The Speculist: Atlas Hugged

logo.jpg

Live to see it.


« FastForward Radio -- Achieving Friendly Artificial Intelligence | Main | Friday Videos -- Apollo 11 Anniversary Edition »


Atlas Hugged

Speculist blogger Stephen Gordon has written a very interesting essay on the Atlas Shrugged phenomenon, exploring whether Ayn Rand's novel is, in some sense, coming true in our world today. He decided to publish this piece on a different blog because of the political nature of what he wrote. But, hey, I'm linking to it from here because anything Stephen writes is worth taking the time to read.

As I have noted before, Atlas Shrugged is essentially a science fiction story. It was set in the near future at the time it was written, a world we would now consider a past-future. The plot relies on technologies that didn't exist at the time the book was written -- Reardon metal, recovery of oil from exhausted wells, the torture gizmo the bad guys use on John Galt, etc. Rand also implies that in the future the US government has been restructured, referring repeatedly to a monocameral "national legislature."

However, in Rand's fiction, the story does not arise from technological or historical developments. It is driven by philosophy. So a discussion of whether her novel is being realized in the world today naturally turns into a discussion of political philosophy, which is not Speculist material. It's interesting to note, however, that Stephen describes, in addition to the political / philosophical form, an emergent, economic form of Atlas shrugging that is somewhat orthogonal to Rand's concept. In that model, it is a different Atlas who shrugs.

Likewise, I think there are some other "Atlas Shrugs" scenarios that are well within the realm of the topics we explore here. For example Stephen writes:

Welfare recipients have two barriers between themselves and a better lifestyle. They have the first natural barrier that all people face - they have to find the energy and ambition to work harder, or get an education to work for better money. Recipients also have an artificial barrier – they would lose the largess that is making their lives fairly comfortable. A marginal improvement in their productivity could actually result in a net loss of income. So it would take a significant improvement of their productivity before they'd see any benefit to their lifestyle at all. That's a bigger obstacle to productivity than some people can overcome. So they, quite rationally, work less than they would have otherwise.

Forget the politics of welfare for a moment. What's interesting to me here is the notion that some individuals do better economically when all productive activity is outsourced. Is this a social problem that needs to be remedied or just a sneak preview of the future economic life of all of us?

Over the past couple of centuries, human economic productivity has increased in unprecedented ways, deriving from "outsourcing" of productive labor to machinery. Since the machines have, up to now, mostly needed human beings to operate them, it wasn't always clear that outsourcing was taking place. But the machines keep getting smarter, and working their way further and further up the management ladder.

The digital revolution is only now truly being felt in the productive sectors of the economy. Things get very interesting as the complexity of intelligence embedded into the machinery of production continues to grow. We could be 10-20 years away from the equivalent of a human intelligence falling below Kurzweil's magical $1000 price point. At that point, EVERY activity currently performed by human beings could reasonably be done more cheaply and efficiently by a computer -- assuming that it is possible, legal, ethical, etc. to "buy" that capability for that price.

At that point, Atlas might shrug again, this time when the truly productive sector of the economy excises its least contributing part: humans. You may think these are completely different circumstances, but maybe not. In situation A, producers reject the high cost of doing business brought about by taxation and stop producing for ungrateful consumers. In situation B, producers reject the burden of pretending that the (perhaps perfectly grateful) consumers need to be part of the production cycle at all.

Then what happens? We ALL become welfare recipients? For that to work very well, we might have to outsource the running of government to the machines along with everything else. If that image is too scary -- and it will probably seem a lot more rational when the time comes, seeing as we will have witnessed human-or-greater artificial intelligences in action at that point,rather than trying to imagine having our world be run by some massive IVR system -- consider the alternative in which each of us owns our own outsourced means of production in the form of a universal assembler. Either way, you have the same result -- a world in which none of us has any more motivation to be "productive" than Stephen's welfare recipients.

What do we do then? Take your pick. I'll probably spend a lot of time blogging and podcasting, or rather the future equivalents of those activities. Also, I'll probably do a lot of traveling -- on earth, in space, and in virtual worlds. Later on, if we get tired of amusing ourselves and decide we want to become productive again, we'll have to think about massive cognitive enhancements to enable us to operate at the level of our robot overlords. In a world run by them, true productivity will have to do with increasing knowledge, increasing capability, and creating beauty, and those activities will be occurring at a level that it is impossible for us to imagine in this world.

Most likely, those of us who want to remain productive on those terms will just join up with the machines. It will be the last chapter of the story: Atlas Embraces.

UPDATE: An anonymous reader points out that that last line should really be "Atlas Hugged." Now how did I miss that? Thanks for the new title, reader!

Note: I need hardly mention that if you want to leave a comment about how "liberals" or "wingnuts" or the president or the obstructionist republicans are ruining the world, this is not a blog where we talk about those things. I believe Stephen is accepting comments over at The Last Pragmatist, however.

