81–90 of 604 posts
Oh wait, I don’t have to imagine. That’s what Anthropic does. A nice preview for what is in store for those who chose to turn off their brains and turn on their AI agents.
Yesterday was a realization point for me. I gave a simple extraction task to Claude code with a local LLM and it "whirred" and "purred" for 10 minutes. Then I submitted the same data and prompt directly to model via llama_cpp chat UI and the model single-shotted it in under a minute. So obviously something wrong with coding agent or the way it is talking to LLM. Now I'm looking for an extremely simple open-source cod…
I use AI, but only what is free-of-charge, and if that doesn't cut it, I just do it like in the good old times, by using my own brain.
Earlier quoted context omitted.
It feels more and more like OpenAI/Anthoropic aren't the future but Qwen, Kimi, or Deepseek are. You can run them locally, but that isn't really the point, it is about democratization of service providers. You can run any of them on a dozen providers with different trade-offs/offerings OR locally. They won't ever be SOTA due to money, but "last year's SOTA" when it costs 1/4 or less, may be good enough . More quantit…
Open Source isn't even within 50% of what the SOTA models are. Benchmarks are toys, real world use is vastly different, and that's where they seriously lag. Why should anyone waste time on poorer results? I'd rather pay my $200/mo because my time matters. I'm not a poor college student anymore, and I need more return on my time. I'm not shitting on open weights here - I want open source to win. I just don't see how t…
When was the last time you used any of them? Because, a lot of people are actively using them for 9-5 work today, I count myself in that group. That opinion feels outdated, like it was formed a year ago+ and held onto. Or based on highly quantized versions and or small non-Thinking models.
Do you really think Qwen3.6 for a specific example is "50%" as good as Opus4.7? Opus4.7 is clearly and objectively better, no debate on that, but the gap isn't anywhere near that wide. I'd call "20%" hyperbole, the true difference is difficult to exactly measure but sub-10% for their top-tier Thinking models is likely.
This is what worries me. People become dependent on these GenAI products that are proprietary, not transparant, and need a subscription. People build on it like it is a solid foundation. But all of a sudden the owner just pulls the foundation from under your building.
Luckily local AI is becoming more feasible every day.
Yes, it's possible to run tiny quantized models, but you're working with extremely small context windows and tons of hallucinations. It's fun to play with them, but they're not at all practical.
I feel like I'm using Claude Opus pretty effectively and I'm honestly not running up against limits in my mid-tier subscriptions. My workflow is more "copilot" than "autopilot", in that I craft prompts for contained tasks and review nearly everything, so it's pretty light compared to people doing vibe coding. The market-leading technology is pretty close to "good enough" for how I'm using it. I look forward to the da…
(but I guess they're not really conflicting, if the "solution" involves upgrading to a higher plan)
My max20 sub is sitting unused since april mostly now, codex with 5.4 (and now 5.5) even with fast mode (= double token costs) is night and day. Opus is doing convincing failures and either forgets half the important details or decides to do "pragmatic" (read: technical debt bandaids or worse) silently and claims success even with everything crashing and burning after the changes. and point out the errors it will mak…
GLM always feels like it's doing things smarter, until you actually review the code. So you still need the build/prune cycle. That's my experience anyway.
I feel like I'm using Claude Opus pretty effectively and I'm honestly not running up against limits in my mid-tier subscriptions. My workflow is more "copilot" than "autopilot", in that I craft prompts for contained tasks and review nearly everything, so it's pretty light compared to people doing vibe coding. The market-leading technology is pretty close to "good enough" for how I'm using it. I look forward to the da…
I find it incredibly difficult to saturate my usage. I'm ending the average week at 30-ish percentage, despite this thing doing an enormous amount of work for (with?) me.
Now I will say that with pro I was constantly hitting the limit -- like comically so, and single requests would push me over 100% for the session and into paying for extra usage -- and max 5x feels like far more than 5x the usage, but who knows. Anthropic is extremely squirrely about things like surge rates, and so on.
I'm super skeptical of the influx of "DAE think Opus sucks now. Let's all move to Codex!" nonsense that has flooded HN. A part of it is the ex-girlfriend thing where people are angry about something and try to force-multiply their disagreement, but some of it legitimately smells like astroturfing. Like OpenAI got done pay $100M for some unknown podcaster and start hiring people to write this stuff online.