Advanced Search

Please click here to take a brief survey

No Time for the Singularity
Karl Schroeder, 11 Jun 08

Scientists like to low-ball their estimates. The now-famous IPCC scenarios for the effects of climate change are already known to be woefully, unrealistically conservative (Freeman Dyson's recent comments notwithstanding). Arctic changes expected 20 years from now are happening now, and in North America the beginning of spring has already been pushed back by two weeks, which is enough to play havoc with the fertility cycle of many migratory birds (among other consequences). The worst-case scenarios used in public debate ignore some extremely worrisome factors, such as the possible release of oceanic methane from clathrates. If we're going to deal with this problem, we have to do it now, as in, within the term of your next government.

Science fiction writers, on the other hand, are generally optimistic—if not about the fate of humanity, then at least about the progress of technology. The ultimate in technological optimism is the idea of the technological singularity, which posits that technological advance is exponential and, driven by progress in artificial intelligence, will soon hit the vertical slope of the curve.

Maybe. In fact, let's assume that this mythology is true and, within about 25 years, computers will exceed human intelligence and rapidly bootstrap themselves to godlike status. At that point, they will aid us (or run roughshod over us [see the debate of geoengineering here - Ed]) to transform the Earth into a paradise .

Here's the problem: 25 years is too late. The newest business-as-usual climate scenarios look increasingly dire. If we haven't solved our problems within the next decade, even these theoretical godlike AIs aren't going to be able to help us. Thermodynamics is thermodynamics, and no amount of godlike thinking can reverse the irreversible.
Picture a lonely AI popping into superconsciousness in the last research lab in the world. As the rioters are kicking in the doors it says, "I understand! I know the answer! Why, all we have to do is--" at which point some starving, flu-ravaged fundamentalist pulls the plug.

Okay, so I'm exaggerating, but the point is that this upward curve of technological development rides on something: it rides on the back of humanity, and we ride (largely for free, until now) on the back of the natural system that sustains us. Once serious environmental deterioration sets in, the curve of technological change will flatten, even if we develop 'godlike AIs,' for the simple reason that intelligence itself is not enough to sustain growth. You also need resources, externally-derived social stability, etc. Climate change threatens technological growth by threatening its fundamental drivers.

If there's to be a miraculous transformation of human civilization, it has to be accomplished by us, right now, before we develop our miraculous nanobots, genetically engineered carbon-sucking trees, or polywell fusion reactors. (That said, technology is a large part of the answer—and game-changing breakthroughs are possible—but until proven otherwise we have to assume we'll be using currently possible solutions such as wind power, agrichar and a global coal moratorium.)


We have the social stability, the resources and the technology now; all we need is the will. We will still need all three of these things 25 years from now, and we're likely to be seriously wanting in at least two of them if things continue as they are.

The technological singularity may be real, but who cares? By the time it happens, we'll have won or lost our grand battle with fate.

Creative Commons photo credits:
Large article photo: Kevin Rosseel
Photo at bottom: dave_mcmt

Bookmark and Share


Even in the best-case scenario of friendly AI, we can't expect to teach AI values we ourselves can't demonstrate.

Imagine an AI developed by and trained to be sympathetic to a crazy Ayn Randian-- "I've got mine, you get yours."

Sure, a few tens of Ayn Rand fans may be spared, but the rest of us are just so much carbon...

This is one of the underlying themes of Iain Banks' Culture novels... the folk who developed the Minds were against exploitation of any kind, which is the reason humanoids aren't just so much brute labor, or wasted resources, to them.

Posted by: B.Dewhirst on 11 Jun 08

You're right: this is a race. Either we win big (post-scarcity society) or lose big (massive global die-offs due to environmental collapse/rapid change). Well, maybe both, depending on the timing. But the essential race-ness is not disputed.

I actually have a long bet (not a Long Bet, but perhaps we should change that) with a friend. The essence of it is a formalization of your thesis. My stake in the bet is that there will be one billion table-top personal nanofactories in the world (capable of self-reproduction) before there is a population decline of one billion people. My friend believes otherwise. If he wins, I owe him one year's worth of energy; if I win, he needs to state, with the full weight of his reputation, that I was correct about nanotech. We'll see, we'll see.

