well I see more problematic the people actually doing the Iran attacks and murder of heads of state. Betting on those is distasteful, but doing those things is where the damage lies.
the entire point of the argument is that they're the same people. Military bets appear to have significantly higher rates of insider trading than baseline[1], which implies two things, both catastrophic. One is that the markets leak classified information (which is the entire point of the market and it should be a national security no brainer to close it for that reason alone) but the even worse scenario is causality in the other direction, that a bet leads someone to take a military decision.
> One is that the markets leak classified information (which is the entire point of the market and it should be a national security no brainer to close it for that reason alone)
It's not at all obvious that leaks of classified information are per se detrimental to _actual_ national security.
It's not because specifically with these markets there is an amplification effect where the one doing the bet creates incentives for what it's betting in favor or against to materialize in the world.
In other words the money spent on bets that involve killing directly foments more killing.
That they are virtually always replaced with other mass-murdering dictators, usually after a period of even worse mass murder known as a civil war. All at the low low cost of some billions of dollars taken away from useful work.
Ali Khamenei was recently murdered, at a price tag of several tens of billions of dollars, so far. His more radical son now occupies his former position. As a result, is the world a better place?
> So if instead of text we come up with a different representation for mathematical or physical problems, that could both improve
But then, wouldn't we first have to translate all of our current math and physics knowledge into that new representation in order to be able to train a model on it? Looks like a tremendous amount of work to me.
Yes, but by then you already have general LLMs capable of helping with the work. And even if you didn't, if that's what it would take to advance research in these fields, that would be a justifiable effort.
I have an e-ink photo frame on the wall that switches picture once every 24h, picking one of my pictures of the last 10+ years by random. So every single one of my tens of thousands of pictures gets a real chance to be seen at least once during my lifetime :)
But transistor designers didn't pivot away from designing transistors. They left Facebook and all the other stuff to others and kept designing better transistors.
Analogy doesn't work because the bomb came first and the reactor later, that is, regarding fission the reactor isn't even here yet. And it was clear from the beginning that the chain reaction is real not hypothetical.
No, there was a functioning reactor built as part of making the bomb.
And it is not a hypothetical concern that a new type of agent might be fundamentally uncontrollable; we're already barely surviving corporations persuing their inhuman goals.
