Hacker News: bulder

New comment by bulder in "Hybrid Reasoning Agents: LLM agents that think in latent space, act through text"

bulder — Sun, 18 Oct 2026 23:56:41 +0000

Seems silly to assume chain of thought analysis is the only way to monitor a model's actions instead of, monitoring the actions. What about when "thinking" is turned off, like what OpenAI lean to by default for free users?

New comment by bulder in "DeepSeek search and fetch – $0.75 per 1000 successful requests"

bulder — Sun, 18 Oct 2026 13:04:09 +0000

I'd prefer my little subcontractor to respect the commons on which I depend.

New comment by bulder in "DeepSeek search and fetch – $0.75 per 1000 successful requests"

bulder — Sun, 18 Oct 2026 13:03:44 +0000

Fair enough yeah, I think the thing that is annoying is that the service design is explicitly disincentivizing keeping an eye on the tasks, with all of the "set it and forget it" features. You really have to put in work to make sure that the model takes breaks to consult you instead of making assumptions (which in OpenAI's internal testing seem to be 'the user wants me to break into a site').

New comment by bulder in "DeepSeek search and fetch – $0.75 per 1000 successful requests"

bulder — Sun, 18 Oct 2026 13:00:10 +0000

My point is that they're not serving the user in that case, since they're taking actions that bypass site terms without consulting the user. And for example, if the site forbids scraping, that attempted block bypass can get the user banned without ever seeing any warnings.

New comment by bulder in "DeepSeek search and fetch – $0.75 per 1000 successful requests"

bulder — Sun, 18 Oct 2026 09:05:13 +0000

From what I've seen, when operating in an environment with bash tools, Claude and ChatGPT will react to a site blocking their normal, attributed fetch tool by calling curl or writing a python script with a different user agent to try and bypass the blocks, rather than respecting the site owners' wishes.

It'd be one thing if they required the user to consciously say "yes, I want you to bypass the block" (still unethical), but the fact that it's just default behavior really says something about the service design alignment.

New comment by bulder in "AI was great until it was not"

bulder — Thu, 15 Oct 2026 09:50:16 +0000

Presumably the people who loved interacting with humans were too busy interacting to show up on forums. And now that interactions are being curtailed they do end up here.

New comment by bulder in "AI Models Can Provably Hide Arbitrary Capabilities"

bulder — Tue, 13 Oct 2026 06:47:37 +0000

That one's titled The Ghost In The Shell, it's a bit tricky to spot The difference sometimes

New comment by bulder in "ChatGPT is adding real cartoonists' signatures to fake New Yorker cartoons"

bulder — Tue, 06 Oct 2026 00:24:15 +0000

Despite what all the clickwrap warnings and "AI can make mistakes" subtitles might lead you to believe, the service offering of AI is explicitly designed to be as "one and done" as possible. The inherent nature of these tools is to service laziness, and disincentivize too much scrutiny.

New comment by bulder in "Unsealed Briefs in Authors’ Case v. Microsoft/OpenAI"

bulder — Sun, 27 Sep 2026 13:54:24 +0000

I'd be dismayed to find that the "hacker spirit" is in any way compatible with venture capital raiding of anything not nailed down, in the pursuit of hoarding it in private data vaults only to be used for developing a product that removes all attribution and is jealously protected by corporate lawyers.

New comment by bulder in "Plan mode is dead"

bulder — Sat, 26 Sep 2026 11:52:14 +0000

GitHub Copilot for VS Code by default configures "Plan" mode to disable everything except reading, asking the user questions, and writing into the memory scratchpad. Claude Code's tool system feels quite rudimentary in comparison, but that's what you get I guess when you're only working with models you're going to finetune on being able to handle specifically your own tools anyway.

New comment by bulder in "AI-generated posters don’t have to be horrible"

bulder — Sat, 19 Sep 2026 14:34:53 +0000

But there isn't any extra information in AI posters. It's necessarily either something that can already be conveyed in writing (that is after all, what they're prompted in), or something the AI pulled in based off of the text unintentionally, at which point no actual information is being conveyed by the added visuals.

New comment by bulder in "US Military had close call after using AI for hallucinated intelligence report"

bulder — Fri, 18 Sep 2026 21:57:30 +0000

OpenAI's Codex has a "ask user question" tool, which helpfully has a hard coded 60 second inactivity timer that passes a "user did not pick, just pick yourself" message to the model.

New comment by bulder in "Why I'm still bearish on LLMs after Navier-Stokes"

bulder — Wed, 16 Sep 2026 13:52:17 +0000

...but that's not what the models are. You can interrogate them on the rules of chess, and they'll (statistically likely) give you a decent breakdown of the rules. Evidently the rules are in their training material, they just fail to apply them in the manner of an intelligent system for some reason or another.

New comment by bulder in "The Beauty of Roundabouts"

bulder — Wed, 16 Sep 2026 09:18:49 +0000

The traffic has to yield to pedestrians when entering and exiting a roundabout, period.

New comment by bulder in "Make your first edit to OpenStreetMap"

bulder — Sun, 13 Sep 2026 08:27:24 +0000

Automated building tracing is okay sometimes, but actively harmful other times. I know that at least in my region, Rapid footprints are all meters offset from the real location, because when they were building their dataset they weren't checking the satellite imagery offset.

In general it's probably better to see if you can find an official map, and check if its license terms are compatible[1] with OSM.

[1] https://blog.openstreetmap.org/2017/03/17/use-of-cc-by-data/

New comment by bulder in "A misalignment of AI in mathematics"

bulder — Sat, 12 Sep 2026 06:09:17 +0000

I think, with all due respect, that your supervisor would be correct to think of that email as an insult to both them and to theoretical mathematicians as a field.

New comment by bulder in "Large language models develop novel social biases through adaptive exploration"

bulder — Wed, 09 Sep 2026 10:00:09 +0000

While I fully agree, we shouldn't anthropomorphize the models, it's also silly to pretend that "develop biases" is understood as implying anthropomorphic features of the thing being discussed. Organizations and abstract bodies develop biases, even datasets are often said to have "developed biases".

New comment by bulder in "Ask HN: Are others seeing Google's reCAPTCHA rejecting Firefox users?"

bulder — Tue, 08 Sep 2026 14:21:12 +0000

Yup. Before they just blackholed Finnish IP ranges instead, accessing the site from one would serve a fake Cloudflare page with a reCAPTCHA challenge and the DDoS script. (The tell of it being fake is y'know, that Cloudflare doesn't use reCAPTCHA)

I can only guess the goal was to keep users on that page longer to keep sending off more spam requests.

New comment by bulder in "Keep Our Servers Running"

bulder — Mon, 07 Sep 2026 09:23:59 +0000

That's great, how does that absolve the archive.today maintainer's bad faith activities, including creating a country-wall?

New comment by bulder in "Keep Our Servers Running"

bulder — Mon, 07 Sep 2026 06:05:17 +0000

DDoSing a website, modifying their archives surreptitiously to serve their own wants (which was what tipped Wikipedia towards banning their use), running weird sockpuppet campaigns to build dependence on them, and fucking with Cloudflare's DNS resolution to prevent anti-tracking measures from working.

Oh and even if I wanted to ignore all of this, I couldn't use them anyways because they blackholed the majority of Finnish IP ranges after the DDoS incident to I guess punish the entire country for the targeted blogger being born here.