Note that I am not in any way a fan of Kalshi and similar services, and I personally believe they are simply gambling services. Just curious what the law might say for now.

The point overall is that the "prediction markets" don't act as real insurance, it's just a tangential side effect for a minority of participants.

It seems that overall the CFTC and the NYAG are presenting materially different event time lines, so as an outsider it's a bit unclear what is actually happening.

The reason why this is fundamentally quite hard is to do with routing optimization. Routing tables have to reflect physical connectivity - which means packets destined for F2::67 mustn't be routed to LA if that is the IP of a phone that is now resident in SF. However, changing routing tables on all devices along the way as a phone roams would not be an easy feat.

And no, I don't think humans would make the same errors, the key reason for this failure mode is that there are some facts of human day to day life that are so mundane that they are virtually never written about, thus making them an ideal combination of universal human experiences but out of distribution for the LLMs.

The reason for the mistake is irrelevant here. All that matters is that you catch the agent saying something that basically no human would, especially in this type of game.

I think the clear winning strategy is to find some of the common sense reasoning fails that are still quite common, like the "should I walk or drive to the car wash" one from a few months ago.

I feel this completely misunderstands the problem, and the vast gulf between an LLM and a CPU.

First and most importantly, the set of behaviors of a CPU is extremely constrained, and we have a very simple model for which behaviors are safe and which are not. Writing to addresses between X and Y, executing certain instructions - unsafe; everything else, safe. In contrast, an LLM has a huge array of possible behaviors, and variations of those behaviors, and it's very unclear which are safe and which are not. Is emitting the text "sudo rm -rf /" safe? Yes, in some contexts, such as writing this HN comment ; absolutely not in others, such as generating a command that an agent will execute. How do you check which is which? What if it emits "sudo rm -rf /usr/sbin/../.. ", is that safe?

Secondly, CPUs can absolutely be used to hack other people. Nothing in the permission model helps in any way prevent other computers from being attacked by your CPU. So exactly the part we care most about in AI security is the part that has never been solved, for any computing system ever created.

If potatoes suddenly became $1000/kg, everyone would agree that there is a shortage of potatoes. While sure, people won't starve and will eat other things, it doesn't change the basic fact that there aren't enough potatoes.

The situation with RAM is even worse. If I'm trying to build and sell computing devices, I need RAM. Whether no one is willing to sell RAM at all, or whether they'll sell it only for 10x yesteryears's price, it doesn't make much difference to me, I'm still going out of business. Sure, that will then make demand technically go down, but not the shortage.