chemotaxis

I feel like everyone in this thread is assuming this is a good faith move by Australia to help kids in school and with socialization.

I mean... you can say that about most of things in life. Behind every social movement or policy, it's always a mix of good faith, cynical fearmongering, and opportunism by people or organizations who stand to gain something from it. Does it matter?

If you think that social media and smartphones are harmful to the youth, you (a) should probably be glad that someone is doing something decisive about it; and (b) you get a large-scale experiment that will hopefully prove or disprove that.

Probably because AI appears to work, more or less, and now it's just a race to make it better and to monetize it.

Before ChatGPT, I'd guess that the amounts of money poured in both of these things were about the same.

I've known quite a few people who went to therapy and I'm not sure that's even the right question to ask. I don't think they were paying to get helped as much as they were just paying to have someone to talk to. To be clear, there are people who genuinely need help, but for most, a therapist is probably just a substitute for a close friend / life coach.

And say what you will about this, a paid professional is, at the very least, unlikely to let you wind yourself up or go down weird rabbit holes... something that LLMs seem to excel at.

This wouldn't ban the behavior, just the disclosure of it.

The best part is that this article is almost certainly AI-generated or heavily AI-assisted too.

Before people get angry with me... there's plenty of small tells, starting with section headings, a lot of linguistic choices, and low information density... but more importantly, the author openly says she writes using LLMs: https://www.sh-reya.com/blog/ai-writing/#how-i-write-with-ll...

I think the author is speaking authoritatively about things they may be less familiar with, or where they really want to push a particular doomsday / degrowth agenda (the only prescription at the end the article is that we need to stop technological progress). This paragraph in particular caught my eye:

Bah! Who needs copper anyway, when we have so much aluminum?! Have you thought about how aluminum is made? Well, by driving immense electric currents through carbon anodes made from petroleum coke (or coal-tar pitch) to turn molten alumina into pure metal via electrolysis. Two things to notice here. First, the necessary electricity (and the anodes) are usually made with fossil fuels, as “renewables” cannot provide the stable current and carbon atoms needed to make the process possible. Second, all that electricity, even if you generate it with nuclear reactors, have to be delivered via copper wires.

This seems to be trying to say that we can't make aluminum without copper, but that seems nonsensical. First, power can be delivered by wires made out of aluminum and indeed, it often is - I don't think that much of the transmission grid is copper. Second, the comparatively tiny amount of material needed for electrodes is a completely wacky argument. And renewables not being able to provide "the stable current" needed for smelting?

I'm not cherrypicking here, there's a lot of assertions of this type in the article. Essentially, everything is doomed and there's nothing we can do, because we're going to run out of copper. And fossil fuels. And there's absolutely nothing that can replace them, ever. And therefore, we shouldn't build AI datacenters? That's what it says...

This is not necessarily a fundamental limitation. It's a consequence of a fine-tuning process where human raters decide how "good" an answer is. They're not rating the flow of the conversation, but looking at how complete / comprehensive the answer to a one-shot question looks like. This selects for walls of overconfident text.

Another thing the vendors are selecting for is safety / PR risk. If an LLM answers to a hobby chemistry question in a matter-of-factly way, that's a disastrous PR headline in the making. If they open with several paragraphs of disclaimers or just refuse to answer, that's a win.

It's really getting in the way of all the daily AI opinion pieces I come here to read.

More seriously, there are tens of thousands of people who come to HN. If Fourier stuff gets upvoted, it's because people find it informative. I happen to know the theory, but I wouldn't gatekeep.

4562, 8825, 1065

I think that's common in most places. What's different in the US is that the IRS forces you to proactively provide a lot more information about it, though. I have a rental property and need to enter the same information about the same income and expenses on three different forms, breaking it down in different ways. It's tedious and error-prone, and I guess the philosophy is that it's easier to spot fraud if the numbers on all the different forms don't add up to a coherent story.

Other countries presumably rely on other fraud signals. They might have more visibility into your day-to-day financial transactions, or there might be more of a culture of leaving an anonymous tip if you suspect your neighbor isn't paying a fair share.

I think the biggest mistake people make when thinking about mathematics is that it is fundamentally about numbers. It’s not. Mathematics is fundamentally about relations.

Eh, but you can also say that about philosophy, or art, or really, anything.

What sets mathematics apart is the application of certain analytical methods to these relations, and that these methods essentially allow us to rigorously measure relationships and express them in algebraic terms. "Numbers" (finite fields, complex planes, etc) are absolutely fundamental to the practice of mathematics.

