Ray Kurzweil - Immortality Only 20 Years Away | Hacker News Reader

Ray Kurzweil - Immortality Only 20 Years Away

telegraph.co.uk

20 pointskeltecp1146 comments

I'm suspicious of anything said to be 20 years away. It's far enough in the future that you can be a little hand-wavy about how it's going to happen, but not so far that no one would care. A variant of Occam's razor implies that when something is described in a news article as 20 years away, the 20 came from the demands placed on news articles rather than the nature of the subject.

"Generally the prediction comes as far as possible in the future while still being within the lifetime of the predictor." I don't recall the source of this paraphrase.

Sustainable nuclear fusion has been twenty years away for the last fifty years or more.

It still is twenty years away.

That's a long twenty years.

Saying a research project is X years away is silly, given that you could get it done faster by throwing money at the problem.

The moon landing should have been 30 or 40 years away from Kennedy's famed announcement, but it was given a high priority.

Nuclear Fusion hasn't gotten enough funding, imho. Solar too.

Somewhat true, but there's an important difference between an engineering challenge and a research challenge. I know next to nothing about nuclear science, but I guess that fusion will require breakthroughs. Moon landings - while not particularly easy - simply require a lot of quality engineering. Engineering like that is mostly a function of the bags of money you throw at it.

From my limited understanding, it is just iterative improvements in fusion that are required. You shoot a laser at something compressed in a strong magnetic field, and use the dissipating heat from fusion to boil water to spin turbines to power the laser and magnet. This whole process already works, but doesn't net energy gains. Improvements in efficiency of the laser, magnets, or power gen could make it tip into net energy gains. It's still an engineering problem from then on, as you need to compete on $/Kwh.

The drug approval process alone would take up most of those 20 years. This means that the methods for making us immortal would have to be perfected in 10 years, at most. The only way this is going to happen is if human-level or better AI is finished, like, next week.

Meanwhile, in the real world, medicine has made only incremental advances in the past few decades. Witness the fact that pharma is falling off a cliff because they haven't come up with many blockbusters recently: http://www.fiercepharma.com/story/bail-faster-patent-cliff-a...

Kurzweil is 61. He's fantasizing that something will be ready in 20 years, which is about his life expectancy at this point. Sad.

>The drug approval process alone would take up most of those 20 years.

That's for U.S. regs only, right? They could (dangerously) release for broader testing in other countries.

Yes. Though Europe is just as bad.

They could, but AFAIK they don't (not that I pay much attention to the drug market in Lower Elbonia). Roughly speaking, the looser a country's drug regulations, the smaller its market.

There's not really that much time to be saved here anyway, not compared to the time frame we're talking about. Look at figure 2.1 here: http://books.google.com/books?id=zlzwOEtDu4IC&pg=PA11...

If you skip the Phase III trials and the red tape, you save maybe 5 years. Probably not enough to do much for a codger like Kurzweil, any more than it did for a couple of other dead people who loudly proclaimed that they would live forever.

http://en.wikipedia.org/wiki/FM-2030

http://en.wikipedia.org/wiki/Robert_Anton_Wilson

This same topic came up in a comment thread on HN a while ago... The drug approval process in most countries definitely needs to be shortened.

There are thousands of people that have terminal diseases which are unresponsive to treatment that are willing to undergo experimental therapies. Animal testing is not very nice for the animals and doesn't always provide useful results. Voluntary human trials need to be more accessible...

Though on the other hand most new drugs aren't that much more effective than the old ones. If at all.

Drug approval processes can take anywhere between several months and a decade, the problem here is that a longevity drug would be screwing around with so many systems in our body that the usual 5 year process (for standard drugs) or even the 10 year process (like for Hep C meds and such) would likely look like a blur to the people developing longevity drugs. My personal bet, the drug approval process alone would take up all of those 20 years.

The only way a longevity drug would get through testing quickly is if the benefit seriously outweighed the costs and there was serious public outcry for it.

The only fast way to market for a drug like this might seriously be to sell to China or Africa and prove it in the field. The only problem here is that you won't ever get the big bucks from Westerners.

