6 ms·

Why any of these "AI experts" don't explore in depth the most obvious solution: use the AI research to upgrade human intelligence, or as Kurzweil names it, achi

by joeyspn 12y ago

Why any of these "AI experts" don't explore in depth the most obvious solution: use the AI research to upgrade human intelligence, or as Kurzweil names it, achieve a "trascendent human race"... Wouldn't be much better (and far less dangerous) to upgrade "ourselves" generation after generation (in stages)? http://en.wikipedia.org/wiki/Transcendent_Man http://en.wikipedia.org/wiki/Transcendent_Man

Jach 12y ago
They do.
ZenoArrow 12y ago
I'd like to ask you a basic question, you might not think there's much depth in it at first... What is intelligence good for? EDIT: To whomever marked me down, just say what you want to say.
- joeyspn 12y ago
  This is actually a tricky one... =) I guess the answer is relative to each person but my best reply would be: "to help species thrive and reach cosmological harmony in an universal context". But then for a different person could be: "for making me earn moar bucks". So it really depends on your values/morals...
ggreer 12y ago
They have explored that solution and found it much harder than creating a safe superintelligence from scratch. To use an analogy: We don't build submarines by upgrading fish, and we don't build passenger jets by upgrading birds. The human brain is some of the worst spaghetti code imaginable. It has no API, no documentation, and no abstractions. Worse, human brains do not have stable values over time. Someone at age 60 won't value the same things they did at age 20, and this is not just due to new knowledge. That's a deal-breaker for a superintelligence. Solving these problems is likely much harder than making AI de novo. Bostrom himself devoted a decent chunk of Superintelligence to biological enhancement. His conclusion was that genetic engineering through iterated embryo selection could get us some IQ 200+ minds. These people would have more cognitive horsepower than any mind in history, including greats like John von Neumann and John Conway, but they would not be superintelligent. Also, Bostrom concluded that somatic gene therapy and brain-computer interfaces were unlikely to help. And as a practical matter, humans take 20+ years to reach maturity. Many forecasts put artificial superintelligence only a generation or two away, at which point all biological systems would be superseded. Biological enhancement may help to create better AI researchers, but they're not the end-game.
- joeyspn 12y ago
  Well, we're just getting started to reverse-engineer the human brain. I wouldn't rule out the option of brain-computer interfaces yet. It's certainly worth exploring. The APIs could be known in a decade or less... The EU "human brain project" [0] is focused on that task. Now we can also hack our genome much more easily with the latest breakthroughs in gene editing (CRISPR), so we are slowly unblocking the necessary tools to start building "superhuman upgrades"... I'm not saying that is easier, just that it would be much better for us as species. I see this as the same dilemma with nuclear power. We know that fusion is better than fission for us in the long run, althought it is orders of magnitude more difficult to get... Meanwhile, fission is really convenient (easy and cheap)... Hopefully we won't have any "nuclear (fission) crisis" that annihilates us before we get to master fusion in the coming years. Same applies to AI vs superhumans... [0] http://en.wikipedia.org/wiki/Human_Brain_Project http://en.wikipedia.org/wiki/Human_Brain_Project https://www.humanbrainproject.eu/ https://www.humanbrainproject.eu/
pgodzin 12y ago
I think a lot of that will come in advances in things like medicine, where nanotechnology will be able to assess the health of our bodies from within and repair it using some kind of AI. Changing brain functionality seems like a step too far when we are nowhere close of understanding how our brains work.

by joeyspn 12y ago

Jach 12y ago

They do.

ZenoArrow 12y ago

I'd like to ask you a basic question, you might not think there's much depth in it at first... What is intelligence good for? EDIT: To whomever marked me down, just say what you want to say.

joeyspn 12y ago
This is actually a tricky one... =) I guess the answer is relative to each person but my best reply would be: "to help species thrive and reach cosmological harmony in an universal context". But then for a different person could be: "for making me earn moar bucks". So it really depends on your values/morals...

ggreer 12y ago

They have explored that solution and found it much harder than creating a safe superintelligence from scratch. To use an analogy: We don't build submarines by upgrading fish, and we don't build passenger jets by upgrading birds. The human brain is some of the worst spaghetti code imaginable. It has no API, no documentation, and no abstractions. Worse, human brains do not have stable values over time. Someone at age 60 won't value the same things they did at age 20, and this is not just due to new knowledge. That's a deal-breaker for a superintelligence. Solving these problems is likely much harder than making AI de novo. Bostrom himself devoted a decent chunk of Superintelligence to biological enhancement. His conclusion was that genetic engineering through iterated embryo selection could get us some IQ 200+ minds. These people would have more cognitive horsepower than any mind in history, including greats like John von Neumann and John Conway, but they would not be superintelligent. Also, Bostrom concluded that somatic gene therapy and brain-computer interfaces were unlikely to help. And as a practical matter, humans take 20+ years to reach maturity. Many forecasts put artificial superintelligence only a generation or two away, at which point all biological systems would be superseded. Biological enhancement may help to create better AI researchers, but they're not the end-game.

joeyspn 12y ago
Well, we're just getting started to reverse-engineer the human brain. I wouldn't rule out the option of brain-computer interfaces yet. It's certainly worth exploring. The APIs could be known in a decade or less... The EU "human brain project" [0] is focused on that task. Now we can also hack our genome much more easily with the latest breakthroughs in gene editing (CRISPR), so we are slowly unblocking the necessary tools to start building "superhuman upgrades"... I'm not saying that is easier, just that it would be much better for us as species. I see this as the same dilemma with nuclear power. We know that fusion is better than fission for us in the long run, althought it is orders of magnitude more difficult to get... Meanwhile, fission is really convenient (easy and cheap)... Hopefully we won't have any "nuclear (fission) crisis" that annihilates us before we get to master fusion in the coming years. Same applies to AI vs superhumans... [0] http://en.wikipedia.org/wiki/Human_Brain_Project http://en.wikipedia.org/wiki/Human_Brain_Project https://www.humanbrainproject.eu/ https://www.humanbrainproject.eu/

pgodzin 12y ago

I think a lot of that will come in advances in things like medicine, where nanotechnology will be able to assess the health of our bodies from within and repair it using some kind of AI. Changing brain functionality seems like a step too far when we are nowhere close of understanding how our brains work.