix this. No amount of therapy makes it better.
Where therapy is helpful is learning to live with the disability and to prevent or contain the worst outcomes that can result. Therapy helps manage symptoms, not address the root cause. Which is good because even stimulants are not perfect, and usually by the time you're seeking help there's a lot of damage already. But they're not addressing the same thing.
As I said in my message, it would determine whether I use the project or not. I need to match it to my use cases, and if I find a match I need to know: is it matching because my use case falls in the domain of what the human collaborator wants from this project, or is my use case a hallucination that will go unsupported?
You cannot tell either from the AI written docs. I'd rather OP gave a rambling post about some problem he has in gamedev that led him to think a new language was needed, why the constraints are different, and how this will improve X, Y, or Z in his own words. Then I can infer what this project is and where it is going. Right now, I have no clue.
Goose is written by a game developer. They tend to concern themselves with C++'s niche.
I would throw the entire docs out and write it from scratch in your own words. I get that we live in an era where the machine writes most/all of the code, and I can accept that. What I need, to even consider using and extending this project, is some understanding of what part of the design is yours and what part is accident by Claude. I can't get that from reading Claudese.
"Everything is only a few hundred clicks away!"
Some will, and they will outcompete.
Therapy helps everyone. Everyone should do therapy. I expect that any study comparing ADHD with therapy to ADHD without, will find much better outcomes with therapy. Because therapy works at improving general life outcomes.
But does therapy help alleviate ADHD symptoms specifically? I doubt that very much.
And if that's not bad enough, those side effects can be permanent. This is unlike stimulants, which have side effects some people really don't like, but they go away when you stop.
It's like being invited to play Russian roulette. Literally, as suicidal ideation ironically is a known side effect of the antidepressants...
I have ADHD, I have experienced the side effects, and I wasn't telling you what to do. I hope you find peace.
Those are extremely common side effects. The benefits just make it worth it for most.
You be you and do what works for you. But cognitive behavioral therapy does not work by itself for ADHD. Stimulants are the gold standard treatment.
Stimulants don’t dampen the fun. More fun, if anything. Very few downsides.
But being a Category 2 controlled substance means health care providers are always pushing me towards off-label use of antidepressants that DO have very big and worrying side effects.