For a work claiming to do mathematics without numbers, this paper uses numbers quite a bit.

I think that sort of goes hand-in-hand. "Normal", well-rounded people don't decide that software licensing is the most important thing in the world and don't devote their entire life to that. A normal person would be content with a 9-to-5 software engineering job at Sun, IBM, or Microsoft.

I think you see that with a lot of other revolutionaries. They often take unreasonable positions and behave in unreasonable ways. RMS' tragedy is probably that his side more or less won, so now he's just a weirdo without a cause.

The number of people you want to follow is much smaller than that 0.1%.

We're talking about bloggers reaching their audience. The audience they can reach via Mastodon is much smaller than on Twitter, even if you factor in the consequences of algorithmic feeds.

Sadly, I have realised fewer people actually give an F than you realise; for some, it's just a paycheck.

I found that most of the "people problems disguised as technical problems" are actually generated by people who get far too invested in their work and let it define them. They get territorial, treat any lost argument as an attack on their whole self, etc. They also lose perspective, getting into flame wars over indentation styles or minor API syntax quibbles.

People who show up for the paycheck are usually far more reasonable in that regard.

I think this ignores the codebase churn in Big Tech. The code you write today probably won't be there in ten years. It will be heavily refactored, obsolete, or the product will be outright canceled. You can pour your heart in it, but in all likelihood, you're leaving no lasting mark on the world. You just do a small part to keep the number going up.

Tech workplaces are incredibly ephemeral too. Reorgs, departures, constant hiring - so if you leave today, in 5-10 years, there might be no single person left who still remembers or thinks highly of the heroic all-nighters you pulled off. In fact, your old team probably won't exist in its current shape.

If you build quality furniture for your customers, chances are, it will outlive you. If you work on some frontend piece at Amazon, it won't. I think the amount of pride in your workmanship needs to scale with that.

Mastodon, you'll see all of them.

Alone... look, I want Mastodon to be successful, but revealed preferences don't lie. Mastodon MAU is about 0.1% that of Twitter, down more than 60% from the peak.

The main thing is that no one wants the hassle of keeping up with 50 mildly-interesting blogs by visiting them regularly. You really need a "push" mechanism of some sort. Social media doesn't work for this because if someone subscribes to a content creator on X / Twitter, they most likely won't see most of the creator's posts. Instead, the algorithm will show them cat memes and other on-platform engagement bait.

Many other social venues are gone too. If you're lucky, you can reach your audience on HN, but it's about the only remaining, successful aggregator of this type. Reddit has grown a lot more insular and many subreddits don't allow outgoing links. Where else do you go?

In this reality, the most practical push mechanism is email, but sending email to thousands of recipients is hard. You pretty much need to pay someone for the privilege if you want to have a reasonable success rate. Substack will do it for you for free, and it also lowers the friction because it gives visitors a familiar UI with a pre-filled address and no concern about phishing / spam / etc.

Beyond that, I don't think Substack is actually that much of a community. They built a good brand by attracting (buying) a bunch of high profile writers, then had an issue with neo-Nazis where they took controversial stances... I don't associate the domain with anything especially good or bad, not different from blogspot.com or wordpress.com. I have a special hatred for medium.com because almost everything over there is aggressively paywalled, but that's another story.

And yeah yeah, RSS, but the friction for RSS is much higher.

I'm surprised these newsletter gatekeepers haven't implemented a tip jar where you put in $/year and it gets divided based on readership.

I've seen a bunch of publication with a "tip" button, but I suspect it's not worth the effort. Very few people pay in the first place, so a random one-off payment of $1, $10, or even the "unicorn" $100 is not worth standing up the infrastructure and dealing with the tax paperwork.

On the flip side, if you find 100 people who really like your content and are willing to substantially support it on an ongoing basis with a subscription, you end up with recurring revenue that makes it a better deal.

Better yet, many of the graduates will become politicians, journalists, or prominent tech figures who will be pontificating about morality and regulating it for others.

I think your attempt to rebuke the proof is flawed too. The problem in your reasoning is mixing up "arbitrarily many" and "infinitely many".

There's no convergence after a finite number of steps. But at infinity, the canonical limit of this construction method is a circle. And because it is a circle, the circumference at infinity "jumps" to 2*pi. This is quite counterintuitive but perfectly legit in mathematical analysis. It's just one of many wacky properties of infinity.

My favourite part of climbing the corporate ladder is finally having enough clout to just say "no".

