hinterlands

It is fairly rare to see an ex-employee put a positive spin on their work experience.

I don't think this makes OpenAI special. It's just a good reminder that the overwhelming majority of "why I left" posts are basically trying to justify why a person wasn't a good fit for an organization by blaming it squarely on the organization.

Look at it this way: the flip side of "incredibly bottoms-up" from this article is that there are people who feel rudderless because there is no roadmap or a thing carved out for them to own. Similarly, the flip side of "strong bias to action" and "changes direction on a dime" is that everything is chaotic and there's no consistent vision from the executives.

This cracked me up a bit, though: "As often as OpenAI is maligned in the press, everyone I met there is actually trying to do the right thing" - yes! That's true at almost every company that ends up making morally questionable decisions! There's no Bond villain at the helm. It's good people rationalizing things. It goes like this: we're the good guys. If we were evil, we could be doing things so much worse than X! Sure, some might object to X, but they miss the big picture: X is going to indirectly benefit the society because we're going to put the resulting money and power to good use. Without us, you could have the bad guys doing X instead!

My local Cinemark has tickets for $5.50, $8.50... you're probably in a premium market.

All the 'box office records' since then are the result of charging way more to a continually plummeting audience size.

I don't think that going to the movies has gotten more expensive in real terms. It's just that the records are usually not adjusted for inflation, so a film with the same audience and the same inflation-adjusted admission price will appear to make 80% more at the box office compared to 2002.

Yes. Look, I'm not arguing this is bad, I'm just trying to respond to the original question and capture the essence of the debate.

There are three pertinent points: (1) it's EDTA; (2) it's not that EDTA is safe or not safe, it's that no one applied to have it approved as an OTC medication; (3) you can still (probably) sell EDTA as a supplement in the US, but the FDA grumbled about it, which angered various chelation cranks.

Probably, but the process doesn't work that way. The default is that you can't sell medication to people, period. Some pharmaceutical company applied to have a specific form of EDTA approved as a prescription drug, and that was that.

Separately from this, substances that meet the criteria of being "natural" can be sold as supplements as long as you don't claim they cure anything. EDTA is naturally-occurring and you can buy it as a supplement in the US, although the FDA has some beef with this, which I think is what the original remark might be alluding to.

EDTA is also a common food additive and a laboratory reagent, so people who want to use it can buy it easily, which makes the whole debate basically performance art.

It's about EDTA. It can be legitimately used to treat heavy metal poisoning, plus some other things. Some people (who are probably misguided) want to self-medicate. The FDA won't let you. Hence, drama.

Because what's the headline you're going to get out of it?

If the headline is "Mark Zuckerberg is amassing your data and you know it's for evil", it's an easy sell. If it's "there's an ecosystem of little-known companies that sell transaction, location and lifestyle data to marketers, journalists, PIs, and police departments alike", it's not exactly the kind of a message that spurs people to action. And yeah, the newspaper that would be breaking the news is a customer too.

I'm not quite sure what you're saying here. My assertion is that a visible image persists on the screen longer than it appears in the slo-mo clip. You can just point a camera with an adjustable shutter speed at a CRT and see it for yourself. Here's an example (might need to copy the URL and open in a new tab, they don't like hotlinking):

https://i.sstatic.net/5K61i.png

The brightly-lit band is the part of the frame scanned by the beam while the shutter was open. The part above is the afterimage, which, while not as bright, is definitely there.

I think it's fairly common for technologies to get really good just as they're becoming obsolete. Vacuum tubes, CRTs, optical disks, photographic film... in fact, they're often in some respects better than the early generations of the technology that replaces them.

But OLEDs just have too many advantages where it actually matters. Much lower power consumption, physically more compact (no need for backlight layers), etc.

That slo-mo video is somewhat misleading, though. The phosphor glows for a good while, so there is a reasonable chunk of the image that's visible at any given time.

The problem in that video is that the exact location the beam is hitting is momentarily very bright, so they calibrated the exposure to that and everything else looks really dark.

The Chinese medicinal herb you're thinking about is sweet wormwood, from which we isolated artemisinin. Artemisinin isn't a suppressed secret. It's one of the major treatments for malaria and it netted its discoverer a Nobel prize.

The drug the article is talking about contains artemisinin in combination with another substance.

We know how to deal with malaria, so this isn't some story of Big Pharma hiding the truth. The disease used to be endemic in the US and in Europe. Better treatments save lives, but eradication hinges on economic and political factors... which are in turn not helped by malaria.

Also, LLMs can generate prompt files too.

Sure, but the utility of that for PCB design wasn't demonstrated in the article. This is an expert going out of his way to give the LLM a task it can't fumble (and still does, a bit).

Similarly to how most web dev isn't exactly on the frontiers of computer science, a lot of day-to-day PCB design isn't about cutting-edge analog or radio stuff. It's just putting the same MCU or SoC on differently-shaped boards over and over again.

If you can reliably automate that, it's still a pretty big deal.

While I think that AI tools can be quite useful for coding, PCB design, and other tasks like that, the setup of this experiment makes it really hard for the LLM to fail.

The author's prompt is basically already a meticulous specification of the PCB, even proactively telling the LLM to avoid certain pitfalls ("GPIO19 and GPIO20 on the ESP32-S3 module are USB D- and D+ respectively. Make sure these nets are labeled correctly so that differential routing works"). If you had no prior experience building that exact thing, writing that spec would be 95% of the work.

Anyway, I don't think the experiment is wrong, but it's also not exactly vibe-PCBing!

I do think it's an SFBA / generational bubble. We have plenty of boring, expensive software projects that someone will always bring up in a HN thread. For example, every time there's a thread on PCB design, you have some folks talking about Cadence. What's there to say about Cadence? Well, first and foremost, it costs a lot. Otherwise, it lets you design PCBs. But there are people here who pay for it, use it, and want to talk about it.

