wavemode

Your state has eliminated metal tab cans? Huh. What do you drink 12oz of pop out of?

At the same time, I think some people in this comment section are underestimating the likelihood that this was a hack, because they're overestimating how sophisticated such a hack would have to be.

In my mind, if this was a hack, it was probably not a Stuxnet virus or something. These are smart fridges we're talking about - someone probably just logged into them using leaked credentials or a Web app vuln, and turned them off.

> modern SSD bandwidth is ~5-15GB/s

That's nothing compared to modern memory bandwidth.

No, you're comparing apples and oranges. All an SSD sequential read is doing is copying data from one place to another. So you should be comparing SSD bandwidth to memory bandwidth, not SSD bandwidth to (time it takes to execute some arbitrary algorithm). Or you should be comparing SSD bandwidth when performing millions of tiny random non-sequential reads and writes, to the algorithm time.

What your comment demonstrates is that it is possible in some cases for I/O to be fast enough to not be a performance bottleneck for certain kinds of programs. But not that I/O is not slow.

No, I don't think the way you're characterizing this is accurate. I/O is inherently very slow compared to computation. And many programs genuinely don't have any useful computation to do while waiting for I/O - because the result of that I/O operation contains the information needed for the program to even make its next decision.

Such programs are not necessarily impossible to optimize. One common optimization is to use an event loop, allowing just a few threads to handle thousands of concurrent operations. Because while a thread is waiting for I/O in one request or unit of work, in the meantime it moves on to work on processing another request/unit. Another common optimization is batching/grouping of I/O calls.

> why can’t nested functions just be normal functions inside a namespace

Because GCC's nested functions are closures - they can access local variables within the function.

I feel the opposite way. I very rarely trust the database system to do the right thing. Any query more complex than a basic lookup by primary key requires me to look at query plans and validate that indexes are in place and are being used. Otherwise we risk the production server grinding to a halt.

Personally I'd love a more explicit form of SQL that allowed specifying things like "select via scan" or "select via index lookup". (I don't think this HN submission is that - I'm just saying generally.)

Knowing game codebases, I'm guessing their test coverage was not fantastic. Agents weren't gonna get it done.

Both you and the parent commenter seem to be misunderstanding the point the Bitter Lesson paper makes.

The Bitter Lesson is about general-purpose algorithms vs. specialized algorithms. Historically, chess engines were programmed to look at a chess position and use positional understanding (imparted by the human programmers) to decide what the best move is. But eventually, the chess engines that actually became stronger than humans were instead programmed to just check every possible move and countermove and see which ones lead to a win. (I'm oversimplifying, but you get the point.) So even before Stockfish contained a neural network, it was considered an example of the success of the Bitter Lesson.

As it applies to AI agents, the Bitter Lesson would predict that the best possible agent would simply possess A) a way to do anything it wants, B) a way to evaluate whether what it did was correct, and C) a ton of compute. Then just turn it loose on your task. (The fact that the "brain" of the agent is an LLM is kind of irrelevant - you could also imagine the brain just being a program that generates random syntactically-correct code. What the LLM achieves is that, the random generator would take millions of years whereas the LLM is much more efficient at creating plausibly-working code. This is analogous to a chess engine's pruning heuristics.)

The hard part here is B. We've seen some great agentic successes when rewriting an existing project in a new language, since the agent can just use the project's prior test suite as its evaluator. But when developing a new project, you're still figuring out the finer details of how everything is supposed to work. As the old saying goes - writing a spec that perfectly describes how a program should work, is equivalent effort to just writing the program.

Stock ticker: "BIGCO'; DROP TABLE TRANSACTIONS"

To be fair, a game of this only takes a few minutes. Probably best as a "learn by doing" type of thing.

> Out of all the big orgs, Amazon is probably the one that has the least regard for keeping technically interesting projects alive

I'm curious what you mean by this. I would have said the exact opposite - AWS tends to keep projects around for a very long time. They haven't acquired very many open-source projects, but the few that they have are all still running as far as I know.

The Computer Fraud and Abuse Act explicitly contains "knowingly" and/or "intentionally" qualifications. By definition, you can't accidentally violate the CFAA.

I think "deleted" is still always going to have to be a package state, realistically. A package could contain spam, abuse, copyright infringement and/or illegal content.

I don't mind malware being lumped in to that - the place to look up code for historical or analysis reasons ought to be version control, not crates.io

I don't think parent commenter is complaining about the length. "a lot of words to describe..." is a rhetorical device. Their comment is basically saying "you wrote this whole article complaining about Amazon, but what they're doing is common and systemic."

