I've experimented quite a bit with giving agents python vs sed + awk. They make mistakes with both, a lot. The only thing that has stood out is that agents reach for python too quickly if it's available, and that awk causes the least problems, while sed might take several attempts to get results, similar to python.
A lot of what a custom harness does isn't necessarily interacting with a model, it's the framework around it. Part of harness engineering is figuring out what can be done deterministically to avoid calling an LLM. A generational jump doesn't obsolete those areas, it just means you get the right answer (maybe) faster and easier when you do call the model.
There's a difference between their harness and your harness. The latter can be tuned to your preferences, while the former cannot. A custom harness can do everything that a packaged harness cannot. Good luck getting Claude Code to coordinate with multiple model providers or deeply integrate with your environment.
The point you're making is to be a consumer, which admittedly, is probably best for most people, but it's not going to satisfy the hacker. Normies aren't building custom harnesses, so your argument is DOA for the target demographic.
We might have different views on where the SE as a profession is headed, I think. Again, I might be wrong, nobody knows the future, but we didn’t even use the word “harness” that much a year ago. There’s a decent chance in two years the harness itself as a concept might be obsolete. Depended on how fast your teams move, investing in a custom harness might be a bad resource allocation. Unless, obviously you’re in top tier companies which can benefit from it and pivot fast.
Again, this is all subjective risk assessment. I very much support any engineer to learn how to make harnesses, and I, myself, have a long way to go. We’ll see where we are in a year or two, I guess.
We're not going to suddenly have new robots or operating systems. Those things take time. Just because there's massive change afoot, doesn't mean it's widely adopted or applied at lower-level products. ChatGPT is a product, and software, and a breakthrough. You can have a freeform conversation with your computer about anything, in human language, and ask it to do or make stuff and it will at least try, sometimes with surprising results. That wasn't possible until recently. The robots and products are coming, rest assured.
If you can express something deterministically with code, it's better to do that rather than have an agent do it, because it's faster, cheaper, and deterministic. E.g. you regularly copy file A to file B. You can ask the agent to do it, or you can write a script and have the agent call the script via skill. That's the beginning of a harness.
Eventually you arrive at building custom software that does a lot in the traditional way, but delegates certain tasks to the model where it makes sense or it's non-trivial/impossible to express via code.
That makes sense. Sounds similar to what I am usually doing. I let AI write a python script or similar, review it and then use the python script. I don't think I would let AI do anything important directly.
Check out IPLD (the model that powers IPFS), which translates hash-linked DAGs and handles merkle DAGs (git) natively. There's been several attempts to build Xanadu'ish things on various types of networks, web2, web3, p2p, etc, but it just doesn't seem to be what the majority wants. Maybe someday...
I feel the same, but it's an unavoidable cycle IMHO. Any subculture that becomes popular gets diluted by more average people, and the content becomes more basic and banal. You can flag and downvote, but eventually there are enough people who treat a platform like HN as a general discussion board, without regard for cultural alignment. It seems that stuff like politics and random recipes are on the rise, but it's not against the rules, so maybe it really is just a general discussion board that's evolving into the mainstream.
Just because you might prefer coffee or soylent doesn't mean some other hackers don't like cocktails. Make yourself some https://entropia.de/Tschunk and relax!
You've missed the point. I enjoy cocktails, and I even think the linked page here is nice, both in cocktail selection and page design, but I don't think it particularly belongs on HN.
I'm not trying to win any wars here, I only provided a possible explanation to someone's question. Culture evolves and that's fine, or maybe the ratio of technical vs non-technical content is actually stable on HN? I really don't know, but to me it feels like the the volume of random content and low-grade comments is increasing. Personally, I don't think that represents the spirit of HN, but my comment was downvoted so either I'm wrong, the culture has shifted, or the mainstream is already the majority here.
I'm a fan of the Caipirinha however, so thanks for the Tschunk recipe.
I see what you mean, and agree that that kind of submission taking a place on front page would indeed be better, but this submission is still original, done-by-hand work, while pancakes guy served out LLM slop...
Well, it’s our fault. We had forums with categories and those had some random / chit chat category to have off topic discussions with the community while staying on topic anywhere else.
And a lot of the time, this random section was the more active of the whole forum but that was ok because it didn’t impact the other discussions.
But those were less engaging than social networks i guess (which hn is) so here we are.
On-Topic: Anything that good hackers would find interesting. That includes more than hacking and startups. If you had to reduce it to a sentence, the answer might be: anything that gratifies one's intellectual curiosity.
Off-Topic: Most stories about politics, or crime, or sports, or celebrities, unless they're evidence of some interesting new phenomenon. If they'd cover it on TV news, it's probably off-topic.
I feel like the cocktail recipes site posted falls closer to on-topic than off-topic then, having only Ingredients and Recipes pages, not even a blog...
In Northern California the number corresponded to POP-CORN, so if you wanted to know the time you called popcorn, and got a lovely lady's voice: "At the tone, the time will be...". The POP prefix was a shortcut to the phone company.
reply