The article starts by dismissing scrollsaws as "pretty darn dangerous", but that's a pretty big stretch. They're less dangerous than a sharp kitchen knife. You want to talk to your kids and watch them closely the first couple of times they use it, but you'd be hard-pressed to find any accounts of serious injuries caused by scrollsaws.

This toy doesn't seem bad as a crafts tool that buys you several quiet weekends, but at $250... that's actually more than a miniature desktop scrollsaw (Proxxon 37088).

(Note: the original title was "Computer Scientists Figure Out How To Prove Lies" before being changed by the admin)

I honestly think that Quanta Magazine just found the perfect formula to farm HN traffic. The titles are clearly carefully engineered for this audience: not the outright clickbait of university press releases, but vague profoundness that lets us upvote without reading the whole thing and then chime in with tangential anecdotes.

I don't think they're bad people, but I honestly think they end up on the front page multiple times a week not on the quality of the articles alone.

First, it just reads that way. It's the default style if you ask ChatGPT to write a couple of paragraphs that explain why lightfastness is important.

Second, while I know there are reasons to be skeptical about AI text checkers, the author's earlier (less verbose) style doesn't get flagged at all, while the style in more recent articles gets classified as heavily AI-assisted.

I think there's a fundamental disconnect here: the article says that you should be focusing on strategies that, for the most part, make aging more dignified. The goal shouldn't be even curing cancer. And maybe that's right.

But the reason billions of dollars are poured by SFBA VCs into aging research is probably just that they're getting older, they don't want to die, and they figure that they can put some of their money into anti-aging moonshots. It's not really different from rich people getting cryogenically frozen. If you have more money than you can possibly use, why wouldn't you try?

Although someone will challenge me on that, I'm 100% sure that large chunks of the text are AI-generated. That said, the website itself has been around at least since 2017 (the text just wasn't as verbose - e.g., https://sarahrenaeclark.com/diy-gift-bag/).

So, I suspect it's legit. It's a case of an author leaning on a crutch for writing, but we're here to judge the results, not the phasing.

The problem is that these "small things" are not necessarily small if you're an individual.

If you're an individual pirating software or media, then from the rights owners' perspective, the most rational thing to do is to make an example of you. It doesn't happen everyday, but it does happen and it can destroy lives.

If you're a corporation doing the same, the calculation is different. If you're small but growing, future revenues are worth more than the money that can be extracted out of you right now, so you might get a legal nastygram with an offer of a reasonable payment to bring you into compliance. And if you're already big enough to be scary, litigation might be just too expensive to the other side even if you answer the letter with "lol, get lost".

Even in the worst case - if Anthropic loses and the company is fined or even shuttered (unlikely) - the people who participated in it are not going to be personally liable and they've in all likelihood already profited immensely.

The actual claim is that you can keep the game state for a rather minimal 8x4 game board in two uint32_t, one uint64_t, and one uint_8t.

That's not the entirety of program state - stack aside, there are some loop counters that the author isn't including in the total, along with some implicit variables e.g., for keypress data.

It's a nice optimization exercise, but the result is not particularly mind-blowing to me. You could probably make it significantly more compact at the expense of readability. For one, the shape of the snake on the game board is not random: there's plenty of combinations, such as an alternating checkerboard pattern, that are physically impossible. So the shape is likely compressible.

Are controls uniquely important, though? There are hundreds of things in a car that could be made better (more durable, longer lasting, better looking) for just $10 to $100 extra a piece. But it adds up.

It's not just cost, though. The reality is that consumers like the futuristic look, in theory (i.e., at the time of the purchase). Knobs look dated. It's the same reason why ridiculously glossy laptop screens were commonplace. They weren't cheaper to make, they just looked cool.

The other problem are people doing price arbitrage. You find the item on eBay and think to yourself, "cool, I'd rather patronize a small business" - but as it turns out, the item is drop-shipped from Amazon, Walmart, or the like.

Because most of the "loopholes" aren't actually loopholes: they are created for a specific reason under some specific economic theory. Most often, to encourage people to make certain types of investments, avoid double- or triple-taxing certain activities, etc.

We just almost never talk about it in neutral terms: why was this policy implemented, what are the pros and cons, etc. Instead, it's just political talking points to get people to the voting booth.

More to the point, it's a field where we're constantly told that our experiences from a month ago are in no way relevant and that the latest thing is fundamentally different to what we know. Should expertise degrade just as quickly?

In case you wondered, $1M in cash ($100 bills) weigh approximately 15.4 pounds (about 10 kilograms).

Your answer is incorrect. You asked Claude to calculate $7M, which netted 154 pounds, but you then divided it by 10 instead of 7 to get the weight of $1M.

Further, it's quite irrelevant here, as the display involves $1 banknotes, not $100 bills. The correct answer, without the need for an LLM, is: 1 million bills times one gram = 1 million grams = 1,000 kg = 1 metric ton.

These MCUs are generally targeted for low-power uses and various industrial & automotive stuff. They're not meant for anything high-bandwidth - you'd be using a 32-bit ARM or RISC MCU for that, or a SoC.

So, think of it more in terms of offloading simple automation or logic glue to dedicated hardware to save a microamp or keep the state machines tidy.

It's being used in a way where biases matter. Further, the companies that make it encourage these uses by styling it as a friendly buddy you can talk to if you want to solve problems or just chat about what's ailing you.

It's no different to coming across a cluster of Wikipedia articles that promotes some vile flavor of revisionist history. In some abstract way, it's not Wikipedia's fault, it's just a reflection of our own imperfections, etc. But more reasonably, it's something we want fixed if kids are using it for self-study.

Train it only on Wikipedia

Minus most of history...