I hesitate to post this here but I'm not sure Stripe hasn't become most of those things as well now. It's obviously been very successful and I'm happy for the people who built such a useful service in the early years. But I also mourn the loss of the tidy product, transparent pricing, clear documentation, and legendary support that made it exceptional back then.
Why not? The point of regulation is to fix things when market forces haven't been sufficient. You keep talking about winners and losers but it's unclear to me who you think is actually winning in this situation.
and what market forces haven’t been sufficient here. Demand went up, price has gone up. Price has been fairly stable since the initial run up. It’s unclear to me why you think who the winners or losers matters. My position is it’s incredibly difficult to regulate short term supply and demand. You will always introduce unintended consequence. Regulation in my opinion works best where there are external costs that are hard to measure. This is not one of them.
Market forces have not been sufficient because the funny money moving around in the AI sector is starting to significantly damage other parts of the economy. I'm not sure how I can say this any more clearly.
If the AI sector wants to take a big risk that either pays off or it doesn't then that's one thing. If the danger is contained and the people taking the risks are the only ones who will suffer if things don't work out then that's their choice. The important point is that the danger in this case is not contained. It's not just the AI sector being affected now. It's not just the people choosing to take the risk in the hope of a big profit who are being harmed.
I don't think it's realistic any longer to treat this as some sort of short term supply-and-demand problem with luxury products. It's been building up for several years already. We're seeing public statements from some of the big companies involved that already run through 2026 and beyond that imply the supply chain situation continuing to worsen for everyone else. And the components in question are now necessary for a lot of normal operations and day-to-day life.
If we just wait and see for another 2-3 years as some are suggesting then in the meantime real businesses who were operating responsibly and doing nothing wrong will be failing to grow as they should, putting up prices for their customers, letting staff go, or even failing completely. Governments will be redirecting tax revenues to basic IT infrastructure instead of public services. People who just wanted to buy a normal device for their own personal use won't be able to afford one or maybe to find one at all and their quality of life will fall.
I can see no reason why we should allow these harmful outcomes just because some already rich VCs and some AI tech bros are playing games at the scale of the global economy to try to sustain the unrealistic valuations of the big tech companies for their own benefit. It's increasingly unlikely to work anyway and it's a realistic possibility that the bubble will burst this year so trying to contain the fallout from that as much as possible instead of allowing the bubble to inflate even further is not a bad idea either.
The "scale might be different" matters quite a lot in this case. We're not just talking about demand slightly outpacing supply and resulting in prices going up 10%. We're talking about large parts of our societies and economies no longer having reliable access to technology that we now depend on for normal operation.
We wouldn't allow any amount of investor money to buy out essential utilities and then exploit their natural monopolies to charge the public 10x as much for access to water or electricity.
We wouldn't allow any amount of investor money to buy out all the companies that maintain our roads or rail networks and then charge 10x the established prices for maintaining that infrastructure.
We wouldn't allow any amount of investor money to buy out all the phone networks and then deny people access to communication because they didn't pay some exorbitant protection fee.
No-one thinks regulation of these markets and interference with these kinds of corporate transactions is a crazy idea. Why do so many people here apparently think we should let the funny money funded AI giants distort the entire global tech supply chain in the hope that their silly valuations won't come crashing down for a bit longer?
We can't afford to wait "several years" to see whether the invisible hand will fix the problem. The markets have already allowed this situation to develop over a period of several years. The damage is too severe and it's happening right now and it's getting worse. Governments need to start swinging the regulatory axe now.
What we're seeing is the natural conclusion of VC distortion in a market. There is so much money being pumped into AI speculatively now that it's hurting normal and sustainable businesses in other parts of the economy.
The solution might have to be mandatory rationing of some kind to avoid a situation where only a handful of AI giants are able to buy essential components. We can't just throw the rest of the economy under a bus to support the AI bubble for a few more months.
I'm working with a business right now that would like to buy some new servers for sensible, boring business reasons. It is having trouble because the prices from their normal suppliers are now extremely high - if the components are even available at all. This business has nothing to do with AI or Big Tech and yet it's at risk of being unable to continue normal operations in much the same way that a business would be affected if the phone networks were all switched off or the water supply to its office was cut. We regulate those industries because their continued reasonable operation is essential to make sure everyone else can continue to operate reasonably as well.
