Paradigma11

An interesting explanation I heard is that China had this massive housing crisis that they never allowed to go Boom and clear the system in a few hard years. Instead the problems are lingering with an abysmal local credit market. This is why the local population can't/won't buy much and China either has to produce less or export more. For now they decided to export more to keep the machine going.

The bigger principal problem I have is that I do not understand how the Chinese economy/trade is supposed to work. China is working extremely hard to be completely self sufficient and not import anything. How is trade supposed to work if you do not want anything from your trading partners? What can they sell you besides sovereignty/land...

It will/is be a hard awakening when all those nations start to default/nationalize those projects. That is the point were usually the gunboats show up in the harbor....

In Russia various security services and oligarchs are always trying to win Putins good will. Which is why they will try various low level sabotage actions like paying some losers on signal to firebomb a warehouse or hack into freezers. Remember there is no possible downside for them.

But you only have 2 parties. If you vote for the party that over spends less you show that this is an election winning strategy and will influence governing more.

If LLMs worked that way, you would never have to search for prior art again.

For practical purposes an agent is learning if you put information in its context. So you can create a comprehensive architectural analysis and put it into context and suddenly the agent is familiar with the project. You can also delete information from context, which is a rather hard and imprecise process with humans.

Those integrations are my main reason of not trying it out. You get additional online storage and when your subscription runs out and you are over the limit you wont get any emails (at least that was what happened to me many years ago). What if some automatic chat classification goes wrong and google decides to close your account?

As long as there is no usable customer service and/or ways to limit the blast radius of incidents this is a nogo for me.

Then he would bother and change to chinese models. And OpenAI/Anthropic would lose that profitable subscription which would not really help them at all.

You don't see the tokens served locally by those companies.

"First person to get the wallet wins. If you don’t need the money, consider donating it to those who do: givewell.org."

What risk?

It is very hard to start reasoning what could have happened in those models but I am pretty sure that it was not intentional sabotage.

But are they really compromised? If everybody owns them nobody owns them?

I am pro-AI and am not really participating in the more combative discussions. Why would I fight battles in a war that I am winning anyway and can just sit it out.

But you are mixing up different categories.

Most of your questions are things that are relevant to everybody due to a possible economic crash. And you are right to point this out.

But for me as an AI user those aspects are not really relevant in my decisions. If an AI bubble bursts providers will still want to sell me those tokens at that price and yes development will slow down for a time which does change not a lot for me.

They would most likely be bad sleepers. Here an illustration: https://youtu.be/HBURlcNpfoo

Or buy a cheap Ebike.

I think in many areas we are going to see a renaissance of the "Renaissance man" since AI makes it possible to do so much more as a single person.

What movie physics are you referring to?

If the password database of the website is hacked the attacker knows your password for that very website. With passkeys he knows your public key.

And it is the nation states that fight tooth and nails against more direct elections.

Its ok if the tank is heated up to 60C once per week to avoid Legionella.

Which is why the most important thing is to make sure tokens are a fungible commodity. That is the only thing that counts.

That has not been my experience. Agents love to refactor. They also love to ridiculously overengineer it and run the new version as shadow in the background for several months before adoption.

My experience is that if unchecked very quickly no useful work is being done and every agent is just checking boxes and telling each other what wonderful bureaucrats they are. I am trying to have comprehensive telemetry and checks that limit the overhead to max 10%. Jokes on me since the telemetry turned out the be the most complex item which will finish last. Well, there is always the next project.

Most polls in the EU show around 60% support for military assistance to Ukraine. People are fed up with many things and opposition parties are more likely to be pro Russian to get that vote. If anything the last years have shown Russia's true face and I don't think there is much appetite for renewing those business deals which have been shown to be such a dangerous dependency.

Well, if the policy is to sell its vote membership rights to international adversaries that is to be expected, no?

They lost because they left, they didn't leave because they lost. There is not that much discrepancy between US and Vietnamese scholars about that.

I guess everybody is waiting for the midterms and Trump becoming a lame duck.

MS is actually moving in that direction. They have https://github.com/microsoft/agent-framework and https://learn.microsoft.com/en-us/agent-framework/integratio... which all aim to commoditize llms.

Thanks, will do.

No, but I would be willing to bet against concrete milestones on polymarket/kalshi.