(Michael Anissimov, Glenn Zorpette.)
First, may I welcome Mr. Zorpette to the wild and wacky world of debating the techological singularity! May these discussions be exciting and illuminating to you as they are boring and repetitive to me. (They’re not boring because of your comments, Mr. Zorpette, but merely from my stubborn insistence and decade-plus quest for understanding all the common objections and enthusiasms surrounding the so-called Singularity, despite their ever-repeated frequency.)
The IEEE Spectrum special issue on the Singularity is opened with a critical introduction by acclaimed technology journalist Glenn Zorpette, “Waiting for the Rapture” with the tagline, “Technological convergence will change our lives but won’t make them indefinitely long“, which represents his number one qualm with Singularity discussions, but a relatively minor component thereof: radical life extension. My initial comment would be, “I came for Singularity criticism, but got radical life extension criticism. Can you please write a new article that criticizes the feasibility of smarter-than-human intelligence, and makes criticism of life extension a side note, or submit this article to another magazine that addresses the life extension issue?”
The opening paragraph begins as follows:
“Across cultures, classes, and aeons, people have yearned to transcend death.”
Yeah, we totally have. And not in an exclusively irrational way either. Since the average lifespan back in the day was about 30, and now it’s about 80, I’d say we’ve come a long way.
He continues, “Bear that history in mind as you consider the creed of the singularitarians. Many of them fervently believe that in the next several decades we’ll have computers into which you’ll be able to upload your consciousness—the mysterious thing that makes you you. Then, with your consciousness able to go from mechanical body to mechanical body, or virtual paradise to virtual paradise, you’ll never need to face death, illness, bad food, or poor cellphone reception.”
1. No unified creed. Many singularitarians have different positions on the 22 concepts I listed. However, I do identify as a “singularitarian”, even though the word has lost most meaning, partially because it gives me an opportunity to respond to many of the arguments lobbed in my general direction.
2. Conscious is no more mysterious than life, the planets, etc. Mysteriousness is in the mind, not reality. All of these things seem mysterious until we actually begin to understand them through scientific inquiry. You know that, so why not drop the word “mysterious”?
3. Yes, if the mind is really just what the brain does, and the brain’s functionality can be duplicated in other media (carbon nanocomputers, etc.), then indeed, we will eventually have a future where we “go from mechanical body to mechanical body, or virtual paradise to virtual paradise, you’ll never need to face death, illness, bad food, or poor cellphone reception”. It’s just as foolish to dismiss the plausibility of this possible future scenario for its shock value as it is to embrace it for its superficial “geek rapture value”.
On to the second paragraph. The phrase “rapture of the geeks” is immediately invoked. As my colleague Steven writes, “that image of a shared psychological flaw is itself so seductive that it has distorted people’s view of what the singularity is about into a kind of geek-bible-wielding strawman — singularitarian ideas are assumed to parallel fundamentalist Christian ideas even where they don’t, just because the comparison is apparently so much fun.”
Singularitarian ideas are portrayed as full of shit because Rapture-believers are full of shit. But this guilt by association is unfair. Were the Wright brothers idiotic because they aspired toward controlled flight, and controlled flight had (allegedly) previously been the province of angels? Were the early nuclear engineers idiotic because they aspired to harness the power of the Sun, which had henceforth (allegedly) only been harnessed by God? Were the first genetic engineers idiotic, because they aspired to modify the very code of life, which up to that point had only (allegedly) been done by God? Are AGI designers idiotic because they aspire towards creating general intelligence in a medium other than biology, whereas previously, general intelligence has only been found within biological structures? Are life extensionists idiotic because they seek to ameliorate the causes of aging enough to heal metabolic damage before it causes pathology, and begin this project immediately rather than in a century?
The definition of the Singularity Zorpette uses in his first pargraphs is the “intelligence explosion” introduced by I.J. Good and popularized by Yudkowsky, but this is one of the only mentions it gets in his article. Otherwise, he ignores this scenario and focuses on mind uploading and radical life extension, both of which he severely doubts. This preoccupation with criticizing life extension sets Zorpette alone from other authors in this issue. I wonder if he is aware that of the tens of thousands of people advocating an engineering approach to life extension, only a minority buy into the Singularity visions extolled by Kurzweil?
Zorpette fairly rags on Ray Kurzweil’s upcoming movie, The Singularity is Near, saying, “Without any apparent irony, the picture’s producers call it “a true story about the future”. While understanding the need to make compromises when it comes to marketing soundbites, I agree with Zorpette that this is a poor tagline for Kurzweil’s movie. Please change it, Mr. Raymond Kurzweil! Here are my alternate suggestions for taglines, which, if adopted now (it’s not too late!), might avoid a national media backlash:
“The next step in humanity’s journey.”