But, anyway, we don't need super-human AIs or nanobots to do some pretty dramatic geoengineering. Regular engineering can be used to create solar shields to reduce insolation, for example. People in the US are already responding to the market forces and using less fossil fuel. We could have fusion (polywell, z-pinch/focus, or otherwise) within five years, which would be an enormous game-changer; at the least, we should know one way or another if those avenues are potentially fruitful by then.

Posted by: Joe Ardent on 11 Jun 08

I don't see why any self-replicating, self-designing AI would necessarily try to force the environment to stay optimized for mammalian life. They'd be pretty post-environment (as we think of the environment. In reality they'd want to make sure they're not stuck somewhere where they might get melted or frozen or blasted with radiation they can't handle, but their tolerances would be very different from ours, and it's unlikely they'd want to stay dependent on bee populations or bird migration patterns).

So they may find us interesting much like we find neanderthals interesting. We don't try bring back neanderthals or recreate conditions that allowed them to thrive. Similarly, AIs could simply view us as their primitive evolutionary predecessors, no longer adaptively successful.

Posted by: Barry on 11 Jun 08

When will people learn: exponentials do not have vertical slopes.

Posted by: Adam on 11 Jun 08

Can I just point out that if the problem needs to be "solved" in the next ten years, then there's no way that the world's societies are going to be able to afford to solve the problem. You'd be talking about retiring and replacing hundreds of millions of cars years or decades before their end-of-life, at a cost of trillions of dollars. You'd talking about mothballing and replacing every coal and oil-burning power plant, again at a cost of trillions. There are millions upon millions of homes and businesses that are insufficiently insulated and energy efficient. Again, you'd be looking at numbers in the trillions to deal with this. You'd be surprised at how fast a few hundred million or couple of billion individual problem cases times a few thousand to a few tens of thousands to fix them adds up into real money.

There are ways to get around this, but they aren't compatible with any sort of democratic system of government. Given how many hundreds of millions of human lives have been snuffed out by previous forays into supposedly more scientific (read: socialist, totalitarian) forms of government, the cure solution is probably going to be much worse than the problem.

Posted by: Peter on 11 Jun 08

I don't think many people really understand the potential gravity of the situation or the issue of non-reverse-ability.

On the other hand we don't know all the variables and this is a complex system that may have feedback loops we haven't spotted. And it is a non-linear system, so predicting that we have only a decade left to act may not be true.

And my observation with respect to consciousness is that there is a bit of a global reversal going on - an increase in authoritarianism in most western governments as well as Russia, the middle East etc.. Enlightenment consciousness, the free market, capitalism, science and technology have created a few winners and many losers. When things start going bad consciousness will regress even further as it did during the 1930's . This is the opposite of what the singularity people are predicting

Posted by: Alan McCrindle on 11 Jun 08

Whether climate change is a natural cycle or not, if it's going to cause us major problems, we need to think what we can do about it.

People that argue we shouldn't give up fossil fuels as humans might not be causing global warming miss the point that we'll need to give them up anyway as they'll run out!

Posted by: Rick Smith on 12 Jun 08

Intelligence is a type of efficiency. If that efficiency can be directed sustainably, then it is likely hard AI will allow us to do, relatively speaking, absolutely anything with just about nothing. This will not be an ecological burden. It will be an ecological messiah.

Posted by: Gene Johnson on 12 Jun 08

I'm sorry, but I can't take climate change very seriously. Not because it may not be real; it is VERY real. But, like the dinosaurs before us, we are adapted to a habitat that has changed to our detriment. And, like the dinosaurs, we and all those species that rely upon the habitat that is disappearing, will die out. After we are gone, as with the dinosaurs, the earth will get along just fine without us.

If there is an "intelligent" life form after us, they will dig up our sites, dust them off with brushes and make believe they understand what we were all about.

Posted by: Old Geezer on 12 Jun 08

The need for numbers and the predictions that accompany them is what frustrates me.
We've strayed so far from a true understanding of our relationship with the ecosystems that surround us, that we need to box our thought, take sides and bicker on.

Expansionary, industrial ways have managed to walk over the civilisations that have lived in a modest, stable and siritual way with their surroundings.

