If a fossil of an an alien, or any alien artifact, were put it on display, it would attract millions. Sure some would see it because of its objective importance. But most would come just because it is
Reducing it to a neural network of inputs and outputs is not possible due to nonlinearity, in addition neural networks are not a magic bullet technology where you can train a neural network to do anything, it is difficult enough just to get a neural network to identify cat pictures.
For direct simulationsThe computing power needed to simulate a billion neurons covering every degree of freedom in order to accurately emulate the human brain would be impossible.
It seems Robin is hoping we can emulate humans by reducing the brain to an input-output economy, with neurons instead of industrial sectors. There was much optimism about that approach in the 1960s, but it petered out. I think the problem is interactions are highly nonlinear, so empirical correlation analysis will never get far.
Thus, we need to complete understand human nature, our hierarchy of emotions, desires, affects, and reason. If we figure that out, ems will basically be irrelevant, because we would then know the meaning of life, and all have a clear guide to optimal human flourishing. I'm all for it, but not very optimistic.
Straussian commentary on the world we actually live in. ..
This is what I thought long ago, but I was dissuaded by Robin's futurological earnestness. Can a book be a serious futurological treatise and esotoric commentary on contemporary society? [I don't think the objectives are compatible.]
My confidence in experts is not great. Are sci-fi and futurism experts rewarded mainly for being right, or just being good at telling stories, which involves flattering people's prejudices with the latest scientific buzzwords (Malcom Gladwell)?
In 1950, experts thought Freud and Marx were the most profound thinkers of recent times. Now we think they were mainly just wrong (though, they all have many witty quotes within their ouvre).
How about some experiments, results from consciousness research? I once bought this machine that would read all my receipts and organize them electronically. It took so much time to organize the organization, it was worthless.
Robin, I think I read somewhere in this site, or heard in a podcast, that you expect the numbers of ems running at "steady state" in that future society as extremely high.... Doesn't that situation support a "simulation argument" regarding our own current condition? ie, we should probably be in one of the following disjunct scenarios:
A) we are ems almost for sure, and of a common type
B) we are not ems, and the number of ems to be run in the future will not be astronomically higher than the numbers of "flesh humans"
C) ems may be generated in high numbers, but they are so different to us that we cannot "find us" within that set... Therefore, it is difficult to regard them as human in any sense
I'd like to know your opinion on this... In fact, I wouldn't be surprised if you have already covered this matter somewhere. If so, I will be grateful to have a pointer.
I don't want to be a skunk at a garden party, but help me get over this hurdle. Why should I believes ems are remotely possible? It's hard for me to take interest in the hypothetical if I think the probability of it happening is practically zero.
I've read a lot of sf since I was a kid, so of course sf has influenced me. I've tried to read as much sf on uploads/ems as I could, but most of it was pretty disappointing and didn't have much to add to my book. In a sense a heuristic I've used is to try to look for un-sf forecasts
You need to do more reading on the human brain.
Reducing it to a neural network of inputs and outputs is not possible due to nonlinearity, in addition neural networks are not a magic bullet technology where you can train a neural network to do anything, it is difficult enough just to get a neural network to identify cat pictures.
For direct simulationsThe computing power needed to simulate a billion neurons covering every degree of freedom in order to accurately emulate the human brain would be impossible.
It seems Robin is hoping we can emulate humans by reducing the brain to an input-output economy, with neurons instead of industrial sectors. There was much optimism about that approach in the 1960s, but it petered out. I think the problem is interactions are highly nonlinear, so empirical correlation analysis will never get far.
Thus, we need to complete understand human nature, our hierarchy of emotions, desires, affects, and reason. If we figure that out, ems will basically be irrelevant, because we would then know the meaning of life, and all have a clear guide to optimal human flourishing. I'm all for it, but not very optimistic.
The most succinct discussion of this I've seen by Robin is "Signal Mappers Decouple" ( http://www.overcomingbias.c... )
I'd be interested in what you think of the argument.
Straussian commentary on the world we actually live in. ..
This is what I thought long ago, but I was dissuaded by Robin's futurological earnestness. Can a book be a serious futurological treatise and esotoric commentary on contemporary society? [I don't think the objectives are compatible.]
My confidence in experts is not great. Are sci-fi and futurism experts rewarded mainly for being right, or just being good at telling stories, which involves flattering people's prejudices with the latest scientific buzzwords (Malcom Gladwell)?
In 1950, experts thought Freud and Marx were the most profound thinkers of recent times. Now we think they were mainly just wrong (though, they all have many witty quotes within their ouvre).
How about some experiments, results from consciousness research? I once bought this machine that would read all my receipts and organize them electronically. It took so much time to organize the organization, it was worthless.
Robin, I think I read somewhere in this site, or heard in a podcast, that you expect the numbers of ems running at "steady state" in that future society as extremely high.... Doesn't that situation support a "simulation argument" regarding our own current condition? ie, we should probably be in one of the following disjunct scenarios:
A) we are ems almost for sure, and of a common type
B) we are not ems, and the number of ems to be run in the future will not be astronomically higher than the numbers of "flesh humans"
C) ems may be generated in high numbers, but they are so different to us that we cannot "find us" within that set... Therefore, it is difficult to regard them as human in any sense
I'd like to know your opinion on this... In fact, I wouldn't be surprised if you have already covered this matter somewhere. If so, I will be grateful to have a pointer.
All the best
Ems have been a staple of futurism and science fiction for many decades. The simple argument is that if experts find it plausible, so should you.
I don't want to be a skunk at a garden party, but help me get over this hurdle. Why should I believes ems are remotely possible? It's hard for me to take interest in the hypothetical if I think the probability of it happening is practically zero.
5%?
I expect that there will be one.
I'm curious Robin.
What's your P value for the possibility that human beings cannot be emulated in a computer?
I am looking forward to the book but .... no kindle edition?!
:(
Plans are useless, but planning is indispensable.
I've read a lot of sf since I was a kid, so of course sf has influenced me. I've tried to read as much sf on uploads/ems as I could, but most of it was pretty disappointing and didn't have much to add to my book. In a sense a heuristic I've used is to try to look for un-sf forecasts
Thanks, Robin.
Can these probabilities be supported or are they mere betting odds?