4 ms·

Including the dataset of lives lived under slavery or cut short by genocide makes me come to radically different conclusions with regard to how prone to over-im

Including the dataset of lives lived under slavery or cut short by genocide makes me come to radically different conclusions with regard to how prone to over-imparting consciousness humans tend to be. Adjusting exploited entities by probability of exploitation and adjusting advantaged entities by probability of advantage also leads me to radically different conclusions. For example, Ray Dalio is one advocate for using AI to supplement decisions. He is also exceedingly rich and a large part of his fortune was made with the help of investment decisions predicated on AI reasoning combined with human reasoning. Even though on average people who try to benefit from AI might not, the average benefit can still be massively positive, because the outcome distribution isn't uniform. Losing means losing a small fraction of a small amount, but winning means gaining an extreme amount to such an extent that it pays for the bad bets. This appears to often happen in practice - see the venture industries economics for evidence of this - and I don't see why it shouldn't apply to AI, which is something present in most unicorns that people often point to when discussing this concept.

tuatoru 4y ago
> how prone to over-imparting consciousness humans tend to be gp was talking about humans' attribution of consciousness to inanimate objects, not other humans. Nearly all religions apart from the more abstract big three would be counterexamples to your example. Dryads and naiads, all kinds of shrines and oracles and talismans...
- cutemonster 4y ago
  Also, two wrongs (slavery etc, and imparting consciousness on inanimate objects and machines) don't make a right
  - JoshCole 4y ago
    Okay. How confident are you that I'm wrong? Would you be willing to construct a wager with me? I ask, for the purpose of the wager, that you generalize the claim that overestimation of intelligence is exploitable. The bet will be as follows: $10,000, 1:1 odds. You will play, without the aid of a computer, against StockFish. My contention is that you will (1) lose the game and (2) lose it despite the fact that StockFish assumes you make better moves than you actually do. Your contention is that you will (1) win the game and (2) do this because StockFish predicted moves which overestimated your actual abilities, which makes it exploitable. I don't think you will be confident enough to put disagreement with my epistemics where your money is. However, to encourage you to do so, I will allow you to instead opt to donate the money to a charity of your choosing rather than giving it to me.
    - cutemonster 4y ago
      But this was about: > humans' attribution of consciousness to inanimate objects and in that context, I think the text you wrote is off topic. But StockFish is good at calculating chess moves yes. And a counting stick is good at counting numbers, is it also alive?
      - JoshCole 4y ago
        Game theory states that exploitability(s) is expected_utility(s, best_response(s)). I think you are perhaps confusing this with whether or not a conditional probability distribution is correctly estimated. Your point would be valid if he said that, but he went much much farther than that. He claimed exploitability and not just that but a troubling amount of exploitability. Within the last month, I've implemented a function which calculates the exploitability of a strategy. I've verified my implementation on known games which are solved, reproducing theoretical results. I'm quite confident that exploitability isn't what you seem to think it is and that in actual practice calculating exploitability with the sort of screening off your are proposing isn't as safe to do as you seem to think it is.
      - JoshCole 4y ago
        In your next post, please exploit me. Extract $1,000 from me using this cognitive weakness you are defending the existence of. Notice that if I gave this same sort of request to a magician, they could fool me because they are talking about what is possible in reality. You aren't. You have to do this, cutemonster. It is desperately important to do this! You have to test your beliefs against reality. If not a bet, then do the trial. The only failure would not be putting your beliefs to the test, because then you stay wrong, but think you are right. Put fallibilism to the test. Do you really know what you think you know? I suspect you don't, but maybe I'm wrong. So humor me. If I'm really wrong, better I found out here, then later. You have a hypothesis about the way you think reality works! Test it! Exploit me with the proposed quasi-magical latent schizophrenic assumption thingy you were defending. I probably definitely have that, sure, you'll get me, don't you think? Easy money, right? Go for it. Shoot your shot.
      - JoshCole 4y ago
- JoshCole 4y ago
  > The latent schizophrenic impulse to impart consciousness to any-old-object in the world is here being exploited by peddlers of technology. He is talking about the concept of exploitability. I'm addressing that context by expanding it to focus on other relevant parts of the classification task - focusing on the false positives screens off information that is critical for estimating the exploitability of a strategy. When you include other things, like true positives and true negatives, you get a very different expected outcome. Notable, you can't not do this and be reasoning correctly. One of the interesting and counterintuitive aspects of decision making under imperfect information which differentiates it with decision making under perfect information is that decision points influence your expected value calculations even when you are not in that situation. For example, if you are investing in a startup A and it is going to fail, that doesn't mean you shouldn't invest in that startup, because the results of investing in startup B, startup C, and startup D all have an impact on whether you should invest in startup A. For good writing on this, see Paul Graham. For great writing on this, read research in game theoretic reasoning. But all of this is pointless to talk about if we don't actually disagree. So... If you really really disagree with me, in my world, the best way for us to settle this is with a bet. I'll bet you $10,000 that you, as a human playing without the aid of software, will lose to a chess playing program of my choosing which has the property of reasoning about your moves with the expectation that you are a stronger player than you actually are. I think the contention that an agent which overestimates others being exploitable is false. A happy accident here is that if I'm wrong to be at all impressed with AI, you shouldn't just be able to win because of the "exploitable" tendency it has to assume its opponent is stronger than it should. You should also be able to be much better than it, because it isn't like AI is better than human in decision making... right? All hype, no substance, nothing to worry about?
- JoshCole 4y ago
  By the way, as someone who cares a lot about epistemics, I can't help but notice you seem to have misunderstood the meaning of the word counterexample. The word counterexample does not mean to counter an example with another example. It means to counter an idea with an example. I'll give an example of this. If someone says -3 is a counterexample to 5 that doesn't make sense. What they would need to say is that -3 is a counterexample to the claim that all numbers are positive. So when you say: > counterexamples to your example That doesn't make any sense to me, for much the same reason that someone telling me -3 is a counterexample to the idea of 5 doesn't make sense to me.

