Complaints database
29,255 raw documents, unedited and searchable. This is what everything else on the site is built from — nothing here is generated.
Train sim created by just one person is being called the best ever made
Nothing against Godot, it's a great project, but looking at the wikipedia list of notable games using the engine, there's a single one I have heard of. Meanwhile about 28% of the market share of all video games is using Unreal. So I don't really see the need to make "everyone much happier" honestly. > Sell additional…
Show HN: R3 – A Local Code Review Tool for You and Your AI Agent
Oh man I was looking for something like this for a while, reviewing code in VSCode is starting to get annoying and I began to think that VSCode is no longer suitable for development in the AI era, but I already built my own solution that addresses the two problems I had with reviewing code: - A better alternative to r…
Why American ambulance rides are so expensive
I'm not sure we're talking about the same thing. What I meant with "you have to wait months" is that if you go with the government-provided doctors and processes, they put you on a list and you have to wait for your turn. Getting an exam done can take months, and you can wait years for surgery, unless your condition i…
EU Parliament greenlights Chat Control 1.0
It's confusing because of the indirection. The Internet Watch Foundation is one of these NGOs that makes lists of hashed child porn images and URLs. All the big tech companies subscribe so they can coordinate on blocking child porn, as they are legally required to do. Unfortunately the IWF is also a "charity" and thus…
GPT-5.6
I used to have the same experience until 5.6 sol xhigh. I have instructions in my code review skill and agents.md to encourage parallelism including multiple agents as long as quality isn’t impacted. I additionally instruct codex to not use less capable agents because at least with 5.5 this would seriously increase sl…
Life with Hazard Ratios
The most important part of this for a living human being is touched on at the end. You only die once. Life expectancy is an ensemble mean over a population, and "you are not a population". You need to try to avoid risks that are going to kill you personally, not risks that affect aggregate life expectancy (there's ove…
GPT-5.6
They added this a few weeks ago. Before, they would do their celebrated usage resets while I was behind the already generous weekly quota. If I had 80% left on Friday to spend over the weekend, they would reset and I'd have 100% until next Friday, a potential net negative. OpenAI listened to the complaints.
EU Parliament greenlights Chat Control 1.0
In theory, it can't. But in reality, it does in important cases. Germany has implemented EU sanctions against a German journalist, which deprive him and his family of the ability to conduct basically any economic activity in Germany or even to leave the country. He is not allowed to work. No one is allowed to pay him…
Moving to FreeBSD from Linux
Apple made me realize that I should choose AGPL as a license for my own projects, because I don't agree with the implied contract values of X11/MIT or BSD licenses anymore. The latter licenses I choose now only for things I absolutely don't care about and won't spend any time on. Things that I care about I licens…
Grok 4.5
Steering a single model towards that seems, again, technically challenging, especially if it needs to be obfuscated enough to pass code review (you are doing that for LLM code right?) and especially if targeted towards specific fields of use (e.g. critical infrastructure). I still don't see how the threat model could…
Why American ambulance rides are so expensive
He refused to go in an ambulance the first time. He could have refused the second ride. You can refuse transport or medical care at any time as long as you are able to make rational decisions. As an EMT if you're A&Ox4, alert to time, place, person and event, you make the choices. You do sign a release so we have proo…
Muse Spark 1.1
Sometimes. It depends on the task. $15/Mtok is still a lot cheaper than human written code. It’s probably worth spending more for contract reviews. Tasks don’t have uniform value. If you’re using an AI for entertainment, then frontier premiums are hard to justify. For paid work, they’re a great deal if used reaso…
Show HN: Getting GLM 5.2 running on my slow computer
Most any ticketing system can integrate with ordinary IMAP and smtp email flow, so you can really use any agent that can "do" inbound and outbound email to talk to a self hosted ticket queue.
