2001zhaozhao

5 ms·

61.
by 2001zhaozhao 3mo ago
I think they definitely have the IDE vision in mind and are trying to move off of terminals, based on two parts of the original article > so you can close the application, reconnect from another device, and pick up exactly where you left
62.
by 2001zhaozhao 3mo ago
Right now there is about a 95% chance that when I manually review an agent's code I find nothing that needs changing. I find it plausible that an extra agentic review pass and more testing can bring this number up to the point that one
63.
by 2001zhaozhao 3mo ago
The thing is, one could probably create a specially-prompted "agentic homework cheating" harness that catches all of these professors' trick attempts 100% of the time while also making the response look more human. The only s
64.
by 2001zhaozhao 3mo ago
I wonder how long it will take for this to get fully decensored and for bad, BAD things to happen
65.
by 2001zhaozhao 3mo ago
I write almost all code with GenAI nowadays. I'd like to say it's not slop, but I guess now there's no way to get Codeberg to agree.
66.
by 2001zhaozhao 3mo ago
It is probably from randomness. The benchmark tasks nowadays are so long that you can't really afford to run a large number of samples of them per model & effort combination
67.
by 2001zhaozhao 3mo ago
I thought the cache length was 1 hour, not 5 minutes?
68.
by 2001zhaozhao 3mo ago
Yep i aim to split my work into tickets that are <=1k lines of code each. In the end it all gets squashed into a mega 10k commit regardless but i can keep track of the quality much more that way. (more precisely the sub-tickets merge int
69.
by 2001zhaozhao 3mo ago
I had a bit of this impression when reading the post as well as the authors' product website. A lot of it does seem to be stuck in 2025. For instance I think their post "long-context isn't the answer" on their website po
70.
by 2001zhaozhao 3mo ago
> When I say maintainability, I mean the specific thing where it becomes really, really hard to change one part of the codebase without breaking another part. The corollary of agents being bad at maintainability but good at coding is tha
71.
by 2001zhaozhao 3mo ago
I think the right shape is to review and merge directly from the agent window.
72.
by 2001zhaozhao 3mo ago
They seem to be doing this to disincentive people who spam bug reports with AI where only 10% of them are actually real. Instead, now you probably need to be actually vetting your bug reports yourself so that a large percentage of them are
73.
by 2001zhaozhao 3mo ago
I feel like the simplest Pelicanmaxxing method is just to teach the model that whenever a user asks for a svg illustration of something with no other clarification, it should default to making it as detailed and pretty as possible. This wou
74.
by 2001zhaozhao 3mo ago
bro literally tried to make satire and ended up laying out an actual compelling case for AI management
75.
by 2001zhaozhao 3mo ago
AI 2027 was right.
76.
by 2001zhaozhao 3mo ago
> Cheap > Hetzner cloud > published in December 2025 which means that it's twice as expensive now as it was when the article was published
77.
by 2001zhaozhao 3mo ago
I think a tool with this kind of UX is the future of engineering orgs. It reminds me a lot of JetBrains Space (which I already thought was a good idea back then), but with agents. I don't see a lot of information about the Git hosting
78.
by 2001zhaozhao 3mo ago
I wonder when they are going to start realizing that on-premises self-hosted is the solution over local
79.
by 2001zhaozhao 3mo ago
Don't worry, it will also soon start to exploit security issues that Codex and Fable refuse to exploit... (and if it refuses, the abliterators will fix that "problem" real quick)
80.
by 2001zhaozhao 3mo ago
greg
81.
by 2001zhaozhao 3mo ago
Now there are not one, but two incredibly powerful open LLMs. I think this level of capability makes general prioritization / high level decision making doable with the right harness, and now everyone has hard-to-interrupt access to th
82.
by 2001zhaozhao 3mo ago
Oh, I already integrated this into my agent orchestrator but I can now improve the integration with the source code access. Pretty cool.
83.
by 2001zhaozhao 3mo ago
I really respect the epistemtics work here. It might become an accurate, inexpensive open-weight workhorse for high-level prioritization and decision-making work. (Finance bros will also love this)
84.
by 2001zhaozhao 3mo ago
> Every other tracker still requires a web UI. Planwright is the only PM tool with a full MCP server surface — so the human PM can operate the entire board from any MCP-compatible AI assistant. I don't know about you but I lowkey th
85.
by 2001zhaozhao 3mo ago
In Plan C the government essentially misses the opportunity to implement the multinational deal while the threat of covert projects is still low (fewer latest-gen chips unaccounted by tracking measures, worse models/algorithms to use f
86.
by 2001zhaozhao 3mo ago
This is by far the most realistic optimistic AI takeoff scenario I've seen, and more specifically it's the first one I've read that deals with both the AI alignment and power concentration issues in a sufficient way, even in
87.
by 2001zhaozhao 3mo ago
I'm working on a multi-harness IDE that supports custom agent workflows and skills that are shared between any harnesses it wraps over. I think it might prove handy for a workflow like yours.
88.
by 2001zhaozhao 3mo ago
It means OpenAI and Anthropic are now in a RSI race with each other
89.
by 2001zhaozhao 3mo ago
Huh, a good alternative just as anthropic's 50% weekly subscription subsidy is ending this weekend. Time to see if it's benchmaxxed or actually a strong leap over GPT5.5. They also seem to really not care about alignment, or care
90.
by 2001zhaozhao 3mo ago
Somehow this article doesn't even mention the fact that AI makes software rewrites much, much faster than before and with higher confidence of backwards compatibility. Nowadays, a good AI harness can fairly reliably rewrite a medium co

More ›