Hacker News: Sivart13

New comment by Sivart13 in "Railway Security Disclosure"

Sivart13 — Wed, 21 Oct 2026 22:41:27 +0000

surely once there's been enough patches the code will be perfect though?

New comment by Sivart13 in "Claude Duet"

Sivart13 — Wed, 21 Oct 2026 21:38:08 +0000

tons of people these days reflexively respond to long comments saying they're AI

one could wonder, how many of these reflexive AI replyers are themselves AI

New comment by Sivart13 in "Do dentists know what they're doing?"

Sivart13 — Thu, 15 Oct 2026 02:55:44 +0000

This was the most male gazey article I've read linked from HN. Some real Don Draper vibes.

New comment by Sivart13 in "Towards Self-Driving Codebases"

Sivart13 — Thu, 17 Sep 2026 19:47:29 +0000

Well put. I've found that agents are great at adding features to areas of the code that I already understand, where I can phrase my goals coherently and keep them on the rails.

Any attempt to describe software by "specification" leaves enough ambiguity that the agent will do deranged stuff like add a ton of code to satisfy error cases that can't happen. Or, like you imply, it will confidently create a bunch of nonsense to "solve" a problem in a way that is not really possible.

New comment by Sivart13 in "Pion, an agent designed to run any company autonomously"

Sivart13 — Mon, 14 Sep 2026 17:27:21 +0000

I wish we could go just one day without a company proudly engaging in Torment Nexus related activities.

New comment by Sivart13 in "Feeling Sad about AI"

Sivart13 — Fri, 11 Sep 2026 17:44:56 +0000

> a lot of people (myself included, a few years ago now) have been through this moment of existential crisis and come out the other side

It's not that binary for me. Sometimes I'm on board, sometimes I'm chewed up by existential dread.

It doesn't help that there's no time to really acclimate. The entire world economy is pointed directly toward making this thing as powerful (dangerous) as possible, as fast as possible.

New comment by Sivart13 in "ChatGPT Images 2.5"

Sivart13 — Tue, 08 Sep 2026 19:06:53 +0000

Great idea, get an AI art tattoo so you can always be reminded how AI art looked in 2026.

New comment by Sivart13 in "AI Agents and the Refactoring That Never Happens"

Sivart13 — Wed, 02 Sep 2026 22:51:58 +0000

Test coverage for its own sake is not always a worthwhile goal.

New comment by Sivart13 in "AI Agents and the Refactoring That Never Happens"

Sivart13 — Wed, 02 Sep 2026 22:50:22 +0000

"Confidently wrong" is definitely a standard behavior model for LLMs.

I agree with the other reply that you're likely to get better results if it has some kind of test case to run that's more authoritative than its own reasoning.

New comment by Sivart13 in "AI Agents and the Refactoring That Never Happens"

Sivart13 — Wed, 02 Sep 2026 20:13:05 +0000

> Here’s the problem. AI agents are not bound by human context limits in the same way. An agent can read the tangled function, trace every caller, and make sense of the mess that would have stopped a human cold. It can add the next branch correctly, and the one after that, working confidently inside code that no human on the team fully understands anymore.

They're not bound by the same limits but they're still bound by some limits, yeah?

I'm not an AI expert, so I don't honestly understand why LLM driven agents are as good as they are. But my impression is "trace every caller", most of the time, is still an approximation. Once the code has gotten convoluted enough, cases are going to get dropped.

New comment by Sivart13 in "SteamdDB Joins Nexus Mods"

Sivart13 — Wed, 02 Sep 2026 18:55:20 +0000

> So the real news here is what "sustainable foundation" means.

I think that's answered pretty well by the 5 ads that open when I visit the announcement page on nexusmods.com link in the OP vs the speedy and clean UI that on the steamdb.info version (for now).

New comment by Sivart13 in "The Amazon tax"

Sivart13 — Tue, 18 Aug 2026 21:40:53 +0000

It's completely insane to me that the first search for Okta in the app store isn't the actual Okta app, it's a half-page ad for something that could easily be confused with the Okta app

substitute for literally any business name

Sure, I know the right one to click on, if I'm in a sound state of mind. What if I'm sleepy? Stressed? 65+ years old?

New comment by Sivart13 in "Discovery Loop"

Sivart13 — Wed, 05 Aug 2026 18:10:08 +0000

I guess if we failed to Prevent Nuclear Terror the bunker denizens of the future are gonna need somewhere to hang out.

New comment by Sivart13 in "Discovery Loop"

Sivart13 — Wed, 05 Aug 2026 18:09:30 +0000

The solution to most of these problems lies in policy, not in new tech advancements.

Maybe if our biggest companies did something other than suck up to science denying wackos, some progress could be made in these areas.

New comment by Sivart13 in "Waymo in Dallas"

Sivart13 — Tue, 04 Aug 2026 22:56:40 +0000

Why would we want to have robot cars that only drive as good as humans can?

New comment by Sivart13 in "Claude: Elevated errors across all models"

Sivart13 — Wed, 29 Jul 2026 21:42:38 +0000

no accountability so far

New comment by Sivart13 in "Claude: Elevated errors across all models"

Sivart13 — Wed, 29 Jul 2026 21:05:15 +0000

when I retire I'm opening a bar for AI Agents called The Honest Caveat

New comment by Sivart13 in "Claude: Elevated errors across all models – Resolved"

Sivart13 — Wed, 29 Jul 2026 20:23:23 +0000

If anything Claude writes too many comments. Instead of clean code it dumps walls of text describing a given 'if' as a 'user-grained access-gated control-correcting flow-valve'.

New comment by Sivart13 in "AI changes the economics of software rewrites"

Sivart13 — Thu, 09 Jul 2026 18:44:58 +0000

you don't need to be an LLM to have LinkedIn brain, but it helps

New comment by Sivart13 in "AI demands more engineering discipline. Not less"

Sivart13 — Wed, 17 Jun 2026 23:17:34 +0000

This is the major part of her argument I didn't understand.

As I see it, the "infra as code" transition means going from more ad-hoc changes to less. Predictable, auditable.

Using LLMs to replace mostly-working code with a new, nondeterministically different version seems entirely different to me. You've identified a problem in the code, great, let's rewrite it from spec. Now it fixed the problem! But did it introduce more problems? Hope you used enough tokens, I guess!