Calavar

The classical definition of a clean room implementation is something that's made by looking at the output of a prior implementation but not at the source.

I agree that having a reference compiler available is a huge caveat though. Even if we completely put training data leakage aside, they're developing against a programmatic checker for a spec that's already had millions of man hours put into it. This is an optimal scenario for agentic coding, but the vast majority of problems that people will want to tackle with agentic coding are not going to look like that.

Sure, you can swap a USD tethered stablecoin to another currency. You can also do that with USD itself. What is the value add of injecting an extra step by first converting your USD to stablecoin?

As a physician, I can give further insight. The blood pressure medication the commenter is referring to is almost certainly a beta blocker. The effect on blood sugar levels is generally modest [1]. (It is rare to advise someone with diabetes to stop taking beta blockers, as opposed to say emphysema, where it is common)

They can be used for isolated, treatment of high blood pressure, but they are also used for dual treatment of blood pressure and various heart issues (heart failure, stable angina, arrhythmias). If you have heart failure, beta blockers can reduce your relative annual mortality risk by about 25%.

I would not trust an LLM to weigh the pros and cons appropriately knowing their syncophantic tendencies. I suspect they are going to be biased toward agreeing with whatever concerns the user initially expresses to them.

[1]

I thought most of the major AI programming tools were already doing this. Isn't this what subagents are in Claude code?

Talk about market cap, especially meme stock maket cap, reminds me of that old XKCD comic on extrapolation. Market cap is what you get when you extrapolate the fair market value for the 1% of a company's shares currently on the market all the way out to 100%. But demand doesn't work that way - it doesn't scale linearly.

Reza Pahlavi has also had recent interviews with CBS [1], the Economist [2], and CNN [3] (all within the last 30 days). So how is the existence of Reza Pahlavi interviews on Fox evidence that conservative media is covering this issue more than liberal media?

On the topic of Politicians, Democratic congressmen like Dave Min and Jim Hines have also spoken in favor of US intervention in Iran.

[1] https://www.cbsnews.com/video/full-interview-exiled-iranian-...

[2] https://youtu.be/jk6wfvje8Zo

[3] https://www.cnn.com/2026/01/23/world/video/iran-united-state...

This is a straw man in my opinion. But regardless of that, your theory doesn't explain why conservative media isn't really covering this either - The Iran protests haven't exactly been front page material on Fox News or OAN or Breitbart

Isn't a lack of comments the opposite of the problem you were previously claiming?

> All comments, complaints, corrections, and retraction requests? Unmoderated? Einstein articles will be full of comments explaining why he is wrong, from racist to people that can spell Minkowski to save their lives. In /newest there is like one post per week from someone that discover a new physics theory with the help of ChatGPT. Sometimes it's the same guy, sometimes it's a new one.

Judging from PubPeer, which allows people to post all of the above anonymously and with minimal moderation, this is not an issue in practice.

I don't think any oil execs are interested in this, just like they weren't interested in investing in Venezuela after Maduro's ouster (at least if you believe the Financial Times).

Rather these invasions appear to be the pet projects of neo-imperialist advisors in the government who see national growth as a zero sum game, a Starcraft-esque race for a finite set of resources where powerful countries can generate wealth only by using their power to steal from others. In Steven Miller's own words: "[The world] is governed by force, [is] governed by power. These are the iron laws of the world since the beginning of time."

> Lax is completely open source and uses the MIT license.

I see that there is an MIT license file in the repo, but the readme in the repo still says there is no license and redistribution in source or binary form is explicitly forbidden. This is a discrepancy that should probably be addressed

Danluu has a great piece on why Ballmer was better than people gave him credit for: https://danluu.com/ballmer/

The linked reddit post appears to be about Australia

There are multiple Python 3 interpreters written in JavaScript that were very likely included in the training data. For example [1] [2] [3]

I once gave Claude (Opus 3.5) a problem that I thought was for sure too difficult for an LLM, and much to my surprise it spat out a very convincing solution. The surprising part was I was already familiar with the solution - because it was almost a direct copy/paste (uncredited) from a blog post that I read only a few hours earlier. If I hadn't read that blog post, I would have been none the wiser that copy/pasting Claude's output would be potential IP theft. I would have to imagine that LLMs solve a lot of in-training-set problems this way and people never realize they are dealing with a copyright/licensing minefield.

A more interesting and convincing task would be to write a Python 3 interpeter in JavaScript that uses register based bytecode instead of stack based, supports optimizing the bytecode by inlining procedures and constant folding, and never allocates memory (all work is done in a single user provided preallocated buffer). This would require integrating multiple disparate coding concepts and not regurgitating prior art from the training data

[1] https://github.com/skulpt/skulpt

[2] https://github.com/brython-dev/brython

[3] https://github.com/yzyzsun/PyJS

> If a valid alternative is to halt normal operations and present an alert box to the user saying "internal error 573 occurred. please restart the app", then that is much preferred IMO.

You can do this in your panic or terminate handler. It's functionally the same error handling strategy, just with a different veneer painted over the top.

Paul Graham said the same thing about Python 20 years ago [1], and back then it was true. But once a programming langauge hits mainstream, this ceases to be a good filter.

[1] https://paulgraham.com/pypar.html

You're not the only one who feels that way, but IMHO it's not a valid complaint.

The C++ standard says implementation defined because the weeds get very thick very quickly:

- Are paths formed with forward slash or backslash?