I'm seeing the same thing. I was consulting a group of people in my city that wanted to digitize massive load of old VHS tapes. No AI, no crazy tech, just standard, boring storage+network infrastructure.
I'm looking at the procurement sheet that I made for them a year ago. Half of the items are no longer available, while the other half became so expensive that we'd probably build 10 of such labs with these costs a year ago.
I'm also looking at my home NAS right now - I pray not even a plastic clip breaks inside, because I'd have to shut it down.
While these are still likely the first things that you'd think of being affected, I'm sure the effects are rippling through essentially every industry that utilizes these components in their supply chain. Which is probably - every industry nowadays?
I wish this comment can be on the absolute top of this page. This really is one of my frustrations with the AI bubble.
Fwiw, the days of creating an good ol' reliable hosting provider/Vps provider are over. I looked extensively into it one time out of curiosity but this would be one of the worst times in history to do that.
We would be sort of stuck with the options that we have right now and more and more shops in Lowend are even shutting down or raising prices with the sheer ram crisis and even HDD and storage crisis now.
A provider in LET had a post which said, "what should we providers do to deal with the ram shortage/ram prices"
These providers gave competition/had different unique features too to have chosen them but they were also incredibly price sensitive and the AI bubble blew the sensitivity by raising the prices almost 5 times or more. This would impact real businesses.
Thank you for creating this comment. I hope more people can read this. I genuinely just want this bubble to burst asap so that we can see a sense of rationality back within the market/the market functioning as expected without the immense irrationality/unpredictability of future.
another point is this, from my hosting provider idea, I shut it down. Why? because it literally makes 0 sense to start now, its postponed indefinitely untill the bubble bursts/ram prices are decreased.
How many other projects might be going through something similar. Gck1's comment next to mine also gives an example of a project whose value of cost increased 10 times.
How many of such projects would simply be unable to be built because of the ram inflation can't be underestimated imo.
and forget people who wish to game and many other things too. Basic comodities in the previous year or two feel like luxury now. All because of AI. It's insane.
I think that’s a massive stretch. What we are seeing is a new frontier in tech that nobody knows where it will land yet. Hyperscalers see a future where if they don’t build now that they might be left behind.
Absolutely VC money is flowing around but I think it’s unclear where the cards fall yet.
Not sure what you would regulate here. I hate the tripe that America and China are at war but I do think it’s not a great decision to stop the current work the west is doing as China is pushing full steam ahead.
It's not much of a stretch at all. There are already normal businesses that can't buy normal equipment at normal prices (or can't buy it at all) right now because the supply chain has been redirected to a small number of businesses that can only afford to drain the pipeline like that because of the astronomical scale of speculative investment they've received. Similarly there already individuals who can't buy normal equipment for their own use.
This situation is harmful both economically and for basic quality of life. It is rational - and probably now necessary - for governments to intervene to counter the market distortion and ensure the continued availability of normal products to everyone else.
I am fully aware that the West regulating here would potentially undermine the VC investment model that these big tech firms are relying on. I have no problem with this. Business entities are legal fictions that we allow to exist for the benefit of real people. If the behaviour of those entities is harmful to real people - and I don't think anyone can credibly claim otherwise in this case - then it's time to change the rules they operate under.
If you watch the popular streaming services on a Linux PC then you'll be lucky to get more than 720p and some only seem to provide SD anyway - even for content from the HD era. This despite obviously having access to more detailed versions that they'll serve you for the same money if you watch on one of their preferred platforms.
IME this becomes quite jarring if you're watching on a modern high-res screen - even something like a quarter-sized browser window or PIP on a 4K monitor - with other much sharper content visible nearby. If you're watching either full-screen or in a smaller window but without other more detailed visuals around it then personally I find it less jarring and I can still become immersed if the material itself is good. It's still noticeable if you're used to watching HD or 4K content on other devices but I'd rather watch something good in SD than not watch it. The same goes for old shows that were only made in SD originally.
I am using a projector. Definitely not as sharp as a 4K and higher TV but it makes for a huge screen since I am covering almost all the wall. I find the result smooth and enjoyable regardless of the input resolution.
In other words, immersion seems to be more important than precision/resolution.
If you are watching in in a window / PIP then clearly you don't actually care that much about the looks since you are not even devoting your full attention to them.
Social media, which everyone here is complaining about, is by far the most open and democratic form of mass media that has ever existed.
It would be if it were actually social - if the messages people saw were written by authors those people were interested in because of some kind of social relationship. But of course that's not really the case.
