I think the author is mostly right about the current state of AI, but his future predictions (or worries) are based on a false premise: that the massive LLMs will keep improving in the future.

As far as I have seen the improvements have clearly slowed down, while the energy consumption is rising linearly (or worse). It's like the energy (money) vs. performance graph is logarithmic, and the companies are expending double the energy to get a 10% improvement. Something like that is not sustainable, and the money seems to indicate so.

I really think that LLMs are a dead-end for AI. A really useful dead-end, once the bubble pops and with time, we get a useful working model for them, probably based mostly on local LLMs, maybe using specialized training data.

Using the law of conservation of matter and extending your own logic, matter (mass) can never be created or destroyed, only transformed, so weapons of mass destruction don't exist.

Control and censorship. Imagine a developer distributing an apk that Google doesn't like, let's say, emulators, ad-blockers, VPNs (if they become illegal in some places), piracy apps or many others. Just remove their verified status and voilà.

This is awful. Google is quickly becoming a new walled garden, Apple-style.

The decline began with Reagan, and accelerated with 9/11.

1. Linux runs games well now.
2. New computers are expensive, and Windows runs like poop on old computers.

I don't think Firefox is immune to this, just that because its architecture is different, Anthropic didn't bother coding a bridge for it (given its market share).

The main issue here is that Anthropic violated one of the most important (implicit) tenets of applications in a computer: don't touch other people's shit. Claude.app modified Brave (and others) configuration, adding an extension without user consent. An extension that, by the way, gives full control of the browser to Claude, including reading the DOM for browser tabs unrelated to Claude (for example, the one where you just entered your credit card details).

I think they're a dead-end mostly because of the exponential cost vs. performance. The decreasing returns are obvious, and the companies are trying to adapt by raising token prices, but that will not be enough with the current user numbers (or even double or triple, if we believe the analysts). I think that, at least with these large LLM companies, we're actually beyond the point of economic equilibrium with this technology, at current energy and water prices.

And yes, training is more expensive than usage. That's probably the reason why Anthropic suggested a pause in LLM development (training), supposedly because of the fear that AI could become Skynet, but really because they are getting an IPO soon and if people see their current balance numbers, the IPO would fail and the bubble would probably pop. Which really proves my point a little: the economics of these companies "improving their LLMs" (training) don't make sense at current energy prices.

I don't know if it's my favorite, but Prototype (and its sequel) comes to my mind. It wasn't that badly received, but most people liked Infamous more, in the same generation.

And a game that I really love, my personal best SNES game ever (yes, even beyond Chrono Trigger and Tales of Phantasia, my numbers 2 and 3), is Terranigma, which is a game that many people who've played it don't like (because they get stuck at the beginning in the Underworld or at the fight with Bloody Mary).

I also used gopher, usenet, uucp, archie and all that in the early '90s. And if you were just a little bit careful, catching a virus or trojan was very rare.

Now, everything is monetized. You're under constant surveillance on a scale it was unimaginable at that time, and being careful amounts to nothing (unless you're super paranoid) because your security depends completely on how others secure your data, assuming that they're trying to secure and not sell it.

In this respect, now it's much, much worse than it was, even if in some other ways it's somewhat better.

Peak internet was around 2005-2010, IMHO.

Maybe don't assume everyone on the fediverse is USA-ian.

At least the AMD system management requires physical access (the AMD PSP does not have a network stack). Intel ME / AMT does have a network stack, and it hides its packets inside the host traffic. That's the reason of the black holes on many Intel CPUs when listening on ports 16992-16995 (the host does not see incoming traffic to those ports because the AMT intercepts it).

One could accuse Elden Ring of many things, but clunky controls is definitively not one of them.

It's probably one of the best combat systems I've ever played. When you die, you know it's your fault (usually because of greed), not the system cheating you. It's very fair, unlike many others.

I would amend that to "since WWII", and the answer would still be no.

I disagree. The Nixon government was awful, but so were many governments before. It wasn't a decline per se, it was just part of the constant change (both good and bad) in leadership.

The real decline began when corporatism started to take over the political/legal system, which was mainly a Reagan phenomenon. Large businesses started pulling everything towards their shareholders, nothing for the workers. Syndicates lost power. The political system started being bought by the corporations. Whole laws started being bought by the corporations.

While corruption has always existed, previously it was something to hide, to be ashamed of. But with the corporate takeover of the USA it gradually began to be normalized, and now it's almost expected from any relevant figure.

I know your comment was tongue-in-cheek, but I'll say it anyway: wisdom is not the same as intelligence, and though they are probably very smart individuals (most of them), they are terribly unwise.

Foolish, not dumb.

Stallman was right: https://www.gnu.org/philosophy/right-to-read.en.html

This is from 1997, mainly about books but it mentions debuggers and "free kernels" (linux and... hurd?). Considering that they also want to force OSs to require their users' age and some Linux distros won't do it... illegal Linux distros may be in our future.

That doesn't solve the problem: it only disables the duck.ai assistant, but doesn't disable the search results that are normal websites generated with AI (the real issue here). To be fair, that's not DDG problem specifically but every search engine problem.

The web is now full of these fake websites, which is a real problem because on the search results they look legit and only when visiting them you realize it's AI crap.

And the funny (or sad) thing is that current AIs are being trained on these fake hallucinating sites, so even they are suffering from false information, which in turn is given to humans and used to make more fake websites and... the result is up to your imagination.

It's far from useless. Overblown, for sure.

They made an awesome hammer that can solve many hammer-related tasks, but they're selling it like it can also cook, drive and keep the house clean.

Blame the companies, not the math.

This. Pre-2010 social media, which basically showed a feed with the posts of the people you followed, had a lot less negative impact on people. The engagement maximization strategy multiplies "negative" news, because outrage apparently engages more than good news. And current algorithmic feeds like TikTok destroy attention span and make people dumber.

And while we're at it, ban (or at least heavily regulate) dating apps, the only business where the companies sell the opposite product of what they supposedly offer. If people find long-term partners through the apps they would stop using them, so the companies are incentivized to help you not find love and real relationships.

Though they're a mix of sandstorm/thunderstorm (like in the movie), the Mad Max game from Avalanche.