But the thing is, in big companies, you can keep climbing all your life and never reach that level. You can be a VP and it still doesn't matter, because you're one of several hundred VPs in a company that values lawsuit-proof consistency over giving executives infinite latitude. You're still dealing with the same spreadsheets, processes, and meetings as everyone else. At best, you might be able to send your minions to some, but that doesn't solve the issue, it just messes up someone else's day.

In a company of several hundred people, on the other hand, you don't really need to climb far because relatively little is set in stone. So to folks who are at Amazon, Microsoft, or Google, and who're waiting for the day when they're finally free to say "no" to overhead, I have bad news.

Most importantly, Clippy happened before computers were permanently online. For it to harvest any data, I guess you'd need to mail in a floppy disk.

In a lot of ways, I'm thankful that LLMs are letting us hear the thoughts of people who usually wouldn't share them.

I could agree with you in theory, but do you see the technology used that way? Because I definitely don't. The thought process behind the vast majority of LLM-generated content is "how do I get more clicks with less effort", not "here's a unique, personal perspective of mine, let's use a chatbot to express it more eloquently".

It's not just LLMs, it's how the algorithms promote engagement. i.e. rage bait, videos with obvious inaccuracies etc.

I guess, but I'm on quite a few "algorithm-free" forums where the same thing happens. I think it's just human nature. The reason it's under control on HN is rigorous moderation; when the moderators are asleep, you often see dubious political stuff bubble up. And in the comments, there's often a fair amount of patently incorrect takes and vitriol.

If a company’s crypto holdings are worth more than its market cap wouldn’t that mean I should buy the stock as its enterprise value is higher than its current market cap?

Well, one question to ask is if the company can actually cash out their crypto holdings at the current market price? If they have substantial holdings of thinly-traded crypto, the answer is usually "lol".

Faith in the company matters too. You can't compel them to sell, and if you think the company is going to slowly drive itself into the ground, you probably don't want to give them your $$$. There are quite a few companies that traded below their book value for that reason. IIRC, Kodak is a current example. Want to bet on that?

Yeah, I don't like this. By insisting that the main tell of AI writing is that it's comically inaccurate, it accidentally enforces a pretty bad association: that if something looks accurate, it's not AI slop.

I think you gotta start with a definition of what AI slop is and why it matters. Most of what LLMs generate is not obviously incorrect.

It's why people can't post a meme, quote, article, whatever could be interpreted (very often, falsely) as AI-generated in a public channel, or ask a chatbot to explain a hand-drawn image without the off chance that they get an earful from one of these 'progressive' people.

It happens, but I think it's pretty uncommon. What's a lot more common is people getting called out for offloading tasks to LLMs in a way that just breaches protocol.

For example, if we're having an argument online and you respond with a chatbot-generated rebuttal to my argument, I'm going to be angry. This is because I'm putting an effort and you're clearly not interested in having that conversation, but you still want to come out ahead for the sake of internet points. Some folks would say it's fair game, but consider the logical conclusion of that pattern: that we both have our chatbots endlessly argue on our behalf. That's pretty stupid, right?

By extension of this, there's plenty of people who use LLMs to "manage" their online footprint: write responses to friends' posts, come up with new content to share, generate memes, produce a cadence of blog posts. Anyone can ask an LLM to do that, so what's the point of generating this content in the first place? It's not yours. It's not you. So what's the game, other than - again - trying to come out on top for internet points?

Another fairly toxic pattern is when people use LLMs to produce work output without the effort to proofread or fact-check it. Over the past year or so, I've gotten so many LLM-generated documents that simply made no sense, and the sender considered their job to be done and left the QA to me.

I think this is pretty common for Quanta, and it might be sticking out more because it's a field we're familiar with.

I'm really torn about this, because I think they're providing a valuable service. But their general formula doesn't diverge a whole lot from run-of-the-mill pop-science books: a vague, clickbaity title and then an article that focuses on personalities and implications of discoveries while glancing over a lot of important details (and not teaching much).

There have been studies this year implying that some brain functions rely on quantum interactions.

I'm not in the camp of "the brain is solved science", but come on. All of chemistry depends on quantum interactions and that didn't stop us from understanding a lot of it. It just means we're solving a different set of equations.

It's a big leap to go from "brain electrochemistry is described by quantum mechanics" to "the essence of consciousness / human soul is hiding in the quantum realm and therefore can't be measured or replicated".

There are just two photos in the entire gallery of 150+ that show modern scopes. The rest is vintage CRT equipment, going back to black-and-white films.