- Case sensitive?

- NT style drive letter or Posix style mounts?

- For relative paths, what is it relative to? When there are multiple matches, what is the algorithm to determine priority?

- What about symlinks and hard links?

- Are http and ftp URIs supported (e.g. an online IDE like godbolt). If so, which versions of those protocols? TLS 1.3+ only? Are you going to accept SHA-1?

- Should the file read be transactional?

People already complain that the C++ standard is overly complicated. So instead of adding even more complexity by redefining the OS semantics of your build platform in a language spec, they use "implementation defined" as a shorthand for "your compiler will call fopen" plus some implementation wiggle room like command line options for specifying search paths and the strategy for long paths on Windows

What if #embed steals my credit card data is a pointless strawman. If a malicious compiler dev wanted to steal your credit card data, they'd just inject the malicious code; not act like a genie, searching the C++ spec with a fine comb for a place where they could execute malicious code while still *technically* being standards conformant. You know that, I know that, we all know that. So why are we wasting words discussing it?

"inline" was always just a hint

Quite to the contrary, I'd say this update is evidence of the inner loop being hyperoptimized!

MSVC's support for musttail is hot off the press:

> The [[msvc::musttail]] attribute, introduced in MSVC Build Tools version 14.50, is an experimental x64-only Microsoft-specific attribute that enforces tail-call optimization. [1]

MSVC Build Tools version 14.50 was released last month, and it only took a few weeks for the CPython crew to turn that around into a performance improvement.

[1] https://learn.microsoft.com/en-us/cpp/cpp/attributes?view=ms...

Because Ruby is just a better thought out language than Python is. It had a sane package management story 15 years before Python. It doesn't rely on weird hacks like exception throwing for iterator control flow. It doesn't have nearly as many warts and footguns on basic operations (for example, in Python csv.writer(file).writerows(rows) is broken unless you remembered to set the newline mode on the file; in Ruby file.write(rows.to_csv) just works). Thanks to a sane VM design it's mainline JIT can actually run code faster than interpreted code (something that CPython's JIT can't do [1])

Many Pythonistas are woefully ignorant of what's going on outside their siloed community.

[1] https://fidget-spinner.github.io/posts/jit-reflections.html

Let me counter with this: Can you point out one country in the post world war era that had minimal government investment in science but had very productive scientific output? Or can you point out one country where scientific productivity increased after public sector investment in science was slashed?

Reminds me of a recent story on scammers using search queries to inject their scam phone numbers into the h1 header on legitimate sites [1]

[1] https://cyberinsider.com/threat-actors-inject-fake-support-n...

I think you hit the nail on the head with the analogy to Doctor GPT, but I think you missed it with gatekeeping. I don't think it's about gatekeeping at all.

A freelance developer (or a doctor) is familiar with working within a particular framework and process flow. For any new feature, you start by generating user stories, work out a high level architecure, think about about how to integrate that into your existing codebase, and then write the code. It's mostly a unidirectional flow.

When the client starts giving you code, it turns into a bidirectional flow. You can't just copy/paste the code and call it done. You have to go in the reverse direction: read the code to parse out what the high level architecture is, which user stories it implements and which it does not. After that you have to go back in the forward direction to actually adapt and integrate the code. The client thinks they've made the developer's job easier, but they've actually doubled the cognitive load. This is stressful and frustrating for the developer.

Doctors do, as required by federal law. You can look up any doctor you like:

https://openpaymentsdata.cms.gov

Web browsers, yes. With GUIs and games, it's a less clear. Of course you can write GUIs and games in any Turing complete language but there's still a lot of work to be done in finding the right ergonomics in Rust [1, 2].

[1] https://www.warp.dev/blog/why-is-building-a-ui-in-rust-so-ha...

[2] https://loglog.games/blog/leaving-rust-gamedev/

Because fiddling with Windows firewall settings is a power user feature that only a fraction of a percent of users will touch. If it ever becomes more widely used, then I agree, all bets are off.

> there are much, much larger and controversial line items in the budget to argue about than feeding the poor.

For example, the 40 billion dollar bailout of the Argentine economy.

> Something is very wrong if it takes 20+ years to field new military technologies.

Is it? By what criteria? IMHO the point is to get new tech out quickly enough that you aren't falling behind other major powers in the international arms race. The F35 seems to be ahead of the competition because countries around the world are lining up to buy it over much cheaper alternatives from Russia (Su57) and China (J35).

Not to mention that the Su57 also had about a 20 year development cycle. Maybe that's just how long takes to develop a new stealth fighter?

> Learning what though? When I wrote software I learn the domain, the problem space, the architecture, the requirements, etc

You don't learn these things by writing code? This is genuinely interesting to me because it seems that different groups of people have dramatically different ways of approaching software development

For me, the act of writing code reveals places where the requirements were underspecifed or the architecture runs into a serious snag. I can understand a problem space at a high level based on problem statements and UML diagrams, but I can only truly grok it by writing code.

What happens when a researcher makes a generative art model and publicly releases the weights? Anyone can download the weights and use it to turn a quick profit.

Should the original research use be considered legitimate fair use? Does the legitimacy get 'poisoned' along the way when a third party uses the same model for profit?

Is there any difference between a mom-and-pop restaurant who uses the model to make a design for their menu versus a multi-billion dollar corp that's planning on laying off all their in house graphic designers? If so, where in between those two extremes should the line be drawn?