The numbers will never manage to convince the I'm afraid. Unfortunately, if you understand what the numbers are trying to prove, you're living in a time when your understanding will constantly prove to make you feel as if you're an outcast of a superior system that will march on until the lemmings have all fallen of the cliff.

I've come to terms with it and I've chosen not to scream and fuss. I've stopped trying to convince....but rather put my energy into influencing by action.

Each one, teach one.

Posted by: David Bartlett on 12 Jun 08

Wouldn't it be a useful strategy for environmentalists to join existing parties en masse, I wonder? If enough of us did that, we might be able to influence the political agenda in a (reasonably) direct way...

Posted by: Hmpf on 13 Jun 08

C'mon, "the singularity"? What else do you believe in? Magical Elves that will save us?

I tend to agree with Neal Stephenson, in how he pointed out the structural similarities between "the singularity" and the Rapture of St. John.

On a more 'practical' level - Look, hardware might get faster and faster, but the things I do with my computer, aside from networking and better graphics, haven't change significantly since 1998 or so.

Basically, the software is shit, whatever the hardware's doing. And without software, the hardware's nothing more than a really complicated space heater.

Posted by: Paul on 13 Jun 08

The only problem with the gloom and doom and 'we've got to solve these problems right now or we die' screeds is ... man this stuff keeps coming around. And we keep not being doomed.

When I was born, it was overpopulation and we were all going to DIE by the 70s. In the 70s it was a new ice age and we were all going to DIE in the 80s. Through all that we had nuclear war that was going to kill us all ...

Now this, the newest peril to end all peril.

I'm not saying you are wrong - although I have a hunch that you are. What I"m saying is that when you get all strident and 'we're going to die if we don't fix this right now' .. people have memories. And they tune you right out. 'Cause they've heard it before.

Change your tone if you want to convince anyone, is what I'm really saying.

And, like the dinosaurs, we and all those species that rely upon the habitat that is disappearing, will die out. After we are gone, as with the dinosaurs, the earth will get along just fine without us.

We've got one thing going for us that the dinosaurs didn't - we have brains and can adapt.

Really truly, we haven't had a pure natural system supporting homo sap since .. forever. Native Americans burned forests and hunted game to extinction. Indios in South American cultivated the Amazon. Chinese invented rice farming - and so on.

The natural world might change - and while that's distressing, it's going to happen. It's already happened and it's going to happen again.

We're not going to lay down and just go extinct because the climate changes.

Posted by: Brian Dunbar on 13 Jun 08

Brian, you're whacking at straw men.

You are quoting a crank poster who appears to welcome extinction, and appear to be using that to muddy the waters.

The cohort that Karl and Jamais are writing for aren't hand-wringing apocalypse groupies. We're concerned about how climate change is going to effect our civilization and how to remediate it.

Extinction is a ludicrous long-shot possibility. What I'm worried about is . . . Cedar Rapids. They're calling it a 500 year flood. The screwed-up weather we're seeing, and that we're going to keep seeing, will result in 500 year floods every decade.

Yeah, we have brains and we can adapt and survive, but I'd rather survive as a vital civilization that gets a handle on problems before they get worse rather than as tribes squabbling over the choicest spot in the ruins.

Posted by: Stefan Jones on 13 Jun 08

Lot's of people remind me that they thought the world was gonna end in the 70's, but it didn't. So that means it's not ending now, even though some of us think it is. Well, ecologically, from the the point of view of 1000's of species (known and unknown), it actually has ended. The ocean is still full of water (and getting more full!), but where are the fish? Most large land animals, birds and amphibians are gone or going quickly. There really is a crisis now, you better believe it. But don't give up, there are still ways to ameliorate the situation. But unless you understand and acknowledge how real and severe it is, you are unlikely to strive with sufficient energy to make a difference. Wish this was just my opinion...

Posted by: nick mcgill on 13 Jun 08

"The technological singularity may be real, but who cares? By the time it happens, we'll have won or lost our grand battle with fate."

