Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
Reminded me of a blog post I read this morning by Peter Diamandis about self driving cars, check out how much he thinks they will drop in price compared to classic ones (hint: they are way cheaper):
I know 2030 seems like a optimistic date but considering that information technology advances exponentially its really not. Why the pace of advancement today is so much faster then it was just 10 years ago and why it will be even faster in the next 10.
Some people can not see the impact of computers advancing exponentially no matter how much it is explained to them. When 2030 comes you will finally get it.
Raw processing power isn't sufficient to do the tasks you want. You need mechanisms to control the system under discussion. Additionally, those mechanisms must be non-destructive.
Prove that a million billion C64:s won't have the processing power needed... For the rest, it is unclear what you are talking about.
Quote:
Originally Posted by ncole1
For some tasks such as drug delivery, this may be fairly straightforward; however, for others, this is not so easily so.
Like tapdancing, yes most likely.
Quote:
Originally Posted by ncole1
One I'd be particularly curious about is metabolic pathways. I'm pretty sure it would be fairly difficult to alter them in a sufficiently controlled way to do what one wants, without undesired side effects.
Who is talking about metabolic pathways? Not me.
The base idea of Dr De Grey (for example) is to do away with research about it completely and repair cells, not tinker with metabolism.
Quote:
Originally Posted by ncole1
That's a lot of disparate goals there! (Nanobot migration, chemical processes, neurology).
Are you assuming the nanobots would be programmable?
What??
You mean those programmable nanobots that are programmable today will be it in the future?
Quote:
Originally Posted by ncole1
Yes, however, there's a big difference between applying known laws of physics to extrapolate projectile motion and forecasting technological development using arbitrary parameter choices.
Even something like the stock market is at least partially predictable using established statistics, but quite to the contrary, these technological forecasts are not.
The genome project had completed 1% after 7 years (they were halfway through on their 15 year quest), were they on track?
Oh yes because it's an information technology!
I'm with Josseppie here, you just don't see the disruptive change exponential progress brings.
Quote:
Originally Posted by ncole1
I know you don't think this matters as much, but the CPU speed and performance measures as discussed in the Herb Sutter article have broken their previously exponential trend - why do you expect your preferred performance measures not to do so as well?
Check the trends, unbroken since 100 years+
The vacuum tubes are no longer advancing exponentially, does that mean processing power isn't either?
Well no.
CPU speed is no longer advancing, but have never been exponentially IIRC so your point is invalid from the beginning but whatever, is that a show stopper?
Well no!
Because here is the thing, you can actually check the History, read up what happened since 2003, that's over 10 years of what have actually happened!
So what happened? Did CPU speed limitations stop the exponential trend as you suggest?
No it didn't.
Quote:
Originally Posted by ncole1
I won't comment on this point since I don't want to put words in your mouth...
I posted this link before, but it might be relevant again. For a fairly in-depth analysis of the sigularity concept, I believe that this article will provide the types of arguments you are looking for: The Singularity: A Philosophical Analysis
Beware that it is long and almost insanely meticulous in defense of every detail, but if you wade through it and still think there is nothing defensible about the singularity concept, then I'd enjoy hearing your responses to Chalmers specific claims. (BTW: Chalmers tends to be more flexible with the timeframes - thinking it might be more like a 100 years than 30, but he defends the basic ideas with considerable rigor.)
It is long and a bit daunting at times but at least it (tries to) deal with uploading and death which I think is very interesting (have almost finished, wanted to post before not too long though).
I feel there is a slight sloppiness or forgiveness (if people think zombies are in fact their living friends and relatives, then it is "okay"...) here and there, but then he gets back and proves himself, if not wrong, that there is no real truth about it.
I personally believe we are a City of Minds (IIRC there have been proofs about 2 concious, simultaneous states in left-right non communicating brains with damaged corpus callosum, which hints about mindstates and not one exclusive mind) so maybe we are some sort of changing mindsets. You are in a mindset when you work out, later you just remember the memories from that mindset. Maybe we "die" in a sort of way every time we change mindset and the new mindset is created from the emerging complexity...
I also think that most probably you'd die if you "upload" yourself or well you create some new other entity (a "pessimist" in the .pdf ^^)
IMO for this to be wrong (and an upload your mind-state to a machine gives You continue to be You) you must have some sort of "soul" (or other non known force/thing) and as there isn't really a hint about it except that we are in fact alive (we all can know this about ourselves, cogito ergo sum) but even if there is one, why would it split / follow the uploaded data?
Also, what would happen when you make a copy of yourself (in cyberspace say)? Which one will be you? 1.000 copies? Is the biological body still in on it or has it's "soul" been robbed away / weakened?
Maybe we'll know when people actually will do this...