Perhaps the best market for drugs like this would be to sell explicitly to countries with universal health cares and form an agreement that a subscription constitutes a 20 year contract. Instead of selling on a premium or by quantity, you sell by longevity and nail entire countries on dependence.

>a longevity drug would be screwing around with so many systems in our body that the usual 5 year process (for standard drugs) or even the 10 year process (like for Hep C meds and such) would likely look like a blur to the people developing longevity drugs. My personal bet, the drug approval process alone would take up all of those 20 years.

I think we'll see a host of longevity drugs working on different pieces of the puzzle - a longevity Unix toolkit rather than one monolithic drug. Each of them will have a tractable clinical endpoint to measure, and in the aggregate they will extend lifespan. Of course, we won't really know that until we see a lot of 100+ year old people, unless we get AIs first with models of the human body so good that we can be really confident that they're right. I'm not holding my breath.

>The only fast way to market for a drug like this might seriously be to sell to China or Africa and prove it in the field. The only problem here is that you won't ever get the big bucks from Westerners.

I don't know how patent law would work here - I think your patent would expire in the West while you were field testing your drug in China, so you would fail to enhance shareholder value.

a) He's not talking about drugs, the word 'drug' is not even mentioned in the (hype-filled 'article').

b) Even if he was basing his ideas on one or more drugs, what sort of stumbling block does 'government approval in the USA' present if you're a multimillionaire with contacts in high-tech industries, looking to avoid death? Not much of one.

c) The article claims 'could be as little as twenty years' not 'in twenty years exactly'.

d) His predictions are based on transformational technologies such as nanotechnology and/or AI, they wont follow on linearly from current ailing pharma, so that's not relevant.

e) He has made pots of cash from software that depended on improving technology for many of it's improvements (voice recognition), has predicted when computing power would hit certain levels in advance and designed and made inventions to meet the predictions and his predictions were accurate enough to do that, he has written whole books presenting and justifying his claims, he works part time from a doctor's office while being treated with the currently best available anti-aging and rejuvenation treatments and has changed his life around gunning for longevity. To dismiss him as merely a sad fantasizing old man is completely unjustified as well as being an ad-hom attack.

>He's not talking about drugs, the word 'drug' is not even mentioned in the (hype-filled 'article').

I think his nanobots would be regulated by the FDA as drugs.

>Even if he was basing his ideas on one or more drugs, what sort of stumbling block does 'government approval in the USA' matter if you're a multimillionaire with contacts in high-tech industries, looking to avoid death, present? Not much of one.

The same stumbling block that you would have getting an iphone if they were illegal. They would never have been made in the first place.

>His predictions are based on transformational technologies such as nanotechnology and/or AI, they wont follow on linearly from current ailing pharma, so that's not relevant.

Right - that's why I qualified my statement with "The only way this is going to happen is if human-level or better AI is finished, like, next week." It could happen, I just wouldn't bet that way.

>Ray Kurzweil has counted to infinity... twice

Sure, he's a smart guy. I could dig up other smart guys who work closer to the field who don't think immortality will ever be achieved. Smart guys can be right about one thing and wrong about another.

I think his nanobots would be regulated by the FDA as drugs.

I think that would be like the FDA trying to classify a programming language as good or bad.

The same stumbling block that you would have getting an iphone if they were illegal. They would never have been made in the first place.

They will be made so they can be used on inanimate objects and used or tested on animals. They will be made so they can be tested for FDA approval. Therefore will be available on the black market for the right people with enough money.

Right - that's why I qualified my statement with "The only way this is going to happen is if human-level or better AI is finished, like, next week." It could happen, I just wouldn't bet that way.

That isn't the only way it will happen. Specifically mentioned are artificial organs used to replace failing organs. That is starting to happen right now and could extend life expectancy.

Sure, he's a smart guy. I could dig up other smart guys who work closer to the field who don't think immortality will ever be achieved. Smart guys can be right about one thing and wrong about another.

Whether he's right or wrong isn't the issue, your dismissive comment as if he had just thought of it yesterday and his ideas were obviously meritless and unsupported is part of the issue (it's unfairly misrepresentative) and your personal attack that he's just an old man grasping at straws is just bad form.

