Sign In

Remember Me

Merge Or Purge

The futurist Ray Kurzweil and I have crossed swords several times in the media in the past on the question of whether humanity will “merge or purge” our increasingly intelligent machines.

This essay presents the two main views on what is likely to happen to humanity as our machines become artilects (godlike, massively intelligent machines trillions of trillions of times above human levels). I have labeled these two views the “merge” view (Kurzweil’s) and the “purge” view (mine) for simplicity. I begin by discussing the merge view.

The “Merge” View (Kurzweil)

Kurzweil thinks that the most likely scenario for humanity as our machines become increasingly intelligent is that, to quote him: “We will merge with our machines.” He thinks that in subsequent steps, the human side of our cybernetic selves will soon be swamped by the vastly greater machine capacity and that we will become increasingly machine-like artilects. He is not alone amongst prominent commentators on the species dominance issue, as many others hold this view. Another prominent figure with a similar take is cyberneticist Kevin Warwick in the UK.

The “Purge” View (de Garis)

I think it is much more likely that the Terrans, the people opposed to the creation of artilects, will go to war against (purge) the Cosmists, who want to build artilects, as well as against the Cyborgists, who want to merge with their machines as Kurzweil does. This means that the Terrans will also go to war against the artilects themselves before the machines become too intelligent. This war will be waged with advanced 21st century weaponry, killing billions of people – hence the term “gigadeath.”


Kurzweil is skeptical of my scenario because he thinks that if it finally does come down to a war between the Terrans and the Cyborgs/Artilects/Cosmists, it would be “no contest.” The latter would be much smarter and more capable than the (human) Terrans. As Kurzweil puts it colorfully: “ (It would be) like the US Army fighting the Amish!” (For those of you unfamiliar with the Amish, they comprise an American religious sect that doesn’t use any technology later than the 19th century. They drive around by horse and buggy and abstain from using telephones or the Internet.)

Of course, Kurzweil is correct in thinking that if the Terrans wait until the artilects and cyborgs come into being before hitting back, it would be no contest. For this reason, I believe the Terrans will take Kurzweil’s argument to heart, reasoning that they will have to strike first while their intelligence levels are sufficient to have any chance of winning.

So, Kurzweil’s “no-contest” argument can thus be refuted. But there are other points that can be made regarding a potential conflict between the Cosmists, Cyborgists and the Terrans.

What if Cyborgism proves so popular, that there are, in effect, no Terrans left since everyone follows the cyborg route? If this were to happen, then there would be no gigadeath-scale artilect war because it takes at least two sides to wage war.

The heart of the issue, and the major thrust of the remainder of this essay, is just how popular Cyborgism will prove to be.

Kurzweil is an inveterate optimist (to the point that he is now publicly defending himself in the media, saying that he is well aware of the potential hazards of the artilects’ rise). His guiding purpose is to better humanity’s quality of life through his inventions such as his handheld text-to-voice reader for the blind. No one questions the value of his inventions and I can only praise him in this regard. However, where Kurzweil is weak is in his inability to equitably weigh the pessimistic with the optimistic consequences of the rise of the artilect that seems inevitable this century.

It is not enough to be optimistic. Optimism is fine if it doesn’t conflict with realism. When it does, taking a “Pollyanna” view of the world will only get you into trouble with the cold-eyed political realists who have so much experience of the past horrors of humanity. For example, the 20th century was the bloodiest in history, killing about 200-300 million people for political reasons such as wars, genocides, purges and ethnic cleansings.

So, taking a cold-eyed look at the rise of the cyborgs and the artilects, what do I think will most likely happen?

I do think there will be cyborgs, probably billions of them. In the early stages, Kurzweil may be right. There may be very few pure Terrans left. Nearly everyone will be adding memory enhancers to their brains so that they can, for example, learn a language in a day and be able to look up facts from a huge nanoscale database in their heads. This view is widely held by many people in the species- dominance community. But it is at this moment in history that the problems really start.