GPT-5.6
I did the side by side between Claude code (effort medium) and Cursor (auto). Asked Claude to prepare a plan and asked Cursor to review it and it found tons of gaps in the plan. Cost-wise, it came out better too. I have been using Cursor daily (along with Claude) and the former has been 20-30% cheaper despite me spend…
GPT-5.6
Not parent poster but: I probably spend at least 2/3 of my tokens on code review & QA. At least at my workplace, that's the culture.
Volkswagen to Slash Half Its Models as Sales Fall
> In recent days, German press reports had suggested that the company was preparing to lay off 100,000 workers by the end of the decade and close four factories in Europe. A big round of applause to the EU bureaucrats who forced upon european carmakers an agenda impossible to deal with properly (only giving about two…
Rewriting Bun in Rust
Sorry, but I don't think you understand how many different skills are involved in building software that large companies can actually rely on. Writing code is only one part of the job. I use Claude Code every day and I love it. It has made me much more productive. But I still have to guide it carefully, review every c…
Surgeons face higher infertility rates. UC may make it even harder to have kids
SO and friends are in the industry. I will never understand why they simply don't train and accept more Doctors. If they can have 1 person working 80 hours, they can have 2 working 40, or better yet 3 and see more patients and reduce wait times. Yes there is a cost to training Doctors but you can't tell me its worth t…
My thoughts on the Bun Rust rewrite
Isn't this what makes it so strange, though? You didn't see K&R publish "so-and-so writes bad C", or Stroustrup decrying the Boost maintainers as hacks. Linus used to do this sort of thing, but mostly to things that directly affected the kernel, and even that eventually led to changed behavior and a code of conduct. T…
Train sim created by just one person is being called the best ever made
For the same reason that the Vegas attraction “dig a bunch of holes with industrial diggers” was so popular: people want to do jobs they think are cool without years of training cost up front, and this is a way for them to do so. Farming simulator and Car mechanic simulator are both in my todo list, because those are…
GPT-5.6
I run my AI agent as a different user (in addition to using the sandbox functionality provided by cc/codex). It does not seem possible to run the Codex GUI as a different user. I can run the TUI (/Applications/Codex.app/Contents/Resources/codex) but it has the shortcoming that remote cont…
GPT-5.6
Anthropic just refuses to allow Fable to properly code review my projects. It's so obnoxious. If OpenAI's Fable equivalent is better at this, that'll get me to cancel my Anthropic subscription and switch.
The glass backbone: Why the Army's logistics will break in the next war
It's important not to overreact: Early concepts of aircraft in warfare, between WWI and WWII, often said aircraft would make battle lines irrelevant. They assumed nothing could stop aircraft. It didn't work out that way. Now people say the same about small, uncrewed aircraft (drones). It's based only on a few years of…
How to build a GitHub code review agent
I'm currently building one based on: https://blog.cloudflare.com/ai-code-review/ The keys here are 1. Use mechanical scripts at the start to gather a context and at the end to make the comments. Don't have an agent/llm do this part (make API calls), I do have the comment creation agent write a…
Show HN: Getting GLM 5.2 running on my slow computer
I love seeing that kind of tinkering
Groundtruth – checks your AI coding agent's claims against the Git diff
I've had some success with using a more expensive model to plan a decent work breakdown structure in a document and then a low-cost model to implement the plan and keep following-up, but wonder about subsequent need for code review and code quality. Can a lower-cost model verify completion? Iff tests and test coverage…
I think I have LLM burnout
We end up having to review the copious tests. In addition to the code (we still review it, right)?
GPT-5.6
Interesting idea. I think they're getting more wordy over time, personally, so I think it's more to do with the training than the raw data.
GPT-5.6
Genuine question/not a critique-are you actually reviewing all that code or just sending it and hoping for the best? I just can't imagine someone is reading/reviewing that much code every day, but maybe I'm wrong?
GLM 5.2 is nearly as accurate as a human book keeper
If and when a large number of companies blindly turn over their accounts payable workflow to some AI agent system, it'll be very interesting to see the "social engineer the LLM" methods that fraud people use to get money sent to them. Basically the same idea as the ancient "send a fax with a bill for an unsolicited de…