And we keep automating stuff that really shouldn't be. Classic example is customer support phone lines. I cannot remember the last time I used one of those and didn't have to ask for an operator. Believe it or not, most people call those lines because we need something to be done, rather than information that we almost certainly already have or can easily obtain via the website (which you keep advertising over and over again while we're on hold, and even before we ask the question, but has no options to contact a human other than calling this number.)

I kinda feel we need the entire commercial world to take a time out, and revisit everything we do that involves interactions between the different parties involved in commerce. Is it really the right approach for companies to value shareholder value over long term sustainability? Should customers and employees be valued just as much as the investors? And if you (not you Morromist, the generic you) are about to respond with "But Friedman said" or some other "economic" argument, why do we, as a society, allow shareholders to exist? Is our aim to ensure those shareholders can make money, or is our aim to provide funding to ensure the institutions needed to do everything from develop technologies to providing food on our tables can exist and get the funding they need?

Because I can tell you, I sure as hell wouldn't sign on to the entire that we create an entire legal structure aimed at ensuring people with money can make more of it, and damn the consequences.

I'm not entirely sure how a rant about automated customer service lines turned into this, but now I've written it I'm going to submit it anyway.

Good lord, we agree on something.

Anyway I think the issue is that there are idiots everywhere, and a lot of people who jump on bandwagons and are too proud to jump off once the evidence comes out it's a scam. And while I hate the idea personally, it sounds worse than flying, I would assume a lot of people look at the transportation speeds involved, and just assume it would be successful for that reason, especially in an era where a significant number of people believe America's railroads collapsed due to "flying", when the story is way, way, more complicated (essentially a systemic shock from all sides, over regulation, poorly thought out taxes that weren't applied to rival transportation systems, poor management, and out of control dumbass unions. A perfect storm of crappiness.)

The bubble will need to last 5 years before there's any chance of new players.

40 years ago, sure, you might have been right: silicon fabrication was still a relatively cheap industry to get into. Leaving aside environmental concerns (our old friend MOS Technology of the 6502 managed to end up as a superfund site...) it was quite simply a lot easier to make foundaries that worked on a micrometer scale than on today's nanometers. Gordon Moore said that while his eponymous law predicted a doubling of transistors per IC each year, he also saw a doubling of how much it cost to build a new foundary.

If it were cheap, even in countries that already have fewer regulatory hurdles, someone would have build new factories anyway. It's not like there wasn't any profit in memory production until a month ago. They're not commodity devices.

The article talks about a rise in temperature of half a degree causing a societal collapse over multiple centuries.

By comparison the Earth has heated by 1.1-1.2C in the last 150 years and is still rising, currently at a rate of about 0.2C per decade.

So, meteoric rise vs slight rise within the Earth's usual temperature fluctuations. This should tell you two things:

1. The current rise is, indeed, unnatural, and not due to the Earth's usual changes.
2. This is going to cause a hell of a lot more damage than one civilization smaller than the size of a typical European city collapsing.

> * rise of AI that takes control

I think that's unlikely, but I think the AI crap we're currently following suggests a different path.

The LLM fad is most probably going to eat itself, but take down a lot of things with it.

Let's take a look at it: We had the World Wide Web.

The web was built over a period of a couple of decades maybe (by the mid-2000s it could be considered the primary source of knowledge for everyone in the developed world), with virtually everyone switching to it en-mass. Newspapers went online and closed their paper versions. Magazines closed, to be replaced by websites that were vaguely related. People who once might have written books now wrote blogs or maintained websites with information that was dear to them. Manuals went online. Programmers I think know this more than most, it's easier to search for "Java list to array" than it is to actually go to a physical book and, even with the help of an index, find the method that does that. (Spoiler: Java's version is an ugly hack! I wish Java had kept the original versioning scheme as that'd have made it easier for them to make a Java 2 that didn't rely on the way Java 1 does things, breaking some backward compatibility but implementing things properly. Anyway, I digress.)

So then the first hit came, social media. Not a major hit, and not an obvious problem at first as the first social media sites were glorified blogging platforms. And that was fine, except one of these, Facebook, felt that it was in its best interests to hinder search engines from working with it, which undermined the web.

Then the next hit, the iPhone and Android. Again, not initially a major problem, but over time they encouraged massive amounts of content to be locked behind "apps", with any web interface being a third class option or worse.

Then Facebook came up again, they REALLY didn't like Google. How could they kill Google? By making normal content unsearchable, not just content posted behind Facebook's registration wall. So again, not caring about the web, they came up with a scheme to encourage everyone to post everything as videos. They lied about their own metrics, claiming it showed videos caused a crazy amount of engagement, and the rest of the content providers saw this, and put autoplaying videos on every web page, very often (as was Facebook's intention) without any readable text. Google's search couldn't actually search the videos at that time, so it undermined Google.

Meanwhile... Google was fucking around too. They intentionally made their own search engine less useful, noting that their competitors were apparently just copying them rather than trying to make their own search engines more useful than Google's. So now a Google search is very, very, unlikely to give you relevant answers without a lot of adding quotes to things.

Finally, LLMs. LLMs are the nail in the coffin. They remove any incentive to post anything on the Internet aside from documentation, advertising copy, and rants. There is decreasing incentive at this point to post anything you'd be paid to write or just to make the world a better place. Whether it's journalism, fiction, research, or even a Wikipedia page, fewer and fewer people will ever read what you have to say, but an LLM will, combine it with eleventeen other sources, and produce a summary that's inaccurate and dubious but is "good enough" for people looking for a quick answer.

