I’ll be speaking twice this January in New York City on my upcoming book The Age of Em: January 6, 7pm, at Brooklyn Futurist Meetup, Geraldo’s Cafe in Brooklyn Law’s Feil Hall (1st Floor), 205 State Street, Brooklyn, $5.50 fee. (Video
I attended the talk in New York (the junto club). You were treated so incredibly rudely I could not stand it. I left at about 8:45 (after you had only even been up there for 15 minutes!), after I couldn't stand that host interrupting you anymore. I'm very sorry you had that experience. I hope the rest of the talk went better.
The em is a copy anyway: barring some unknown weird quirk of fundamental physics (basically it would require something like god transferring your consciousness to a copy once your brain dies) you won't be able to experience the em's life. The fact that you can create multiple ems from a single mental image adds to this point.
The inability to know whether an artificial intelligence (ems aren't normal AI, but they are housed in artificial computers) will be conscious is a universal problem, you can't measure consciousness and quite possibly we will never be able to measure it. Many people might only stop doubting the consciousness of a machine if that machine were made of biological neurons (even though that too is no guarantee the machine will be conscious).
So, ems will be intelligent, but not necessarily sentient? Doesn't that throw a (possible) wrinkle into the vision of em society? If you get downloaded, would it not matter to you if (with destructive imaging techniques) you would be turned a non-sentient thing?
My 2 cents: it does not depend on the computational theory of mind. You will get a functioning intelligence no matter how the fundamental physics behind minds works, maybe not with silicon transistors, but certainly with some kind of computer. Whether that intelligence will be conscious, or whether consciousness is a guaranteed byproduct of intelligence, or whether consciousness is some grand illusion somehow, are things we can probably never know for sure anyway (just as we can't be sure all humans are conscious, we just sort of assume that if evolution made one human conscious it made all of them conscious, or some of us assume it's all an illusion).
Maybe I should wait for the book, but I wonder if anyone has insight into this question: does the possibility of uploading minds (as with ems) depend on the computational theory of mind? [Put differently (perhaps unnecesarrily controversially): if Searle is correct, could minds be uploaded to digital computers?]
I'm pretty sure you are missing an important step in your idea of robots ruling the world and EM.That step is non-self-conscious AIs that can, nevertheless, emulate the predictive power of humans, but with greater accuracy in many areas.
Consider IBM's Watson, on health -- first on cancer. Already today Watson is as good as a top doctor with predicting what is the disease, and often better at suggesting further tests, plus equal to prescribing the best known medicine to take. For that individual.
So let's say within 10-20 years, all hospitals replace most of their doctors with Watsons. The administrators and nurses rule, not the robots. The robot (/slaves) do all the work, have all the knowledge available, and even are constantly recommending how to reduce cost, increase service -- but humans rule. Humans make the decisions.
And once robot slaves do all the work, why would humans choose to stop making the decisions? Then the human competition becomes who can get appointed to these top decision making jobs.
I attended the talk in New York (the junto club). You were treated so incredibly rudely I could not stand it. I left at about 8:45 (after you had only even been up there for 15 minutes!), after I couldn't stand that host interrupting you anymore. I'm very sorry you had that experience. I hope the rest of the talk went better.
I've added a link to video of the first talk to the post.
The em is a copy anyway: barring some unknown weird quirk of fundamental physics (basically it would require something like god transferring your consciousness to a copy once your brain dies) you won't be able to experience the em's life. The fact that you can create multiple ems from a single mental image adds to this point.
The inability to know whether an artificial intelligence (ems aren't normal AI, but they are housed in artificial computers) will be conscious is a universal problem, you can't measure consciousness and quite possibly we will never be able to measure it. Many people might only stop doubting the consciousness of a machine if that machine were made of biological neurons (even though that too is no guarantee the machine will be conscious).
So, ems will be intelligent, but not necessarily sentient? Doesn't that throw a (possible) wrinkle into the vision of em society? If you get downloaded, would it not matter to you if (with destructive imaging techniques) you would be turned a non-sentient thing?
My 2 cents: it does not depend on the computational theory of mind. You will get a functioning intelligence no matter how the fundamental physics behind minds works, maybe not with silicon transistors, but certainly with some kind of computer. Whether that intelligence will be conscious, or whether consciousness is a guaranteed byproduct of intelligence, or whether consciousness is some grand illusion somehow, are things we can probably never know for sure anyway (just as we can't be sure all humans are conscious, we just sort of assume that if evolution made one human conscious it made all of them conscious, or some of us assume it's all an illusion).
Maybe I should wait for the book, but I wonder if anyone has insight into this question: does the possibility of uploading minds (as with ems) depend on the computational theory of mind? [Put differently (perhaps unnecesarrily controversially): if Searle is correct, could minds be uploaded to digital computers?]
Hanson's claim is that the technology for uploading human minds will be available long before entirely engineered intelligence.
I'm pretty sure you are missing an important step in your idea of robots ruling the world and EM.That step is non-self-conscious AIs that can, nevertheless, emulate the predictive power of humans, but with greater accuracy in many areas.
Consider IBM's Watson, on health -- first on cancer. Already today Watson is as good as a top doctor with predicting what is the disease, and often better at suggesting further tests, plus equal to prescribing the best known medicine to take. For that individual.
So let's say within 10-20 years, all hospitals replace most of their doctors with Watsons. The administrators and nurses rule, not the robots. The robot (/slaves) do all the work, have all the knowledge available, and even are constantly recommending how to reduce cost, increase service -- but humans rule. Humans make the decisions.
And once robot slaves do all the work, why would humans choose to stop making the decisions? Then the human competition becomes who can get appointed to these top decision making jobs.