Cryonics

Cryonics is the idea that you can freeze a body (or just the brain) and keep it at low temperature until such time as the person can be revived (or rebuilt) using an advanced technology such as nanotechnology.

Alcor is one of the companies that will do the freezing, and Marginal Revolution today linked to an article on Alcor's site discussing the probability that this will work. The article breaks the issue down into a number of factors, gives "gut level" odds for those factors, and comes up with an overall probability ranging from 15% chance of success, to negligible. The comment on Marginal Revolution was "I think simply no one will care to thaw you out." I've gotten this same reaction when I've discussed cryonics with friends.

I think the assumption is that there will be a very long delay (hundreds of years) between when you are frozen and when you can be revived. And that there will be so many people frozen during that interval that you will be one of thousands (millions!) waiting to be revived. And that revival will be so difficult that only a few, if any, will be revived. So even if it were all possible, your odds of survival would be low.

On the other hand, one comment to the MR blog was that your descendents would be there to revive you, and would do so because they in turn hope to be frozen and revived.

I don't think any of these assumptions are correct. Here's my note to Marginal Revolution:

There are few diseases (including aging) that could damage your brain more than having your head cut from your body and frozen. The ability to repair this huge damage implies the ability to repair nearly any damage in a living person. So no one will be frozen after the point where revival becomes possible (because no one will die!) In fact, since repair of lesser injuries would be possible before that, cryonics would already be an unused technology (no new brains frozen) long before revival is possible.

This has some consequences -- if you are revived, it will be to a society of immortals. Second, cryonics will stop being used before it has been proven to work, even if it does work (because no one dies and is frozen.) So only a limited number of people will ever be frozen, and they will all have to take on faith that the technology can work.

So I see cryonics as purely a transitional technology, to get you from our current period, where people all die, to the "immortal" period, where almost no one (involuntarily) dies. There will only be a limited number of people frozen. As time goes on, the cost and difficulty of reviving them will drop, so that eventually all of the frozen who can be revived will be. The only challenge will be for your frozen brain to survive the period until the technology becomes available. This really depends on how long it takes to develop revival technology.

Scanning Brains

There's no way to know how long it takes to develop this level of technology, but there are a couple of general observations you can make. First, it's going to be simpler to scan a brain and extract all relevant information than it is to rebuild/repair the frozen brain and body. Scanning is a prerequisite for repair.

If you wanted to do a destructive scan, you could probably do it now. Microscopes are more than good enough to see each neuron, dendrite and synapse. You could scan a strip, burn or scrape the strip off, and repeat until you've consumed and scanned the whole brain. Assuming the brain is frozen, you can take your time about this.

I'm not sure anyone knows exactly what it would be important to record about the state of a neuron. It would also be a lot of data. Just recording which neuron (of 100 billion) connects to which other neuron (assuming an average of 1000 connections per neuron) would be hundreds of terabytes! Add some more details about the geometry and chemistry of the situation, and you are talking sizeable storage requirements. But not impossible, even with current technology.

More difficult would be simulating the scanned brain, to produce some simulation of the person you've scanned. Computers are nowhere near fast enough (yet) to tackle this, even if we knew exactly what a neuron does, and what a network of them does when you learn. And we also have to provide some inputs (simulated eyes and ears) and outputs (simulated vocal chords at least) so your simulation can interact with the world.

But no matter how difficult this simulation is, it still strikes me as easier than rebuilding a brain. Any nanotechnology capable of repairing 100 billion neurons in a reasonable amount of time is also going to provide a huge amount of computational capability. In fact, a huge amount of computation is required to guide all the little robot assemblers moving through the frozen brain, repairing the damage.

So in my view, true artificial intelligence, including the simulation of scanned brains, comes before any revival of people frozen with cryonics. So they are not only waking to a world of immortals, but also to a world of artificial intelligences. In fact, I expect they would wake as artificial intelligences.

Mike's Newspaper
Writings
Projects
Sea of Memes
Photos