Parent commenter never said it was Spirit. (I kind of doubt that they ever operated in Brazil.) Just that this is the kind of data they would be worried about being sold.

While I agree with the spirit of what you're saying, it is very similar to whenever anyone complains about things where they work. "Well, why don't you just raise it with management?" You're assuming that management

1) Understands the problem ("what are comments?")

2) Accepts that it is a problem ("how can more comments be a bad thing?")

3) Cares enough to solve the problem ("is this issue really a priority to solve right now? just accept the PR and we'll go back and fix if needed some other time cough never cough")

4) Believes you (this can take many forms, but the most common is, subconsciously, "this other engineer says it's not a problem, so I'll just assume it's not since that's easier")

All of these logistical, political and social factors are "the ocean"

LLMs are not the only form of AI.

The principal side effect is that you eat less.

Eating less can have positive effects (losing weight), but can also make you feel like shit (in a wide variety of ways), as your body is running on less energy intake than it has been accustomed to.

Depending on your activity level and the makeup of your diet, eating less can also lead to muscle loss and nutrient deficiencies.

His argument is that the government is going to start forcing tech companies to install backdoors in their tech, now that (hypothetically) they can't use hacking anymore.

This "going dark" scenario would require new legislation. With secure enclaves, modern smartphones can't be cracked open in the manner that the FBI wanted in the 2016 case. So there's no such thing as "court order tech company to crack phone" anymore. It would have to be "outlaw tech companies from producing phones that they can't crack open", which is very different and does not fall under any existing US statute.

What you said:

> A modern consumer GPU can crack "four random English words" in a day

What you're saying now:

> My mid-tier laptop GPU can crack the usual wordlist in under 12 hours against MD5 (46 billion guesses per second)

So your response to the plain mathematical fact that, no, your consumer GPU cannot crack "four random English words" in a day, is that your consumer GPU can crack four words chosen from a list of a few thousand, in a day. Followed by personal attacks. Okay.

> A modern consumer GPU can crack "four random English words" in a day.

This is just completely false.

> And before anyone blames the parents, parents will always be shitty. They are an impediment to children's health society must work around.

...huh? This is completely backwards. We should be making laws that punish parents for failing to parent properly. Not laws that punish everyone.

If we pass a law that children under a certain age may not use social media, then giving your child a non-child-locked device should be treated the same way we treat giving your child alcohol or cigarettes.

Personally I don't think that a release announcement needs to be "marketing heavy" in the first place.

That's really the heart of the problem - in an effort to sound exciting, many parts of the article are just wasting the reader's time rather than simply explaining how things work clearly and directly.

A stock exchange is nothing more than a marketplace. Anyone can start their own (the challenge being, getting a bunch of companies to want to list their stock with you).

That freedom is actually why it's hard to create a Bloomberg competitor - in order to provide their data stream, Bloomberg has to have partnerships with hundreds of companies worldwide who run exchanges for stocks, bonds and other instruments. If exchanges were government-controlled rather than free-market, there would be more centralization and getting the data together would be easier.

Yeah it's interesting how in some contexts they're interchangeable yet, culturally we very often use the words differently.

Personally I tend to feel like your second bullet - ethics are based on rules whereas morals are based on emotions.

Offtopic but, it always amuses me that the words "provably" and "probably" are one letter apart (and I frequently misread them for each other).

Probably never. This has been debated to death by both C and C++ standards committees. The consensus is that, since signed overflow is almost always a sign of a bug in the program, keeping it undefined enables compilers to optimize by assuming it never happens, and also allows sanitizers to continue to flag it to developers so that they fix their bugs (though it could be argued that, a sanitizer doesn't really have to strictly adhere to the standard, the committee apparently didn't feel that way).

It depends on the kind of company, and the context. An ethicist at a company doing drug/medical research probably mostly oversees that the company's research practices are legal and safe (and that participants give informed consent).

At typical corporation, something like a Chief Ethics and Compliance Officer (CECO) would be largely focused on legal compliance - making sure company personnel aren't engaged in bribery, fraud, insider trading etc. The value of this role (and the medical example) is to prevent lawsuits and/or prosecution.

For a company like Meta, an ethicist is likely concerned with the inherent ethical concerns of running a social media company, which impacts the mental health of its users, many of whom are minors. For AI companies, the ethics of releasing models that are capable of doing dangerous things. In name, the job probably involves giving recommendations about how to ethically implement features in the platform to minimize harm. In practice, the job probably accomplishes very little.