“The harmony of technology and biology.”
“When technology improves biology.”
“Where is humanity headed?”
“Is the power of technology exponentially increasing?”
“Can the human body be enhanced?”
These taglines are provocative without invoking futuristic determinism, which Mr. Kurzweil has been heavily criticized for in the past and will continue to be criticized for in the future. This is a weak point, an Achilles heel, that Kurzweil could do without. The thing is, many of Kurzweil’s arguments are strong enough without the deterministic component. Viewed as probabilistic arguments, they still carry plenty of weight. It’s just that as a matter of presentation and marketing, Kurzweil seems linked to the deterministic approach (although he softened it in his recent book), which can be discarded without too much harm (and substantial gain, in fact).
Zorpette writes, “There’s also a drumbeat of respectful and essentially credulous articles in the science press.” Yes, there are! This reality causes me to snicker when Singularity critics try to portray the ideas as fringe, when they have been considered by journalists in the top magazines of the country, some of which I have had the pleasure to talk to personally.
He then writes, “Why should a mere journalist question Kurzweil’s conclusion that some of us alive today will live indefinitely? Because we all know it’s wrong.” Buh. This is where Zorpette shows that his big problem is with radical life extensionists in general, not just “singularitarians”. How does he know that all cryonically suspended persons will never be revived, even in 1000 years? Or that we will not reach longevity escape velocity by 2070, when I will will happen to be “only” 86? In his tone, Zorpette appeals to bioconservative biases like those of Leon Kass. But even Kass takes the “danger” of indefinite life extension seriously, while thinkers like Zorpette do not.
The body is a biological machine, and like antique cars, it should be possible to arrange its indefinite upkeep. Not by eliminating the sources of aging, but merely cleaning up damage before it accumulates to the point of causing pathology. This is the mantra of the Methuselah Foundation, and it makes sense.
Zorpette then says, “The singularity debate is too rarely a real argument. There’s too much fixation on death avoidance.” The unfortunate thing about this statement is that it shows that Zorpette’s focus is most on point #21 than on the other (more complex and difficult) 22 points. It must be somewhat new and startling to him, because many of his objections focus on it. The “Singularity”, a messy meta-concept containing over a dozen constituent concepts, can serve as a lens to examine those that support or criticize it, because they immediate seize upon those sub-concepts that they most detest or support. For Zorpette, his major hangup is the near-term feasibility of anti-aging therapies that clean up metabolic damage faster than it accumulates. It’s a little unfortunate, because it seems easier to argue against, say, mind uploading, or hard takeoff superintelligence. Why make your task more difficult than it needs to be?
Zorpette then quickly shows his respect for the power of technology by writing, “in the coming years, as computers become stupendously powerful—really and truly ridiculously powerful—and as electronics and other technologies begin to enhance and fuse with biology, life really is going to get more interesting.” Here, he seems to embrace transhumanist ideas in the abstract, while rejected the alleged extremism of “singularitarianism”. I welcome Zorpette to read the Transhumanist FAQ and see if his own views can be described as transhumanist in nature.
He then says, about their selection of article writers, “with a few exceptions, we found people who are not on record as either embracing singularity dogma or rejecting it.” It’s funny how this statement oddly tries to sound neutral, while at the same time calling singularity “dogma”. To simplify matters, I’m assuming that Zorpette is mainly mirroring the opinion of his friend John Horgan, who may have thought about the issue more than Zorpette has.
Regarding the role of technology journalism, Zorpette is influenced by Horgan’s idea that a “cranky”, rather than deferential approach, to the career is appropriate. This is fine by me! I’d prefer a mix of deferential, neutral, and enthusiastic journalism, covering all bases. Fortunately (?) for singularitarians, most journalists we talk to are enthusiastic and positive about our ideas. Which is why I welcome explicitly negative coverage (and that’s what this is) as a change of pace and an opportunity to respond to many of the common arguments.
Next, Zorpette summarizes the authors and their contributions. He writes, “Vinge’s 1993 essay “The Coming Technological Singularity” that launched the modern singularity movement.” I would argue that this is wrong. Eliezer Yudkowsky’s 1996 essay “Staring into the Singularity” started the modern singularity movement. As a card-carrying Singularity “cultist”, I think I know what I am talking about.
Zorpette then writes, “That movement has evolved since then into an array of competing hypotheses and scenarios [for a rundown, see “Who’s Who in the Singularity,” in this issue]. But central to them all is the paradoxical yet weirdly compelling idea of a conscious machine. Arguably, no other technology-related concept resonates with such intellectual and philosophical force.” It’s rather odd, because I don’t think that the idea is central to all the competing hypotheses and scenarios. For instance, Kurzweil admits that machines may not actually be conscious the same way as us, and still have a huge impact on the world. Rather, the idea of smarter-than-human intelligence (not consciousness) is central to the Singularity. That’s why you can talk about the Singularity while talking about biological upgrades alone. Conscious robots need not apply. Zorpette misses this, I think, because his high-octane career (which I respect!) doesn’t give him enough time to really dig very deep when it comes to Singularity discussions online.