Humanness Destroyed

If everyone modifies himself or herself in the same way at the same speed, then hypothetically, the whole of (post) humanity could march in lockstep into an artilectual future without any real problem. But that is totally unrealistic. What is far more likely is that some people will cyborg themselves fast and heavily, while others do so more slowly and more moderately. It’s also virtually certain that there will be a wide variety of ways to cyborg yourself, offered by a slew of different cyborging companies.

This will lead to what I call the phenomenon of “cyborgian divergence.” There will be a huge variety of quasi-humans in the environment among families, couples and friends. Early cyborgs will then wake up viscerally to the fact that traditional humanness is being destroyed and that the emotional price is extreme, causing major alarm bells to go off.

For example, I consider it likely that the artilectual components being added initially to human brains will allow significant memory enhancement with not much, if any, intelligence increase. This may have unanticipated side effects, such that personality and behavioral changes occur, enough to make people feel that they have lost their friends or loved ones.

There is ample scope for Murphy’s Law (if something can go wrong, it will) to operate during this historical period of “cyborgian divergence.”

As more and more early cyborgs begin to wake up to the huge emotional and human cost they are paying, they will learn to value humanness a lot more and will start to make strategic political decisions.

I believe it will take time for the cyborgian components added to people’s brains to move up from being quantitatively superior (faster) to being qualitatively superior (allowing higher intelligence). It will take several decades at least for neuroscience to attain a quasi-full understanding of the neural nature of human intelligence. Cyborgism could be operating several decades before such full understanding is attained and incorporated into cyborgian components.

This gradual increase in qualitative capability will allow the Terran-inclined early cyborgs to keep intellectually competitive with the non Terran-inclined cyborgs.

I see these Terrans (or early Terran cyborgs) arguing now along my traditional lines. They will choose to remain essentially human, feel a visceral rejection to what they see happening and organize politically.

They will be fully aware that time is not on their side. If they wish to remove the risk that they will be superseded by a growing tide of artilects and artilect-like cyborgs, they will have to organize quickly and strike first, purging the cyborgs, the artilects and the cosmists so that the existential threat of humans being wiped out in the future is removed.

Surveys on Species Dominance

Predicting how the mix of Terrans, Cosmists, Cyborgists, and artilects will interact with each other will be complicated, especially as views on species dominance begin to polarize. It would therefore be helpful to be able to work with some real opinion data on this issue; this is something professional sociologists can do.

I think opinion pollsters should start making regular polls on the question of species dominance. Since it is this year (2011) that the issue of species dominance is going mainstream in the US media, the general public can begin to think about where they stand in the Terran/Cosmist spectrum, giving fairly informed opinions to the pollsters.

Once you have the data, then more realistic policy decisions can be  made by the strategists and intellectuals of the various competing parties.

Some early surveys have already been made, and the results are interesting. I know from the lectures I’ve given over the past two decades on species dominance that when I invite my audiences to vote on whether they are more Terran than Cosmist, the result is usually 50-50.

At first, I thought this was a consequence of the fact that the species dominance issue is too new, causing people who don’t really understand it to vote almost randomly – hence the 50:50 result. But gradually, it dawned on me that many people felt as ambivalently about the issue as I do. Typically, the Terran/Cosmist split would run from 40:60 to 60:40 (although I do notice that with my very young Chinese audiences in computer science, the Cosmists are at about 80%).

I can give two quasi-official poll results on the Cosmist/Terran split. One is by the BBC in Oct 2006, when the  general public was invited to vote between Kurzweil’s optimistic “merge” scenario (about 60%), and my “purge” scenario (about 40%). Another vote took place a year before on the popular US radio show “Coast to Coast.” At the end of my interview, listeners were invited to vote their preference, Terran or Cosmist, and the split was 55% Terran to 45% Cosmist.

This more or less 50:50 split will only make matters worse, I feel. If the split were 10:90, or 90:10, then one group could wipe out the other if it came to a war. Humanity would not be nearly as traumatized as with a 50:50 situation. The 50:50 split, if it is maintained, could not be worse. It shows how profoundly divisive this species dominance issue is, only increasing passions and the size of the final horror – a gigadeath-scale artilect war.