tuatoru 4y ago

> how prone to over-imparting consciousness humans tend to be gp was talking about humans' attribution of consciousness to inanimate objects, not other humans. Nearly all religions apart from the more abstract big three would be counterexamples to your example. Dryads and naiads, all kinds of shrines and oracles and talismans...

cutemonster 4y ago
Also, two wrongs (slavery etc, and imparting consciousness on inanimate objects and machines) don't make a right
- JoshCole 4y ago
  Okay. How confident are you that I'm wrong? Would you be willing to construct a wager with me? I ask, for the purpose of the wager, that you generalize the claim that overestimation of intelligence is exploitable. The bet will be as follows: $10,000, 1:1 odds. You will play, without the aid of a computer, against StockFish. My contention is that you will (1) lose the game and (2) lose it despite the fact that StockFish assumes you make better moves than you actually do. Your contention is that you will (1) win the game and (2) do this because StockFish predicted moves which overestimated your actual abilities, which makes it exploitable. I don't think you will be confident enough to put disagreement with my epistemics where your money is. However, to encourage you to do so, I will allow you to instead opt to donate the money to a charity of your choosing rather than giving it to me.
  - cutemonster 4y ago
    But this was about: > humans' attribution of consciousness to inanimate objects and in that context, I think the text you wrote is off topic. But StockFish is good at calculating chess moves yes. And a counting stick is good at counting numbers, is it also alive?
    - JoshCole 4y ago
      Game theory states that exploitability(s) is expected_utility(s, best_response(s)). I think you are perhaps confusing this with whether or not a conditional probability distribution is correctly estimated. Your point would be valid if he said that, but he went much much farther than that. He claimed exploitability and not just that but a troubling amount of exploitability. Within the last month, I've implemented a function which calculates the exploitability of a strategy. I've verified my implementation on known games which are solved, reproducing theoretical results. I'm quite confident that exploitability isn't what you seem to think it is and that in actual practice calculating exploitability with the sort of screening off your are proposing isn't as safe to do as you seem to think it is.
    - JoshCole 4y ago
      In your next post, please exploit me. Extract $1,000 from me using this cognitive weakness you are defending the existence of. Notice that if I gave this same sort of request to a magician, they could fool me because they are talking about what is possible in reality. You aren't. You have to do this, cutemonster. It is desperately important to do this! You have to test your beliefs against reality. If not a bet, then do the trial. The only failure would not be putting your beliefs to the test, because then you stay wrong, but think you are right. Put fallibilism to the test. Do you really know what you think you know? I suspect you don't, but maybe I'm wrong. So humor me. If I'm really wrong, better I found out here, then later. You have a hypothesis about the way you think reality works! Test it! Exploit me with the proposed quasi-magical latent schizophrenic assumption thingy you were defending. I probably definitely have that, sure, you'll get me, don't you think? Easy money, right? Go for it. Shoot your shot.
    - JoshCole 4y ago
JoshCole 4y ago
> The latent schizophrenic impulse to impart consciousness to any-old-object in the world is here being exploited by peddlers of technology. He is talking about the concept of exploitability. I'm addressing that context by expanding it to focus on other relevant parts of the classification task - focusing on the false positives screens off information that is critical for estimating the exploitability of a strategy. When you include other things, like true positives and true negatives, you get a very different expected outcome. Notable, you can't not do this and be reasoning correctly. One of the interesting and counterintuitive aspects of decision making under imperfect information which differentiates it with decision making under perfect information is that decision points influence your expected value calculations even when you are not in that situation. For example, if you are investing in a startup A and it is going to fail, that doesn't mean you shouldn't invest in that startup, because the results of investing in startup B, startup C, and startup D all have an impact on whether you should invest in startup A. For good writing on this, see Paul Graham. For great writing on this, read research in game theoretic reasoning. But all of this is pointless to talk about if we don't actually disagree. So... If you really really disagree with me, in my world, the best way for us to settle this is with a bet. I'll bet you $10,000 that you, as a human playing without the aid of software, will lose to a chess playing program of my choosing which has the property of reasoning about your moves with the expectation that you are a stronger player than you actually are. I think the contention that an agent which overestimates others being exploitable is false. A happy accident here is that if I'm wrong to be at all impressed with AI, you shouldn't just be able to win because of the "exploitable" tendency it has to assume its opponent is stronger than it should. You should also be able to be much better than it, because it isn't like AI is better than human in decision making... right? All hype, no substance, nothing to worry about?
JoshCole 4y ago
By the way, as someone who cares a lot about epistemics, I can't help but notice you seem to have misunderstood the meaning of the word counterexample. The word counterexample does not mean to counter an example with another example. It means to counter an idea with an example. I'll give an example of this. If someone says -3 is a counterexample to 5 that doesn't make sense. What they would need to say is that -3 is a counterexample to the claim that all numbers are positive. So when you say: > counterexamples to your example That doesn't make any sense to me, for much the same reason that someone telling me -3 is a counterexample to the idea of 5 doesn't make sense to me.