If you asked me whether I’d take the opinion of a bright layman who has read about the Singularity online as a hobby for three years, or a high-octane journalist who spent a few weeks researching it for his special issue, I think I’d take the former. No offense to Zorpette, but I know how most high-achievers lives work — work, work, work, relax, sleep, then repeat. Not really enough time to sit around and absorb all the subtle stuff. After all, no one really pays you for it. (I’m very impressed by high achievers, like Dr. Jones and others, who invest the time to read blogs like this one, which occasionally rant and segue into random topics, despite their very busy schedules.)
In his next paragraph, Zorpette embraces functionalism, and pretends that everyone does: “Consciousness seems mystical and inextricably linked to organisms. What happens in the cerebral cortex that turns objective information into subjective experience—that turns chemical and neuronal activity in the mouth and nose into the taste of watermelon? pressure waves into the sound of an oboe? We don’t know, but we will someday. No one argues that consciousness arises from anything but biological processes in the brain.” Well, unfortunately, they do, including some of the people who commented in your issue, like Stephen Pinker apparently. In fact, I think the strongest phalanx of opposition to Singularity ideas consists of those, including many so-called scientists, who believe that consciousness cannot be traced to brain signals, but rather some immaterial pixie dust/soul/aura. After all, if you believe in Christianity, Judaism, or personal watered-down versions thereof (like 80% of Americans), views like this are common currency.
He says, “The brain is nothing more, and nothing less, than a very powerful and very odd computer.” I agree, but try proclaiming this at your next IEEE staff meeting, or your personal blog, and witness the numerous objections. Most Singularity skepticism is motivated by exactly these objections.
To wrap up, Zorpette writes, “What we do know is that the brain’s complexity dwarfs anything we’ve managed to fully understand, let alone build. Koch, Tononi, and Brooks are all confident that consciousness will arise in a machine, but they are less sanguine about death-defying uploading, and especially about it happening in time to allow people alive now to preserve their minds in some sort of digitally created Eden.”
This is amusing. If consciousness can arise in a machine, then why could a machine not be made in the human image, and harbor human consciousness? What seems to be occurring here is a reversal of the appropriate order of analysis.
First, you look at the requirements, not the consequences. Requirements of mind uploading:
1) finite state nature of the human brain and mind
2) computers fast enough to run that finite state program
3) substrate-independence of conscious experience
4) interface devices between the program and the external world
Consequences of mind uploading if it is possible:
1) Near-immortality
2) Digital Eden
3) other stuff that sounds religious but are straightforward consequences if uploading is possible
Zorpette and others look at the consequences first, then evaluate the requirements in light of the stated consequences. But wait! This is premature. Ignore the consequences. Look at the requirements on their own merits. If the requirements are fulfilled, then no amount of squirming will free us from extreme consequences, like copying Stephen Hawking’s brain a thousand times and letting them loose on the Internet.
High-profile figures like Koch, Tononi, and Brooks have a tremendous interest to discuss possible requirements on their own merits, but specifically avoid conclusions that lead to anything like uploading. So it’s no wonder that many of them are sanguine about possibilities of uploading. Of course, there are high-profile figures like Marvin Minsky, who I’ve been fortunate enough to get significant one-on-one time with, who do embrace the technological requirements of uploading and its likely consequences. But these are the minority.
My final impression? This is a critical article that attacks life extension and uploading, while mostly ignoring the other 20 points I address as components of this useless “Singularity” word. As for my personal emotional impression of the article, I find it difficult to get excited either way, maybe because I believe the Singularity concept is already as popular as it really needs to be. What is needed is to enhance understanding among those already exposed, not necessarily expose it to a wider audience.
As stated before, however, I would have preferred if the article focused on a critique of the nearness of smarter-than-human intelligence, rather than a critique of (the rather intuitive and heavily-supported) position of “death avoidance”, or critiques of the consequences (rather than the requirements) of mind uploading.
Zorpette views the Singularity as a special case of old-as-history death avoidance, when in reality it is quite a new vision of the dynamics of self-improving and nonbiological intelligence entirely distinct from death avoidance. For instance, most “singularitarians” believe that the Singularity could kill us all as easily as it makes us live indefinitely. Only Ray Kurzweil, and a lesser-known transhumanist futurist, John Smart, have seemingly suggested that the Singularity is necessarily a good thing. Vernor Vinge, Eliezer Yudkowsky, and at least 100 other Singularity advocates I know personally would argue against that.
12 Responses »