FDA: drugs = "articles intended for use in the diagnosis, cure, mitigation, treatment, or prevention of disease." That would include nanobots.

>That isn't the only way it will happen. Specifically mentioned are artificial organs used to replace failing organs. That is starting to happen right now and could extend life expectancy.

What happens when (not if) your brain fails? This would just buy us a few more years before we became senile. Worthwhile, but not immortality.

>your personal attack that he's just an old man grasping at straws is just bad form.

Maybe. It's also true.

FDA: Silly, will happen, will be worked around.

Years: Immortality isn't one step. Extra years are all you need for immortality, as long as you last long enough for the next treatment that gives you extra years. What happens when your brain fails? Hope there's a cure by then. Said failure wont be magic, it will be biological/physical/chemical/electrical and one day repair/replacable at some level.

Straws: Not true. He's not grasping at straws because he's old, he's pursuing a several decades long interest in technology, health and longevity. Also, wanting a longer life isn't 'sad' nor does it make him a 'sad' person.

Years: Immortality isn't one step. Extra years are all you need for immortality, as long as you last long enough for the next treatment that gives you extra years.

Aubrey de Grey has termed this concept actuarial escape velocity.

Would be a pity if Kurzweil's predictions follows the path of fusion energy - always, eternally, twenty years away.

At any rate, I question the practicality of /physical/ immortality. If we're swapping out blood and organs with effectively superior equivalents, why shouldn't we reexamine the necessity of a fully human-equivalent body? If the entire purpose is to establish life support for the brain and preserve and augment the human sensorium, then why keep blood? Or the liver? Or the stomach?

I agree... I think it's quite likely that once we have the tech to replace all the components of the human body the possibility will also exist to do away with it entirely.

I would definitely opt-in for a robotic body! I think I would prefer if it still looked human, but we can debate that in 30 years or so!

People'll pay for a humaniform body for at least the first few decades.

Then the Grinders get involved, and the cultural body image paradigm starts to shift ever so slightly...

Grinders?

Grinders, modders, etc - "Grinders" is Warren Ellis's term for it, and I've been rereading his stuff recently, so...

Basically, anybody who starts tinkering with the physical form itself - whether by small modifications, such as improved eyesight, or large modifications, like swapping your legs out for a tank chassis.

Why not just build a clone and "flash in" the contents of your brain?

Because a clone will take about two decades to mature? And if it's a full clone, will in all likelihood have its own ideas about who occupies its head? And will probably physically develop different from you, as development is partially environmental?

All sorts of problems with the cloning path to immortality.

Or you can assume that by the time you figure out how to download the contents of a brain, you'll have figured out how to speed up the cloning process. Or just start growing a "mindless" clone when you're 30 and start inhabiting it when you turn fifty and then repeat the process.

Technically, these are all doable. Now, ethically and morally... that's a completely different story.

This is all nonsense anyway, but if you're going to do it then I would suggest a way that does not introduce generational noise between copies, or you might find after 'n' generations that you are not immortal after all.

Best to keep 'cloning' from your original cells, not from the ones that have gone through possibly many such cycles.

Alternatively, save the entirety of your genetic code as a data file and update it to the latest storage medium every few decades. Even the best-preserved original cells will become unusable after a while.

Writing in The Sun, Mr Kurzweil said:

This is the UK equivalent of 'Writing in the National Enquirer...'. I like Kurzweil, I own a few of his books, but most Sun readers think a transhuman is a man who wears dresses.

Here is the entire Sun 'story'

Robo-swot 'will arrive' by 2028

MACHINES will “soar past” human intelligence within two decades using a combination of computers and biology, a visionary scientist claimed yesterday.

Dr Ray Kurzweil said nano-robots – microscopic units of artificial intelligent hardware – will eventually become “deeply integrated” in the environment, our bodies and brains, he told a meeting of top boffins in Boston, US.

Speech recognition pioneer Dr Kurzweil said two dozen regions of the brain had already been modelled and simulated by “reverse engineering” intelligence.

Better informed now? No, didn't think so. BTW, a swot is a nerd and a boffin is a professional scientist, in UK gutter-speak.