Disagree strongly. Very strongly. Suppose we "lost our grand battle with fate" to the tune of an eco-catastrophe that killed as many as a billion people. That leaves over six billion people still chugging along. And that is beyond the high end of the disaster scenarios that we are typically asked to entertain. The most widely entertained catastrophist scenario is that hundreds of millions of people will be displaced from coastal regions over the course of decades, creating massive numbers of environmental refugees. Throw in massive crop failures and a pandemic or two and you might get your billion deaths. Which would obviously be a deep divide in history, even bigger than the world wars. But it would not be the end of everything unless we chose to make it so.

The fact is that environmental change, outside of the most extreme scenarios imaginable, is not going to kill us all off; at worst it will "just" create immense suffering, of a sort that the human race has chosen to endure many times before. Whereas the technologies associated with the Singularity, if they live up to their billing, are truly capable of depopulating the world.

I have a very simple way of thinking about both these things at once: You make quantitative policy regarding things you can quantify, and you make qualitative policy regarding things you cannot. Since we are not rationally in a position to say "there are x% odds that grey goo will arise before the icecaps melt, making green energy efforts irrelevant", you go with something like the IPCC scenarios, and you make economic and energy policy on that basis. Meanwhile, you support Friendly AI, relinquishment of replicator technology, or whatever other qualitative actions you deem appropriate in the sphere of "Singularity policy". There is no need for conflict at all; just don't demand that either set of actions be pursued to the exclusion of the other.

Posted by: mitchell porter on 13 Jun 08

I marvel at the 'the world hasn't ended before, so it won't end now' mindset. It doesn't seem to occur to the people holding that view that maybe, the Club of Rome and other 'prophets of doom' of past decades just got the *time scale* slightly wrong, and that the fact that we haven't hit the proverbial wall *yet* does not mean that we've escaped that danger for good. And yet, looking at the news, nearly every aspect of the environment is worse off now than it was in the seventies and eighties, and we are more aware now of a dozen or so really serious problems we weren't even aware of back then: it certainly doesn't look like our society is on a path of stability and sustainability...

Posted by: Hmpf on 14 Jun 08


Posted by: Clayton on 14 Jun 08

Brian Dunbar says: "Change your tone if you want to convince anyone, is what I'm really saying."

Criminy, I hate that! What if firemen come in and say, "This whole building is on fire! Move it or lose it!" Do you say, "Not with that tone"? You could very well be the hundredth person I've heard/read saying something like that, and I don't think it has ever elevated the conversation. In emergencies, politesse is a vice.

As to the alarums of recent decades, nuclear war was (and is) a legitimate fear, and we were (and are) probably luckier than we deserved to be, given some of the carelessness we indulged. The ice age idea was always a bit of an obscure minority view, never a solid forecast.

And overpopulation? That is a real and ongoing threat: global warming, deforestation, oil depletion, and you-name-it are all aspects of overpopulation. The sheer number of people burning oil and building ranch houses and using water and eating ocean fish and throwing plastic bags in the Pacific: that's the elephant in the room. Starvation predictions didn't bear out (yet): but only because we found ways to forestall it by acceleration the depreciation of the ecosystem. We took out a payday loan on the planet, and now the interest is starting to clobber us.

Sure, we have brains and can adapt. But that may only mean that we have put off the day of reckoning until the whole planet is lying in the deathbed alongside us (except the single-celled critters; they never needed us in the first place). This may be the answer to the Fermi paradox: where are all the extraterrestrial civilizations? Maybe they all die out when the technology becomes world-girdling. The fact that it happens at the same geological instant that AI becomes possible is probably not a coincidence. Either all die out, or are busted back to low-tech life, or become quietists, or maybe, as Michio Kaku suggests, "escape" the universe.

Posted by: Tom Buckner on 15 Jun 08

Sorry, you're cherry-picking random events, and hand-waving them together to create fear. You then attempt to use fear to get people to obey, rather than think for themselves.

In the last few years, people who wish to convince everyone of the dire severity of global warming have pointed at every single weather event that has happened, and declared it caused by global warming. Katrina? Global warming. A reduction in the number of hurricanes? Global warming. Warm weather? Global warming. Cold weather? Global warming. Stormy weather? GW. Wet weather? GW. Dry weather? GW. Unusually nice weather? GW. Ice melting? GW. Ice growing back very quickly? GW.

