Synthetic Life Is Here Already

Synthetic Life Is Here Already

Synthetic Life Is Here Already

Hi, I'm Tim Tyler - and today I'll be discussing the idea that synthetic life is here already.

Evolution has been trying to improve on existing organisms for billions of years - and in modern times some humans have taken up the task - using breeding programs and genetic engineering to improve on the traits of a variety of creatures.

However, there are limits on what can conveniently be done using incremental improvements to existing organisms, so some have taken to performing more radical redesigns.

Synthesising living organisms from scratch has been a dream for centuries - and has been promoted in modern times by people such as Chris Langton, and Eric Drexler:

[Eric Drexler footage]

However, it has long been recognised that there are down-sides to the practice.

Mary Shelley's 1818 novel, Frankenstein vividly dramatised some of the risks of the enterprise.

In particular creating organisms with a lot of power, seems potentially problematical - if the creatures come to elude the control of their creator.

[xxx footage]

One of the problems is self-replicating agents getting out of control. Here is Bill Joy describing the problem:

[Bill Joy footage]

It is concerns such as these that led to Eric Drexler back-tracking on his original vision:

[Eric Drexler footage]

One idea that seems to be widespread is that you can avoid problems by preventing the machines from reproducing. If you don't give the machines the power of self-reproduction, and make them dependent on humans manufacturing them, then they are more likely to act as beneficial symbionts - and less likely to act as parasites or predators.

There are a few things to note here:

One is that growth is almost as problematical as self-reproduction. As has been noted by Steve Omohundro, sufficiently powerful systems are likely to want to expand to to occupy more space/time - and take in more resources - in order to better meet their goals - and this presents almost exactly the same set of hazards to other creatures in the environment that self-reproduction does. It isn't just self-reproduction that you need to look out for.

The next is the idea that this is a scenario is going to occur at some point in the future. In my view, the new replicators are here already - and have been here for thousands of years. We have new replicators in novel substrates copying themselves all over the planet now - and they are swarming in incredible numbers. Their phenotypes are having an enormous impact on the planetary ecosystem - heating the planet up, making it glow at night, and causing the current mass extinction. Synthetic living organisms invading our ecosystem is not some event that we have to watch out for in the future - but rather it's a ubiquitous phenomenon here and now. The new replicators surround us, are inside us - and make us what we are. Were it not for them, we would still be beasts - primitive cave-dwellers.

Lastly there is the idea that forcing the new organisms to be dependent on humans for reproduction will somehow stop them from taking over control of the planet.

Human control over machine reproduction would indeed create a desirable, mutually beneficial symbiotic relationship between the two parties. However, just because you are in a mutually beneficial symbiotic relationship, that doesn't mean both parties are benefitting equally.

Today, we depend on memes - and we don't want to get rid of them - they are good for us. We can't just decide to ditch ideas like language - we value what they offer us too much. The result is good for us - but also good for the memes, which grow in numbers as a result. It's the same with machines - we don't want to throw our mobile phones away - we value what they give us too much.

Nucleic replicators and memes coexist in the same ecosystem, and ultimately compete for the same space and the same resources. If the memes get a bigger fitness boost out of the symbiosis than the nuclear genes do, then they will increase in numbers relative to those genes - and so come to represent an increasing fraction of the biosphere by weight. That describes what has been happening so far.

If the new replicators continue to get more out of the symbiosis than the nucleic genes do, a likely outtcome is a society numerically dominated by intelligent machines - which are much better at propagating memes than humans are.

In such a scenario, there may still a mutually-beneficial symbiotic relationship between the parties involved - and those humans that opt to continue the symbiosis with the machines might still do systematically better than those who attempt to put an end to the affair. However, humans as a whole are not doing terribly well.

Bill Joy describes this kind of scenario as follows:

[Bill Joy footage]

You might say that the road to a world dominated by intelligent machines is paved with corporate profits.

At each step, the memes offer the nuclear replicators a kind of Faustian bargain, which gives them short-term benefits, but at the price long-term ruin.

So, to summarise, preventing machine self-reproduction in the future seems like closing the stable door after the horse has bolted. The self-replicating entities that are most likely to dominate future ecosystems are already here; they are reproducing in an uninhibited and unrestrained fashion; and we can't stop them - because we don't want to.


Tim Tyler | Contact |