Maybe slow migration will work if we only change from neural pathways and neurons when they actually change, say you learn something new and instead of growing biological cells, nanobots already in place will open a path in the already present network...
Hopefully De Grey will help us live biologically for some time though so we can learn more about conciousness and so but there are other problems like people uploading themselves and even if there is no more conciousness, they will leave the rest of the population behind in an intelligence explosion (for example).
Or maybe we actually are living in a simulation... If I had computronium, I'd build universes for sure so it is quite probable this might be the case...
I know 2030 seems like a optimistic date but considering that information technology advances exponentially its really not. Why the pace of advancement today is so much faster then it was just 10 years ago and why it will be even faster in the next 10.
I hope we see it by 2030 because then I might live long enough to see it happen. I won't be totally shocked if it happens by then because, like you, I expect that the exponential trends will continue. The only reason I don't feel confident about the 2030 timeframe is because of what I know (or believe that I know) about the "hard problem" of consciousness. I'm working on a theory of consciousness that would solve (or side-step) the hard problem, but even if my theory is successful, it is more of a philosophical theory than a scientific one, so I don't yet know how helpful it will ultimately be in overcoming the technical issues of creating a conscious system. My theory depends on our ability to find a great number of "neural correlates" to qualitative conscious experience. I think that this will probably happen by 2030, but then the next step is the real doozy. The raw correlation data needs to be ingeniously tied into a high-level theory (roughly parallel to the way in which Darwin's theory of evolution tied together masses of taxonomic data in biology). I'm working on this too, but it's hard to imagine how it will really work before having the correlation data to work with. My worry is that we might find ourselves in a "catch-22" situation. We might need the help of truly conscious machines in order to solve the question of how to build conscious machines. If that turns out to be true, then we might have to rely on "dumb luck" in our efforts to build the first conscious machines, but that's scary for all sorts of reasons. In any case, my concern is that the crucial creative leap of understanding that is required in order to build conscious machines without the help of dumb luck may take longer than 2030, and dumb luck might take longer than 2030.
All of this is based on my premise that consciousness requires more than just programming complexity. That's a deeper discussion, but, of course, I could be wrong about this. It could be that sheer processing power and increasingly interwoven "expert systems" designed to solve practical problem-solving behavior could end up just being conscious, and that's all there is to it. If that is the case, then I'd be more optimistic about the 2030 date. I suspect, however, that processing power and programming are not quite enough.
I hope we see it by 2030 because then I might live long enough to see it happen. I won't be totally shocked if it happens by then because, like you, I expect that the exponential trends will continue. The only reason I don't feel confident about the 2030 timeframe is because of what I know (or believe that I know) about the "hard problem" of consciousness. I'm working on a theory of consciousness that would solve (or side-step) the hard problem, but even if my theory is successful, it is more of a philosophical theory than a scientific one, so I don't yet know how helpful it will ultimately be in overcoming the technical issues of creating a conscious system. My theory depends on our ability to find a great number of "neural correlates" to qualitative conscious experience. I think that this will probably happen by 2030, but then the next step is the real doozy. The raw correlation data needs to be ingeniously tied into a high-level theory (roughly parallel to the way in which Darwin's theory of evolution tied together masses of taxonomic data in biology). I'm working on this too, but it's hard to imagine how it will really work before having the correlation data to work with. My worry is that we might find ourselves in a "catch-22" situation. We might need the help of truly conscious machines in order to solve the question of how to build conscious machines. If that turns out to be true, then we might have to rely on "dumb luck" in our efforts to build the first conscious machines, but that's scary for all sorts of reasons. In any case, my concern is that the crucial creative leap of understanding that is required in order to build conscious machines without the help of dumb luck may take longer than 2030, and dumb luck might take longer than 2030.
All of this is based on my premise that consciousness requires more than just programming complexity. That's a deeper discussion, but, of course, I could be wrong about this. It could be that sheer processing power and increasingly interwoven "expert systems" designed to solve practical problem-solving behavior could end up just being conscious, and that's all there is to it. If that is the case, then I'd be more optimistic about the 2030 date. I suspect, however, that processing power and programming are not quite enough.
A lot of those issues will be solved once we reverse engineer the brain and that should happen by 2020, give or take a year. That is why people like Ray Kurzweil say we will have consciousness in AI 5-10 years after we successfully reverse engineer the brain and that will be around 2030.
As we race to the singularity it will not only impact humans but animals as well. Because of that one of the discussions is should we engineer animals to be as smart as humans. I am all for it especially mine.
Should we engineer animals to be smart like humans?
Please register to post and access all features of our very popular forum. It is free and quick. Over $68,000 in prizes has already been given out to active posters on our forum. Additional giveaways are planned.
Detailed information about all U.S. cities, counties, and zip codes on our site: City-data.com.
Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