One problem here IMHO is that the meaning of terms like "press" and "media" has shifted significantly with modern Internet trends. Freedom of the press used to be an extension of freedom of speech. The principle was essentially the same but it acknowledged that some speech is organised and published to a wider audience. Neither has ever enjoyed absolute protection in law anywhere that I'm aware of because obviously they can come into conflict with other rights and freedoms we also think are important. But they have been traditionally regarded as the norm in Western society - something to be protected and not to be interfered with lightly.
But with freedom must come responsibility. The traditional press has always had the tabloids and the broadsheets or some similar distinction between highbrow and lowbrow content. But for the most part even the tabloids respected certain standards. What you published might be your spin but you honestly believed the facts in your piece were essentially true. If you made a mistake then you also published a retraction. If someone said they were speaking off the record then you didn't reveal the identity of your source. You didn't disclose things that were prohibited by a court order to protect someone involved in a trial from prejudice or from the trial itself collapsing. Sometimes the press crossed a line and sometimes it paid a very heavy price for it but mostly these "rules" were followed.
In the modern world of social media there are individuals with much larger audiences than any newspaper still in print but who don't necessarily respect those traditional standards at all and who can cause serious harm as a direct result. I don't see why there is any ethical or legal argument for giving them the same latitude that has been given to traditional media if they aren't keeping up their side of the traditional bargain in return. We have long had laws in areas like defamation and national security that do limit the freedom to say unfair or harmful things. Maybe it's time we applied the same standards to wilful misinformation where someone with a large audience makes claims that are clearly and objectively false that then lead to serious harm.
If anyone wants an honest answer to that question it is fairly simple. Polling has suggested - very consistently and over a long period of time - that a majority of the British public (though often a fairly slim majority) tend to support authoritarian interventions by our governments in the name of protecting the public. Most of the time our governments and government agencies do appear to use such powers responsibly and so they tend to maintain that public trust. There has always been a significant minority who were more cautious on civil liberties grounds and there has always been an issue that the supportive majority aren't always very well informed about what could happen if the laws were applied more strongly in practice.
As a personal observation - I think this might start to change over the next few years and the current positions of MPs and government might start to look very out of touch. We are seeing the fall of our long-standing "big" political parties and the rise of a very right wing populist party that is increasingly looking like it might actually win significant power at the next general election. I think awareness of the potential for abuse by the next people to run the government and agencies is growing among the general public. Whether it grows enough to stop some of these policies from becoming law in the near future is a different question of course.
I still think it was a mistake for Firefox to dump its old plugin model. The customisation was a USP for Firefox and many useful tweaks and minor features have never been replaced.
Today the ability to run proper content blockers is still a selling point for Firefox but obviously wouldn't be if they started to meddle with that as well. (Has there ever been a more obvious case of anticompetitive behaviour than the biggest browser nerfing ad blocking because it's owned by one of the biggest ad companies?)
Other than customisation the only real advantage I see for Firefox today is the privacy angle. But again that would obviously be compromised if they started breaking tools like content blockers that help to provide that protection.
It doesn't require consent for cookies or similar data that are strictly necessary to do what the user has asked for - a token for logging in or the contents of a shopping cart are the two canonical examples.
It certainly does require informed consent in other situations though and the dreaded cookie banners were the industry's attempt to interpret that legal requirement.
No, it's now entirely accurate. Nothing in the GDPR requires 'cookie banners', and your Wikipedia link doesn't 'dispell' that 'misconception', but nice try...
My point is that it was never the GDPR that required any sort of "cookie banner" in the first place.
The cookie banner requirement is itself a widespread misconception because the actual rule is neither specific to cookies (it would also cover other locally stored data) nor universal (for example it doesn't require specific consent for locally storing necessary data like session/login mechanics or the contents of a shopping basket).
The requirements for consent that do exist originate in the ePrivacy Directive. That directive was supposed to be superseded by a later ePrivacy Regulation that would have been lex specialis to the GDPR - possibly the only actual link between any of the EU law around cookies and the GDPR - but in the end that regulation was never passed and it was formally abandoned earlier this year.
So for now rules about user consent for local data storage in the EU - and largely still in the UK - do exist but they derive from the ePrivacy Directive and they are widely misunderstood. And while there has been a lot of talk about changes to EU law that might improve the situation with the banners so far talk is all it has been.
reply