• 0 Posts
  • 76 Comments
Joined 16 days ago
cake
Cake day: June 6th, 2026

help-circle


  • If you are talking about GPUs being only a small share of the overall total sum, bad news that the supporting infrastructure is also to a large extend tailor made for that very narrow use case. No one else will need such huge data center facilities designed specifically for GPUs, that includes also the non GPU components. And the infrastructure is the only thing of substance of this bubble. The models aren’t it. Open weight models are on the heels of the closed models. As soon as they are good enough for common applications, the business case for charging billions is slowly evaporating.

    You are also mistaken, I am not worried about GPUs. I am merely stating that they and their server infrastructure (which is tailor made for them) are rapidly getting obsolete equipment by their nature and while the clock is ticking they are largely not even being used. This is fundamentally different from the dotcom and railway bubble.


  • Both rail and communication infrastructure lead to some useless connections but much of it was no useless, in both cases. GPUs are not bolted to the ground but they do become obsolete no matter if you deny it or not. The issue is that the real costs is in using GPUs is very different from these previous bubbles. Those obsolete GPUs will cause much higher operating costs than newer generations, to the point where they won’t be interesting to use even if you gave them away for free. To make matters worse, other infrastructure is much more flexible in its use, one can transport all sorts of things on railways, one can send all sorts of data on communication infrastructure. Those specialised GPUs aren’t very useful for anything other than a fairly narrow use case.

    I think you do not fully appreciate the crazy amounts of GPUs we are talking about here. China has no massive real shortage of GPUs. They managed to get black market GPUs more or less directly from Nvidia just fine. Nor are European universities IT wastelands without compute capabilities. But even if they’d go crazy on expandig compute infrastructure with outdated power hungry GPUs, that would be barely more than a drop in the ocean. Nvidia does have to resort to circular financing to keep the boom cycle accelerating, with GPUs going just to some storage facility if they exist at all. That is not how healthy demand looks like.










  • I am talking about Austria because to compare it with other parliamentary democracies it helps to chose one concrete example, you can chose another one if you like. How about Germany, the largest member state. There Parliament’s position in this regard is actually weaker than in Austria.

    I have no idea where you are coming from but you seem to lack knowledge how parliamentary democracies work if you hold the completely outlandish view that they are on the same level as the Chinese system in terms of democracy.

    Back to the EU Commission. Its election is obviously a system where both, the Council / member states and the EP hold power. (“election” is the word in the treaties btw) This is by design. Power is not centralised. It is common in parliamentary democracies that parliaments elect/consent on members of the government but don’t choose them. However government with members that are not to the liking of a majority in Parliament won’t be elected/voted into power. The same is the case in the EU and there is precedent for that as well. The vote on VdL yielded a paper thin majority im the EP and only because VdL was giving the EP concessions in return. If the EP targets candidates as not acceptable they will not make it into the Commission. Again, there is precedent for that.

    If that sounds like Chinese “democracy” to you, half the democracies (ie all parliamentary democracies) on earth are in reality a Chinese style “democracy”. Seriously?


  • Mistral has recently shown a good trajectory of improvement. It is already an important thing that there is a European mid range open weight model that can compete. (Frontier models need a lot more resources, it is important to compare apples with apples) This is good enough for many applications were data security and sovereignity are prime concerns. Of course, it would be good to have a frontier model, lets see how Large 4 will perform when we get there.


  • The EU commission is elected by the directly elected European Parliament based on suggestions by the Council/member states. The Commission can be voted out of office anytime when it loses support in the European Parliament.

    The Austrian government is elected by Parliament based on suggestions by the Chancelor candidate (the latter chosen by the President). The parliament can vote the government out of office anytime.

    According to you the one thing is utterly undemocratic while the other is not. ok.

    The EU Commission is not the EU, but it is its executive and administration. I f you just kill that, you let everthing derail. Bureaucracy is a dirty word but there without it political entities implode.




  • “Materially speaking is more expensive to send a letter next town than a packet from something like aliexpress.”

    That is wrong, in many cases quite obviously. Microdeliveries are commonly sent as letters within Europe. So they are literally a letter(coming commonly from another European country) + a consolidated flight freight from the other side of the globe. The last leg alone creates more costs than the entire product plus shipment is purchased for.

    Sorry, but if you think this can be done for 0-1 EUR (the latter if we assume the 1 EUR product is worth exactly 0 EUR) I can’t help you.

    Of course this change will incentivise larger but fewer orders. If the platforms would care about that, 1 EUR products with free shipping wouldn’t even exist. They aren’t stupid or incompetent. If there is economic incentive for that, they’ll do it. Removing the advantage of <150 EUR orders, removes the incentives for smaller more frequent orders. This will do a lot to remove a lot of stress from logistic infrastructure, even if total amount of stuff bought in China remains the same. That’s the point. That and systematic mislabeling of shipments that lose their incentive to some extend as well.