FTX made us write a good paper
Some of my scientific papers come about in curious ways, and while I don’t have half a Gazillion citations to my record, this makes my papers more entertaining than the usual products of publish-or-perish mass output, especially if you know their back stories. This is the story of how my own psycho-nautical experience led me to publish on the large-scale connectivity of the human brain:
The latest odd path to a – I think! – quite insightful scholarly publication connected my co author and me to the biggest crypto-scandals, ever. Fortunately the connection is really indirect, but there it goes: As most observers of international news will know, a crypto-exchange called FTX went from a major, major financial institutions dealing with billions of Murican Dollars, and rubbing shoulders/paying to promote major celebrities. People who didn’t do anything useful for the real economy (the crypto folks) paid soccer players (also useless for the common good) to promote their scam-fraud. The head honchos of FTX “borrowed” money from their customers to run all of this. The whole construct crashed, and here is one account of the trail which followed:
FTX founder Bankman-Fried convicted of defrauding cryptocurrency clients
Now, before these fraudsters were caught and they became ex-billionaires they set up a contest. The contest was to show the possibility/impossibility of artificial general intelligence, a version of AI which isn’t just good at a narrow sector of writing summaries/image recognition/.. but at everything which a human can do. And then, an argument should also be made why the artificial general intelligence (AGI) or even super-human intelligence will be dangerous, or not. The best argument would win a good chunk of cash, several 10k US$, as far as I remember.
I found out about this compost via some post on X (then Twitter), leading me to some effective altruism site. Effective altruism, the idea to be as rational/effective with one’s charity, not emotion driven, was one of the FTX founder Bankman-Fried’s main interests. Or, main ways of making himself look good and not greedy. On the effective altruism site the compost mechanics were laid out. It seemed to be serious.
My friend & colleague & founder of the Neurolinx Institute Jay Coggan and I took the opportunity to write up some thoughts about the levels of organization of brains, and of energy use in brains and human-made digital computers which we had pondered & discussed for a while. The prize money would fund (yeah!) our future work on this topic. A cold, hard look at the issues surrounding AI and AGI is oddly missing, in our opinion. With the FTX cash we could dedicate a few months of our time to writing and thinking about all of this.
Just when we had our ideas written up, and the document polished, did FTX crash and the compost disappeared, with all the nice prize money. Shame. Still, Jay and I didn’t let the sudden absence of a lot of cash deter us (we are academics after all! Money is undignified to us!) and nevertheless wrote up our ideas in a paper.
Bankman-Fried will sit in prison somewhere, and given the increasing number of fraudulent startup founders, they might as well start a all-startup-founders prison gang. Maybe they can come up with an app which warns you if you are about to get stabbed in the prison yard.
Jay and I still ponder how we can get funding for researching the foundations of computation & intelligence. We ponder this while gazing at Swiss alpine lakes and the tropical Pacific Ocean, not at prison walls, which is infinitely better, prize money or not. Still, to whom this might concern: fund us!
