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.
Another breakthrough that will help have reverse aging.
Researchers at the Salk Institute have discovered a toggle switch for aging cells. By controlling the growth of telomeres, it may eventually be possible to coax healthy cells to keep dividing and generating even in old age.
The cells in our bodies are constantly dividing, replenishing our lungs, skin, liver, and other organs. Regrettably, most human cells can't keep on dividing forever. Each time a cell divides, a cellular "timekeeper" at the ends of the chromosomes shortens. These timekeepers, called telomeres, are like the aglets at the end of your shoelaces — those important bits of plastic that prevent the lace ends from fraying. But in the case of shortened telomeres, cells are no longer able to divide, resulting in a host of aging-related complications, including organ and tissue degeneration.
Man, it was just a couple of years ago when I choosed the 45nm E8400 over the 65nm Q6600... and now they will make 7nm chips. That's 41 to 86 times moore :-)
No one is surprised by this - but again, size is not a performance measure.
Even Herb Sutter, the author of "The Free Lunch is Over", expected transistor sizes to continue to decrease.
But until you demonstrate the degree of parallelizability of the specific computation required for a "strong" AI which is most relevant to a singularity, the reduced transistor sizes is not an argument for a singularity, given the fact that single-core CPU speed has been flat since 2003.
It is a bit too optimistic I think (if you combine an "Ãœber" and self driving cars, you could cut out 9 out of 10 cars) but what ever number it might be, less cars needed means the cars in use could be more expensive (to create / tend to) for the same user price or just cheaper for the consumer of course.
Well you might be right of course and it is a good prediction using today's tech I think.
Now if we take one of Ray Kurzweils predictions:
2030s: Nanomachines could be directly inserted into the brain and could interact with brain cells to totally control incoming and outgoing signals.
If this comes true (I bet it will, I just won't bet on the exact year), this would actually remove all the medical work to make someone see through a camera.
Very good point - it is likely to happen eventually, but Kurzweil is leaping to conclusions by predicting a year without justification.
Also, there's no good reason to think the nanotech will take the form Kurzweil is envisioning. That's not a reason to think it won't happen of course, but an interesting side comment.
I have been giving this some thought and I think you get caught up in the details of specific technologies while not looking at the bigger picture. To coin a phrase "you can't the the forest through the trees". Now sure you can possibly get technical and site specific technologies that have not advanced exponentially but if you stand back and look at the bigger picture overall computers have.
In the 1970's we had main frame computers that took up entire rooms.
In the 1980's we had desk top computers that had more processing capability then the main frame computers of the 1970's.
In the 1990's we had lap top computers that had more processing capability then the desk top computers of the 1980's.
Around 2005 we had smart phones that had more processing capability then the lap top computers in the 1990's.
Today we have wearable computers that are more powerful then the computers in 2005.
The impact to society by each of these paradigm shifts has been fantastic and at each stage there was a group of people who said yes this happened but this is where it will stop and they had their reasons and every time they were wrong. No different then what is happening today. You have a group of people that say yes we have come this far but it will end now and why. Then, as people who said it would stop before them, they will be wrong and in the 2020's we will go from wearable tech to merging with the tech and it will just get even smaller faster and more common in the 2030's. That will be the so called tipping point as we will really start to advance fast and it will impact things like longevity and that is why I call the 2020's the pre-singularity.
I agree that each decade for at least the next 2 or 3 is likely to bring significant innovation when it comes to electronics. However, once you switch from a "trees" perspective to a "forest" perspective, you lose your quantifiability.
And what the heck is an "overall exponential advance"? The word "exponential" describes a mathematical function, thus, it is a label that can only apply to quantities and not to qualities. The problem I have with your argument is that you start by broadly generalizing from the "trees" (i.e. specific, quantifiable characteristics of computers such as memory storage per $ or CPU speed) to the "forest" which is qualitative rather than quantitative, and then after that you try to claim it is exponential.
But that's nonsense. Only quantities can be exponential! You're not making any more sense than someone who says "The Matterhorn is 3.6705 times more beautiful than Mount Washington in New Hampshire".
To say that that statement is nonsensical is not to deny the beauty of the Matterhorn; it's simply pointing out that you are invoking false quantifiability of a qualitative value. Nor is it denying that there are quantitative measures applicable to mountains; it's perfectly valid to say, for example, that "The peak of the Matterhorn is 8,404 feet higher than the peak of Mount Washington".
Similarly with computers and technology in general. Yes, we are advancing rapidly; and yes, some things about computers can be quantified.
But as in my mountain example, it makes no sense at all to say that technology is advancing exponentially in general. That statement is not more meaningful than "The Matterhorn is 3.6705 times more beautiful then Mount Washington in New Hampshire".
No one is surprised by this - but again, size is not a performance measure.
Even Herb Sutter, the author of "The Free Lunch is Over", expected transistor sizes to continue to decrease.
But until you demonstrate the degree of parallelizability of the specific computation required for a "strong" AI which is most relevant to a singularity, the reduced transistor sizes is not an argument for a singularity, given the fact that single-core CPU speed has been flat since 2003.
I don't understand what you are looking for, the data is everywhere, computer power price is crumbling since 1890 and CPU power goes up! Still does! Just compare a PC from 2005 with one from today...
And if you can't imagine a neo cortex (6 layers of around 90.000.000.000 neurons interconnected) dreaming of parallel computers to run then I can't do anything for you (hint: it's ridiculously easy to treat in parallel).
So please, stop dragging in that "speed stopped @ 2005" because it doesn't matter and you have the proof just by comparing 2014 computers vs 2005 computers.
Very good point - it is likely to happen eventually, but Kurzweil is leaping to conclusions by predicting a year without justification.
Also, there's no good reason to think the nanotech will take the form Kurzweil is envisioning. That's not a reason to think it won't happen of course, but an interesting side comment.
Kurzweil is basing his predictions on exponential trends, as we all know, so it seems most predictable to have the tech in the 2030s.
Today there is DNA origami which can deliver drugs to specific places, can compute (they claim C64 like power ) and is under 1um (the size that can be squeezed into a living cell without wrecking havoc).
They can grow them by the billion too BTW so if they can make these bots to go to neurons (the right place) and get past the blood/brain barrier, they "just" have to get around the communication problem, sending data to billions of nanobots.
If I were to predict a day when nanobots could interfere with neurons (in animals) I'd say in a couple of years if not earlier. When will it be mainstream? 15 years of development for humans and FDA approvals + some years and we are in the 2030s
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.
) and is under 1um (the size that can be squeezed into a living cell without wrecking havoc).