Comments

Phil, you're making a common mistake about leadership: intelligence is not the same thing as wisdom and judgement. Just because a machine is smarter than me, doesn't mean it has better judgement or it is better suited to run the world than I am (not that I am). As a new parent you've not been confronted with this problem just yet, but soon enough you'll be confronted with a creature that is likely as smart or smarter than you who is in no way ready to run your life. AI is nice, but it by itself won't suffice to run/govern human affairs.

Kirk --

Well, first off, my older child is 20 so I'm in the later stages of being a new parent.

Any person (organic or machine) who has greater intelligence than I has a greater capacity for wisdom, judgment, leadership, etc. than I do. That person may lack experience initially, but thinking a million times faster than we do, ought to be able to get caught up pretty fast.

Also, to get back to the idea of addition by subtraction, I'd like to see a candidate for office who doesn't suffer from excessive self-regard and who won't be tempted to sell out to special interests -- having had such proclivities edited out or, better yet, who never had them in the first place. Edit out the love of power and you might have a government run by folks (electronic folks, that is) truly capable of behaving as public servants.

Touche.

I like the addition by subtraction idea (good luck getting any pol who isn't in love with themselves/power), but without the feedback loop of significant consequences actively impacting one's physical comfort, I don't know if good judgement from a machine is possible. The ability to think "faster" is not necessarily "better."

Further, I disagree that anyone who has greater intelligence also has the capacity for greater wisdom and judgement. Some people have no learning curves; additional IQ points aren't going to change that.

I should point out that I love the blog and my quibbling with your latest is in no way a reflection on my appreciation otherwise.

I thought the point of Galt's New Order was that the producers did only what they wanted to do and gave their trust and respect only to those who they deemed worthy. In such an economy 'value' is determined by the producer rather than the consumer. In a near-singularity environment where most needs are met by machines, the ability to produce new entertainment may be the only skill that matters. Whether blogging, dungeon raiding, digital sculpting (from visual arts to software engineering) or simply generating clickstreams - we will either be producers or we will be lost in irrelevance.

"In a world run by them, true productivity will have to do with increasing knowledge, increasing capability, and creating beauty."

How do you figure?

Ben --

To my mind, those three things are what's left for us to produce if the machines produce all physical goods. I think MikeD's warning about being lost in irrelevance is right on the money. If we can continue to contribute in those areas, we continue to matter.

We're already in the early stages of outsourcing all boring, repetitive, dangerous work. I estimated years ago that 90 % of the actual physical labor involved in producing America's GDP was actually carried out by machines. My estimate now is that the % keeps rising.

I agree with Ben's assessment. The thought and creativity of whatever is produced will be the human end of add-on value...at least until we merge with our machines.

Except for the few non-existent technologies, Atlas Shrugged had an amazingly "Fortyish" flavor. That's when Rand began writing notebooks, planning her huge novel.

Transportation was utterly dependent on railroads, there was no television to speak of, nor computers (even the vacuum tube ones that actually existed in the Fifties), no interstates, no jet airlines, and no Elvis!!!

I was 7 years old when the Fifties ended. I simply didn't recognize anything in that novel that I'd recognize from the child-eye view.

It is a genuinely moving novel from the political philosophical vantage point (which I won't get into), but technologically, it was outdated before Rand got done filling out her notebooks.

I will take a chance on edging into political territory by point out one of her economic faux pas, the idea that everyone who really counted as genuine creatives in America could get together at a small, intimate dinner party at Galt's Gulch in Colorado is insane. You'd need at least all the seating in the Denver Broncos' stadium to stuff 'em all in. Actually, a lot more than that.

You're assuming that populations of AI wouldn't in of themselves develop factions/subspecies of some kind. Since everyone talks of AI potentially being on an extremely fast evolutionary cycle one would think that such differentiations among different AI would be inevitable. In such a circumstance, self-interested control over scarce resources may indeed develop as a trait of any AI programmed to survive.

The social and cultural aspects of a Better Living Through AI Automation world have been explored in James P. Hogan's novel Voyage from Yesteryear.

You might want to read 'Sleepless' by Nancy Kress. In it, the uber-capable who had enough money to buy the Sleepless mod have become the runners of the robot factories which provide the rest of the population what they want--so everyone else is a 17th century aristocrat. There is government because people have differing desires.

Its been a long time since I read this book so forgive me if I got some details wrong.

Post a comment

(Comments are moderated, and sometimes they take a while to appear. Thanks for waiting.)






Be a Speculist

Share your thoughts on the future with more than

70,000

Speculist readers. Write to us at:

speculist1@yahoo.com

(More details here.)



Blogroll



Categories

Powered by
Movable Type 3.2