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.
Ray Kurzweil is not any more of a qualified authority on medicine than I am.
You seem to think just because your cherry picked technological parameters advance exponentially means that therefore anything goes.
It doesn't work that way.
Ray Kurzweil is a expert on this subject. Since I do not know who you are I will not say he is less or more qualified then you. However I have yet to see any real proof from you that shows what Ray Kurzweil talks about it wrong.
That being said he, and others I post about, do not cherry pick and the proof "is in the pudding" to use a phrase. I mean I can see it in my life time as things are advancing a lot faster in the last 10 years then they did in the previous 20.
Last edited by Josseppie; 09-09-2014 at 11:29 PM..
Even if that were true in general, your conclusion does not follow. It's a non-sequitur.
That is not my conclusion that is what I have been reading about the impact of computers helping out with treatment options. Just read the paper as that is what the Mayo Clinic expects and they are one of the top hospitals in the world.
Just because Rhodin says it will double every 73 days doesn't make it true.
This type of precision on such a future prediction should be taken with a grain of salt.
He is not the only one just one of many and its all because medicine is not becoming a form of information technology and computers are getting involved more and more.
Valmond did not address the issue of heat dissipation and speed of light problems.
And this still leaves aside the problems of increasing quantum and thermal noise as components shrink or the flatness of processing per unit energy since 2003 and therefore limits by energy consumption.
Also, this doesn't address the issue of more or less parallelizable computational tasks. Just because one calculation can be massively parallelized with very little serial constraint relevant to Amdahl's Law, does not mean the same applies to other calculations or algorithms.
For the Amdahl's "law", I already told you it isn't hampering today's algorithms. They solve our problems twice as fast every year so even if theoretically you can argue about it, in real life it doesn't matter and todays algos would happily crunch data with a billionfold parallelisation extremely efficiently so it leaves kind of a big leeway before Amdahl's "law" would even start to kick in.
But actually, we just disagree, you think the future computers won't be faster. I do. There you are.
Good point. I can see a time in the not so distant future when all of our organs are replaced with artificial ones. Since this technology is advancing exponentially it really could be in the next 16 years.
Technical questions like that I will have to leave to people in the IT field. I can only say that so far they have not been a problem as computers have shrunk a lot all the way down to Google glass and even smaller as I have seen and posted articles on them small enough to fit in the human body. Plus people like Ray Kurzweil do not see that as a issue moving forward.
Computers may have shrunk, but size is not a measure of performance.
No matter how much volume-related limits can be circumvented, if computation per unit energy consumption flattens, computation hits a brick wall. This is discussed in the Herb Sutter article I referenced earlier.
I don't care if you could install Windows on something smaller than a blood cell, or even an atomic nucleus for that matter, it does not change that basic fact. Size is irrelevant to that potentially limiting factor.
Quote:
Originally Posted by Josseppie
That is not my conclusion that is what I have been reading about the impact of computers helping out with treatment options. Just read the paper as that is what the Mayo Clinic expects and they are one of the top hospitals in the world.
No. You cannot get from facts about treatments currently available at Mayo Clinic to the conclusion that actuarial escape velocity will happen before 2020, without invoking hidden premises in your argument.
I am only asking you to state explicitly your auxiliary assumptions and justify them.
Quote:
Originally Posted by Josseppie
He is not the only one just one of many and its all because medicine is not becoming a form of information technology and computers are getting involved more and more.
Yes, computers are increasingly involved in medicine.
However, it doesn't follow from that (without additional unstated assumptions) that ANYTHING will double every 73 days in the year 2020.
To make an analogy, computers are increasingly involved in cars also. Does it therefore follow that in the year 2025 a car will be able to go 100,000 miles per hour on an interstate highway? Of course not!
The question is, what are the limiting factors in the medical treatment, and how will those specific factors change over time? It doesn't suffice to simply say that parameter X of technology A is increasing exponentially, and technology B is increasingly making use of technology A, therefore Y will become possible with technology B by year Z.
That argument is formally logically invalid and therefore is incomplete without discussion and justification of how they relate to each other and at what point that relation should be expected to no longer hold.
Ray Kurzweil is a expert on this subject. Since I do not know who you are I will not say he is less or more qualified then you. However I have yet to see any real proof from you that shows what Ray Kurzweil talks about it wrong.
That being said he, and others I post about, do not cherry pick and the proof "is in the pudding" to use a phrase. I mean I can see it in my life time as things are advancing a lot faster in the last 10 years then they did in the previous 20.
Kurzweil is very qualified on computer science, but Kurzweil is not a qualified authority on medicine.
For the Amdahl's "law", I already told you it isn't hampering today's algorithms. They solve our problems twice as fast every year so even if theoretically you can argue about it, in real life it doesn't matter and todays algos would happily crunch data with a billionfold parallelisation extremely efficiently so it leaves kind of a big leeway before Amdahl's "law" would even start to kick in.
But actually, we just disagree, you think the future computers won't be faster. I do. There you are.
NO. I think future computers will be able to implement highly parallelizable algorithms faster than they can today.
But I am not convinced all computational problems are that parallelizable - in fact, I'd argue that given the halting problem, that cannot possibly be the case.
Once you admit that some computational tasks can only be done serially, then you can no longer claim that simply because something is dependent on computation means it will therefore continue to improve. The argument requires an additional assumption - that the specific computational needs for that specific problem are fully parallelizable. And this is an assumption that must be justified.
Kurzweil is very qualified on computer science, but Kurzweil is not a qualified authority on medicine.
Actually Ray Kurzweil was on the genome project and is an expert.
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.
