4 ms·

This quote, and the kind of thinking that engenders it (and is engendered by it), bother me. Yeah, a cheap laptop can do arithmetic faster than 'the smartest m

This quote, and the kind of thinking that engenders it (and is engendered by it), bother me. Yeah, a cheap laptop can do arithmetic faster than 'the smartest mathematicians'. And the smartest mathematicians can do arithmetic faster than many children. Does that make mathematicians and children 'categorically different things' as well? In some trivial sense, sure, but it doesn't preclude any kind of connection between the two, or require some deep new ontological commitments to model. I'm all for the current practical approach of using 'AI technology' as a human supplemental. But I'd rather not frame it as (what I perceive) as some kind of mystic, dualistic argument. At least not until we know more about both. I also don't really think Peter Thiel is worthy of being the keystone of any kind of argumentum ab auctoritate in this particular field.

yters 8y ago
We can train children to become mathematicians.
- dmreedy 8y ago
  Yes, that's precisely my point. I've read the rest of what you have to say on this, and it is precisely the kind of mysticism that bothers me. There are no means of disproving it, it reduces to dogma in the end, but it's philosophically disingenuous to assert that because computers and people feel like different things in some cases, they must be, and then to argue backwards from there; it's question-begging in the original sense.
  - yters 8y ago
    Seems pretty plausible with a max entropy + Bayesian hypothesis selection. And it has empirical implications: https://www.am-nat.org/site/halting-oracles-as-intelligent-agents/ https://www.am-nat.org/site/halting-oracles-as-intelligent-a... If I'm right, then all AI only companies will be beaten by a AI+human company, so we can make market predictions and propose research directions based on the hypothesis. Doesn't seem entirely without technical/financial merit. I make a note of the economic implications at the bottom of the following proof: https://www.am-nat.org/site/law-of-information-non-growth/ https://www.am-nat.org/site/law-of-information-non-growth/ Why do you say it a form of mysticism?
  - vlovich123 8y ago
    I mean it's clearly obvious that today, "AI" on computers != human. Unsupervised learning is making leaps & bounds. Even supervised learning isn't able to solve all problems better than humans even if it can do so on a staggeringly large amount of problems. I don't think anyone is positing that that's true forever and all time. It seems reasonable to bet on AI at some point becoming sophisticated enough to outperform AI+human. I think it'll happen shortly after the point where AI can identify a new problem (or class of problems) by itself that it hasn't been taught about & then build new tools to help itself tackle that problem. After that it's the singularity because that process repeats ad infinitum. The only value-add of humans after that is if our creativity is somehow better/different than & can explore problems in ways the AI can't (& even that feels like a very short-lived advantage unless there's some crucial physical/mathematical impossibility standing in the way).

yters 8y ago

We can train children to become mathematicians.

dmreedy 8y ago
Yes, that's precisely my point. I've read the rest of what you have to say on this, and it is precisely the kind of mysticism that bothers me. There are no means of disproving it, it reduces to dogma in the end, but it's philosophically disingenuous to assert that because computers and people feel like different things in some cases, they must be, and then to argue backwards from there; it's question-begging in the original sense.
- yters 8y ago
  Seems pretty plausible with a max entropy + Bayesian hypothesis selection. And it has empirical implications: https://www.am-nat.org/site/halting-oracles-as-intelligent-agents/ https://www.am-nat.org/site/halting-oracles-as-intelligent-a... If I'm right, then all AI only companies will be beaten by a AI+human company, so we can make market predictions and propose research directions based on the hypothesis. Doesn't seem entirely without technical/financial merit. I make a note of the economic implications at the bottom of the following proof: https://www.am-nat.org/site/law-of-information-non-growth/ https://www.am-nat.org/site/law-of-information-non-growth/ Why do you say it a form of mysticism?
- vlovich123 8y ago
  I mean it's clearly obvious that today, "AI" on computers != human. Unsupervised learning is making leaps & bounds. Even supervised learning isn't able to solve all problems better than humans even if it can do so on a staggeringly large amount of problems. I don't think anyone is positing that that's true forever and all time. It seems reasonable to bet on AI at some point becoming sophisticated enough to outperform AI+human. I think it'll happen shortly after the point where AI can identify a new problem (or class of problems) by itself that it hasn't been taught about & then build new tools to help itself tackle that problem. After that it's the singularity because that process repeats ad infinitum. The only value-add of humans after that is if our creativity is somehow better/different than & can explore problems in ways the AI can't (& even that feels like a very short-lived advantage unless there's some crucial physical/mathematical impossibility standing in the way).