tetronian2

Wiki Contributions

Comments

Sorted by

Is it currently legal to run a for-money prediction market in Canada? I assume the answer is "no," but I was surprisingly unable to find a clear ruling anywhere on the Internet. All I can find is this article which suggests that binary options (which probably includes prediction markets) exist in a legally nebulous state right now.

Yep. I don't think I was ever aware of the name; someone threw this puzzle at me in a job interview a while ago, so I figured I'd post it here for fun.

I guess I didn't formulate the rules clearly enough--if the coin lands on tails, you exit with the stake. For example, if you play and the sequence is HEADS -> HEADS -> TAILS, you exit with $4. The game only ends when tails is flipped.

Suppose someone offers you the chance to play the following game:

You are given an initial stake of $1. A fair coin is flipped. If the result is TAILS, you keep the current stake. If the result is HEADS, the stake doubles and the coin is flipped again, repeating the process.

How much money should you be willing to pay to play this game?

It appears they are going with some kind of modal logic, which also does not appear to deal with incomplete information. I also suspect "moral" will be conflated with "utilitarian" or "utilitarian plus a diff". But then there is this bit in the press release:

Bringsjord’s first step in designing ethically logical robots is translating moral theory into the language of logic and mathematics. A robot, or any machine, can only do tasks that can be expressed mathematically. With help from Rensselaer professor Mei Si, an expert in the computational modeling of emotions, the aim is to capture in “Vulcan” logic such emotions as vengefulness.

...which makes it sound like the utility function/moral framework will be even more ad hoc.

Possibly of local interest: Research on moral reasoning in intelligent agents by the Renssalear AI and Reasoning Lab.

(I come from a machine learning background, and so I am predisposed to look down on the intelligent agents/cognitive modelling folks, but the project description in this press release just seems laughable. And if the goal of the research is to formalize moral reasoning, why the link to robotic/military systems, besides just to snatch up US military grants?)

So, to follow up on this, I'm going to announce the 2015 tournament in early August. Everything will be the same except for the following:

  • Random-length rounds rather than fixed length
  • Single elimination instead of round-robin elimination
  • More tooling (QuickCheck-based test suite to make it easier to test bots, and some other things)

Edit: I am also debating whether to make the number of available simulations per round fixed rather than relying on a timer.

I also played around with a version in which bots could view each other's abstract syntax tree (represented as a GADT), but I figured that writing bots in Haskell was already enough of a trivial inconvenience for people without involving a special DSL, so I dropped that line of experimentation.

Wow, I was not aware of that. I saw that the last one got some minor attention on Hacker News and Reddit, but I didn't think about the outreach angle. This actually gives me a lot of motivation to work on this year's tournament.

Is anyone interested in another iterated prisoner's dilemma tournament? It has been nearly a year since the last one. Suggestions are also welcome.

Load More