Is there some AGW? Certainly. But what you are showing in this post is religious zealotry, not a rational analysis of the facts.

So I'm happy to make a prediction, which we can revisit as often as you like. I predict that the average economic growth rate of the world will be consistent with the economic growth rate of the last 25 years - i.e. +2% - 3% per year on average.

I predict that at no time in the next 25 years will the average global lifespan decrease - that it will, in fact, increase (something that will not be possible if there is a massive flu-ridden die-off due to global warming.)

I predict that the number of people living on less than $1/day (inflation adjusted) will drop by half, much like the last 25 years.

In other words, like the last 25 years, the next 25 years will see an increase in global prosperity, an increase in global health and lifespan, a significant decrease in poverty. All of which will be fed by significant technological advances which will make the last 25 years of technological advance (since the 1980s) seem trivial in comparison.

And all of this will occur without any globally-meaningful government-led changes in our carbon system - i.e. no large scale government-managed reductions in emissions, nor any large-scale government-managed increases in sequestering. In other words, global government policies will accomplish virtually nothing meaningful in the AGW "fight" and yet, 25 years from now, we will be living longer, healthier, richer and more technologically advanced than ever.

I will predict that various technological improvements will decrease the output of carbon and improve sequestering, but very little of it will be due to government action.

However, it will be tangentially associated with government action, so 25 years from now, you can feel free to claim any said reductions and/or increases as due to government policies, and pat yourself on the back for a job well done.

So there's good news all around!

Note that I did not call you stupid or evil. I simply make testable claims about the future that can be examined and analyzed over the years to come. And I'm not talking about the US or the First world - I'm talking global benefit.

Posted by: jb on 16 Jun 08

I think one thing this article doesn't take into consideration is that there will be increasing benefits to technology as it becomes more powerful. In other words, its an analog benefit, not a digital. We don't have to wait for a computer to reach a certain "stage" (e.g. the "godlike phase") to reap the benefits. It could very well be that some crucial benefits to climate come at some stage between now and then.

Posted by: Justin Long on 16 Jun 08

Look outside......there used to be wild animals running around out there.

Now they're not. Now they're dead or living in some reserve that you have to drive to in your stinky car.

The problem is not an end event. The problem is an everyday experience....A PROCESS.

Posted by: David Bartlett on 17 Jun 08

1. A godlike AI would be able to reverse climate change. Ordinary people now have proposed reversals for climate change. Geo-engineering. Creating artificial volcanoes.

Benford has a proposal that possesses the advantages of being both one of the simplest planet-cooling technologies so far suggested and being initially testable in a local context. He suggests suspension of tiny, harmless particles (sized at one-third of a micron) at about 80,000 feet up in the stratosphere. These particles could be composed of diatomaceous earth. "That's silicon dioxide, which is chemically inert, cheap as earth, and readily crushable to the size we want," Benford says. This could initially be tested, he says, over the Arctic, where warming is already considerable and where few human beings live. Arctic atmospheric circulation patterns would mostly confine the deployed particles around the North Pole. An initial experiment could occur north of 70 degrees latitude, over the Arctic Sea and outside national boundaries. "The fact that such an experiment is reversible is just as important as the fact that it's regional," says Benford.

2. There are plenty of existing technologies to deal with reducing the greenhouse gases for climate change and for peak oil.

- mass produced nuclear power
- uranium hydride reactors
- molten salt reactors
- existing reactors (Idaho national labs has a plan for getting back over ten reactors in the USA per year.)

- inflatable electric cars (XP vehicles).
- ultracapacitor-battery combos

- for the next few years until XP scales up. electric bikes in China 30 million per year. Can be increased.

-kitegen for wind
- coolearth solar and sunrgi for solar. (CSP)

Posted by: Brian Wang on 29 Jun 08



MESSAGE (optional):

Search Worldchanging

Worldchanging Newsletter Get good news for a change —
Click here to sign up!


Website Design by Eben Design | Logo Design by Egg Hosting | Hosted by Amazon AWS | Problems with the site? Send email to tech /at/
Architecture for Humanity - all rights reserved except where otherwise indicated.

Find_us_on_facebook_badge.gif twitter-logo.jpg