In 20 years i'll have my own jet pack and rocket car too. Rosie, can you get the door please?

Although he hasn't proposed a total replacement of human parts the hint is there, consider this:

- the inevitable happens and technology has progressed to the level where everything in one's body has been replaced by 'cyborg parts'.

- You are now living past 500 years.

- After 500 years your brain has a total revamp and replaced by chips, it is now more powerful and would only need replacements every x years.

- eventually your total brain is replaced with new technology, ensuring all memories are kept etc.

- In 1000 years there is nothing original in your body except some old 'data'.

- Now that the 'original you' is totally replaced IMHO you have d i e d and a Cyborg took over your identity! As a corollary this Cyborg has now took over your consciousness also!

One consolation would be that if these events materialize at some point the Cyborg would try to use some of your Lisp stored 'data' and suffer infinite recursion and die too! :)

This quote speaks to the validity of the person making this claim. From TFA:

Within 25 years we will be able to do an Olympic sprint for 15 minutes without taking a breath, or go scuba-diving for four hours without oxygen.

BATBOY FOUND!! Page 33! NEW PICS!

If your blood replacement (whatever it might be) was able to store 50 times the amount of oxygen per volume vs real blood then there's no reason why things like this might not be possible.

They've already made headway toward something like this with artificial blood substitutes: http://en.wikipedia.org/wiki/Oxycyte

No known reason, maybe.

You would have to have a way to get rid of the carbon dioxide or you would die of respiratory acidosis.

CO2 + H2O -> H2CO3 -> H+ + HCO3-

This kind of scares the shit out of me.

To me, the human condition seems like a very fragile thing; more so than many seem to appreciate. With predictions like Mr Kurzweil's, I wonder and worry that our possibilities will grow too large too fast and beyond control. I usually assure myself in optimistic spirit that "the humane will prevail". But it's no absolute truth.

I'm reminded of an interview with Douglas Hofstadter that I think was posted here on HN, on a similar topic.

http://tal.forum2.org/hofstadter_interview

Hmmm, we barely understand how human biology works, and won't be able to cure most current deadly diseases in that time. The more we learn about human genetics, the more we realize the less we know. In other words immortality is only 20 years away in the minds of those who believe in their own hype.

I think the best argument contra Kurzweil's timescale is that business organization and the human element in infrastructure processes is incompressible below a certain threshold. I make this argument here:

http://www.fightaging.org/archives/2005/09/reading-the-sin.p...

and here:

http://www.fightaging.org/archives/2005/09/nanotechnology.ph...

I don't think there's a good argument to be made contra what Kurzweil expects to happen. At this point a transhumanist society of enhanced, long-living humans+ seems fairly inevitable, arriving in probably three to four decades rather than two.

Please. This is such an outrageous claim to make that you need evidence to back it up, calling anything on this scale 'inevitable', let alone to stick an upper limit on it less than two centuries into the future is not very realistic.

So, instead of looking for all the excellent arguments that have already been made 'contra' Kurzweil, the biggest of which is that nobody can predict the future, let's hear some of the reasons why this is 'inevitable' that do not contain any qualifications or build upon other technologies that we do not have today.

Psychology is yet to answer how brain is capable of simulating future, how does it work with time or just the "simple" question—why we sleep among other things. Kurzweil overestimates our understanding of how we work.

Seems like with technology, each of our life spans can be maximized. So if any organ fails, it can be replaced (or maybe preemptively replaced so it lasts longer). However, the biggest problem is going to be our brain cells. After 100-120 years, our brains are just about used up. Doubt its worth persisting much longer after that.

Maybe in 500 years they'll figure out a solution to that, but I wont hold my breath.

The organisms currently composed of organic nanobots (cells) are a beautiful, complicated mess. It'll be some feat to augment that mess in the ways he predicts and as quickly as he predicts.

A human is like a giant software package that's been patched for billions of years. It's had a long time to get pretty complicated and to take advantage of every physical phenomena we've yet to understand.

Seen on a T-Shirt: Ray Kurzweil must die.

it's gonna get crowded.....

Reply on news.ycombinator.com