2min chapter

Minds Almost Meeting cover image

Partiality (Agnes Callard & Robin Hanson)

Minds Almost Meeting

CHAPTER

The Importance of Being Partial to Humans

One question about the AI like I want to be partial to humans over AIs is which of those two positions are they taking. Most of the people I talk to are moral relativists enough to say well it matters to me and I'm not sure if it's globally more morally better or something but they just know that they prefer it. In some in actually in elephant the brain we have a quote from Herodotus how basically everybody thinks that their culture is better than all the other cultures absolutelyBasically he said with an eye toward thinking people have a bias in that regard therefore making you cautious about whether your personal sense that your stuff is just better is really reliably tracks what's actually better.

00:00
Speaker 3
That's kind of funny because I think of gambling as this perfect toy problem to think about risk and uncertainty in the world. And to the extent that it's also kind of a greasy business, I'm purely entertained by it. For instance, I love the contrast between guys who understand math well enough to know when slot machines go positive. And yet they spend their time in smoky casinos, elbowing out tourists to get the juicy plays. But not everyone is going to appreciate that grease aesthetic.
Speaker 2
Although I have to admit, you know, the dean at the time, it took a little convincing to put the craps table in there. And then, and he didn't like the idea of me teaching everybody craps, you know, and then I wanted to keep it there forever. And that was a no-go.
Speaker 3
I've seen a short video of Robert's dice rolling machine. And the best way I can describe it is sort of like a crossbow. The dice go in a slot, then there's a spring, and it shoots the dice out. I'll post a link in the show notes because there are pictures of the machine in one of Robert's papers. Also because it has ways to customize the toss, you could tweak the settings until it does something non-random.
Speaker 4
For initially, we
Speaker 2
tinkered with the machine, then we had the high speed camera come down and that's when we optimized it and said, okay, here's how dice throwers are throwing. Here's the height, here's the speed, here's the spin rate and everything. And we dialed the machine into those specifications. And then we left it at those specifications for every roll, for a ton of rolls. And so we rolled a whole bunch on those sort of ideal settings. And we didn't see any randomness. Then we started, then like you said, probably, you know, in some ways where we should have started was we sort
Speaker 4
of said, can we produce
Speaker 2
non-random events? And so that's when we really started messing with the angle and the speed and the spin rate and all those other factors. And so we tracked, you know, we would do, let's say, two or three hundred rolls on a particular setting and then we'd make an adjustment to the setting and do more rolls. And then we'd consistently address that. So we had exactly what the settings and the adjustments and the distance and everything were and we couldn't repeat
Speaker 3
it.

Get the Snipd
podcast app

Unlock the knowledge in podcasts with the podcast player of the future.
App store bannerPlay store banner

AI-powered
podcast player

Listen to all your favourite podcasts with AI-powered features

Discover
highlights

Listen to the best highlights from the podcasts you love and dive into the full episode

Save any
moment

Hear something you like? Tap your headphones to save it with AI-generated key takeaways

Share
& Export

Send highlights to Twitter, WhatsApp or export them to Notion, Readwise & more

AI-powered
podcast player

Listen to all your favourite podcasts with AI-powered features

Discover
highlights

Listen to the best highlights from the podcasts you love and dive into the full episode