So what we're left with are LLMs that will tell you things based upon the latest information as of 2025. They'll tell you the latest information as of 2025 today. And they'll tell you the latest information as of 2025 in five years. And in ten years, unless someone's willing to employ an army of researchers whose sole job is to write well researched and accurate articles to be ingested by an LLM, they'll be useless.

But so will the web.

And we don't have a replacement for the web, which is unfortunate because we don't have a replacement for the things we replaced using the web.

And as should be obvious, that means a world without information.

And I can't see how our civilization survives that.

Biden doesn't have dementia. He was showing signs of normal age related decline, sure, as would anyone his age. But his handlers wouldn't have let him out in public, even to speak from a teleprompter, if he had it. As someone who's lived with two people with two different types, I can attest that if he had it, it'd have rather more obvious than "mumbling", which isn't a symptom anyway. (And reading the transcript of the debate where his mumbling was *such* an issue, I think it's rather obvious he's thinking clearly throughout.)

The only neurological condition Biden had was his long term stuttering issue, which he's always had. Unless you're suggesting he had dementia in the 1980s, when his reputation for "gaffes" started, I think the argument he had it during his presidency is a thin one.

I find it all the more ludicrous Biden's opponents keep bringing this up as some sort of gotcha when their own guy very obviously has a major issue. Does Trump have dementia? Can't say! But certainly his constant incoherent streams-of-consciousness speeches raise alarm bells, and are far closer to what I saw in my FiL than anything Biden said, and Trump's actions appear to be substantial enough that his doctors are interested, hence the constant MRIs etc. And Biden's opponents are the ones that would fight to the death to condemn anyone who suggests St Reagan should have not run a second term.

As for orange suits, Biden followed the law. Trump's only getting away with what he's doing because of a scared Republican congress and a SCOTUS that's made it clear they believe in the supreme executive theory, despite that being literally unamerican, a rejection of the entire foundation of this country.

Right now, we're noticing that Chinese companies are offering us exploitative deals, and we don't like it, and think that tariffs will fix it. But with tariffs in place, we will find that now it's American companies that are offering us the exploitative deals, but they can charge more now, because they're insulated from outside competition. What I'm saying is that intranational capitalism is just as sleazy and brutal as international capitalism - only less efficient, because it's less competitive.

I agree with you about NATO states paying for mercenaries to defend Ukraine. There are many experienced fighters in Africa and the Middle East, and it's likely that their services could be bought at a reasonable price. Without more soldiers in the field, Ukraine's front is doomed to collapse.

I love AI and I would and could pay for it if I had to, but why would I pick OpenAI to pay? Their product is not really better than their competitors' products, and sometimes it's clearly worse. They have the advantage of being the first mover in their field, and that gives them inertia with low-information customers - the new AOL.com. But apart from that, they have huge debts and not much else to distinguish them. Their best employees had left, and their former partners have become wary of the way they operate. Projections of their future profitability must be based on the expectation that their AI will figure out some better business plan than what the OpenAI humans have come up with!

Arduino responded to this recent drama just a few days ago, saying "Our 20-year commitment to open-source is unwavering" with a good explanation of the new T&C.

https://blog.arduino.cc/2025/1...

I believe Arduino is sincere with their statement.

One man speaking with Adafruit's social media accounts seems to feel otherwise. He probably believes he's doing good by raising the alarm. Maybe some of the points have some merit? But the tone really looks like an attempt to stir up drama and harm Arduino's reputation.

Adafruit does have history with Arduino. In 2015 when Arduino had serious internal division and conflict, Adafruit was manufacturing brand name Arduino Uno under some sort of license deal. That arrangement ended sometime in 2016. Adafruit quickly launched a product line of essentially Arduino clone boards named "Metro". Does any of that matter? Maybe, maybe not. But when reading what really looks like an attack on Arduino's long-established reputation coming from official Adafruit channels, best to keep in mind those 2 companies have a history.

I also have some history with Arduino, having made an Arduino-compatible board and contributed code and help over the years. I've personally met the Arduino developers and Arduino leadership folks several times at conferences. They are genuinely good people who've poured a lot of effort into trying to good in the world.

Maybe Arduino change for the better or for the worse with Qualcomm. I don't have a crystal ball. But I'm trying to keep an open mind and not get caught up in fear over basically boilerplate legalese.

Contrary to the scaremongering we hear, the world is not full of terrorists. They exist, don't get me wrong, but the reason we're so afraid of them is that our fear of them is incredibly useful for people who want to control us.

And also, the ones who do exist aren't engineers. So a shooting, or a stabbing, that's pretty easy to figure out how to do. Derailing a train? That's physics. Physics is much harder than pointing a gun and pulling the trigger.

Unfortunately Acela sucks. I took Acela from New York to Boston once. Once. It was terrifying—the tracks aren't really suited for running at (haha!) 100mph. The idea that this is high-speed rail and that anybody takes that name seriously just illustrates what a backwater the U.S. is nowadays.

There is already a rail corridor through western Indiana into the Chicago metropolitan area. And there are already passenger trains running on it. The problem isn't getting a train into the city center—it's that we don't have electrified high-speed rail lines between the cities. Which, given that we do have low-speed (only 75mph max) highways, which are insanely expensive to build and maintain, seems like an eminently solvable problem.

The real problem is that there are huge fortunes dependent on keeping those roads full of cars. But really that's not even the problem. You can see the problem right here in this discussion: if you haven't lived in a place where high speed rail is ubiquitous, it seems really really hard. If you haven't lived in a place where cars are not completely and utterly dominant, it seems inconceivable that things could be any different. Even people who are anti-car tend to think with car brain because of this.

Microsoft is not a person. It is a massive company capable of pursuing many mandates, some of which can either appear or can actually be entirely at odds with each other.

It most certainly did. The summary even discusses it.