Rick Searle reviews *Average is Over*

The review is excellent and interesting throughout, here is one good bit:

Come to think of it, lack of intelligibility runs like a red thread throughout Average is Over, from “ugly” machine chess moves that human players scratch their heads at, to the fact that Cowen thinks those who will succeed in the next century will be those who place their “faith” in the decisions of machines, choices of action they themselves do not fully understand. Let’s hope he’s wrong on that score as well, for lack of intelligibility in human beings in politics, economics, and science, drives conspiracy theories, paranoia, and superstition, and political immobility.

Cowen believes the time when secular persons are able to cull from science a general, intelligible picture of the world is coming to a close. This would be a disaster in the sense that science gives us the only picture of the world that is capable of being universally shared which is also able to accurately guide our response to both nature and the technological world.

Read the whole thing, the pointer is from Arthur Charpentier.

Comments

To be a winner in an economy where labor has the underhand you cannot merely be good or competent, you have to be exceptional. You have to adapt or be left behind

By definition not everybody can be exceptional...which means we are screwed.

Not everybody has been exceptional for all of human history and we've made it this far

Come to think of it, lack of intelligibility runs like a red thread throughout Average is Over

Interestingly enough, this theme also runs through Peter Watts' Blindsight and Echopraxia; have you tried them yet?

Can you elaborate on that? I'm not sure I get what you mean; I thought they were both pretty good. Watts does assume familiarity with quite a lot of material, and his Notes and References section is awesome (haven't seen many writers go to those lengths to try and justify the stuff in their books).

He means that the characters in the books regularly run into situations that are unintelligible (at least consciously) to them, and indeed that the concept of what it means for something to be intelligible is something the books are interested in, not that the books are unintelligible to the readers.

(spoiler alert)
Having read Blindsight a few years ago I think you're referring to the distinction made between intelligence and consciousness as opposed to intelligibility. Intelligence without consciousness is still intelligible.

1) Took a bunch of kids skiing this past week. Every one emerged from their bedrooms for breakfast with iPad attached and glowing. To think this is a 1960s wold, with minor tweaks, is pretty absurd. They inhabit a networked world no one born before 2000 can comprehend. Every app is easy for them. Computer phobia is something mom might have.

You sound much, much older than me, despite having a child not much older.

But the apps are easy. The popular ones work. The paths are very well trod, and no one will pay a premium for someone who can use an I phone, in fact it is the other way around, you pay to use them.

All this stuff does is make the lucrative endeavors more complex and demanding. And those who are skilled at using machines will need to be extraordinarily productive to be able to generate revenue to cover the capital costs. And be smart enough to fix the machines and design workarounds for their flaws.

This stuff isn't impossible; our predecessors were working on the edge of their understanding as well. I had one of my employees complain that they were working on things that he didn't understand. Welcome to the real world. After 5 or 10 years he may have enough understanding to be able to figure things out.

The problem I see is that to be able to function in such a difficult and demanding environment you need to be basically functional as a human being. Having a mother and a father is a good start. For those who aren't there will always be government work. Or teaching.

I think people born before some year (I pick 2000) see the physical world as something real, and computers as something they can do. After some year the distinction vanishes. Facetime was always there, as real as the breakfast table.

Absolutely true that revealed preference suggests that the latest inventions (smart phones, tablets) are as disruptive as any previous historical invention. People in the developing world get a cell phone before indoor bathrooms, or kitchens. I think economists are missing a lot of the data from this since most of the time there is no money changing hands when people interact through electronic means. In fact it can look like less economic activity (less films paid to see, less books paid to read, less land line international calling and so on). The 1960's world was very tangible. Today's world is much less. Trying to compare the two by counting the tangible items created in this world versus the 1960's is going to miss the point.

Also, Rich Searle talks about how we might be going back to a world of inequality like early Industrial Revolution times. The thesis is that a small capitalist owning class basically takes all the gains from the robotic revolution. But 1) why do these robotic engineers not act like regular capitalists and compete with each to serve the consumers? If manufacturing is a path to such riches, why don't we see that in today's world? Try going into business to make something and even if you are first mover you will quickly find other people competing with you (e.g. Apple and Android). And 2) The proper parallel with the industrial revolution is to consider it a process moving from a) pre-industrialization with a very few people rich but the vast majority unimaginably poor moving to b) more people rich, a few people not badly off, and most people very poor (early industrialization) to c) today where we have a few people very rich, most people very well off and only a very small number of people poor (in a relative sense).

The mistake is to compare b) with c) and say how awful it is that there is such poverty. Actually you should be comparing b) with a). In other words everyone agrees it would be great if everybody was well off but the real issue is how do we get there. It is not to stop industrialization at b) since then you never get to c). Arguing from induction we are tending toward a d) with further automation, where the cost of goods and services are such that no-one needs to go short of anything. Of course there are things that could stop us from getting there (like rogue AI) but induction by itself does not suggest a dystopian future.

Good reminder on cell phones and Africa.

On the Victorian future, I think Searle does look too far ahead. We are a few decision points away from either Robo-Socialism (with a minimum income or something) or destitute millions. IF robots do throw millions out of their homes THEN I'd think there would be a political response.

2) The idea that arcane intelligences might reduce the population to superstition is good, but too late. We can't correctly answer global warming or autism or religious fundamentalism because populations do not trust the intelligences they have now, in human experts.

"They inhabit a networked world no one born before 2000 can comprehend."

Nonsense. Most of the people (myself in very small parts) who designed and engineered that world were born well before 2000. We comprehend it pretty damn well thanks.

I was on the 'net with Vinton Cerf myself, but I'm not deluded enough to think that I am a native in the same way as a 10 year old with an iPad, or a 12 year old running a clan raid.

Well, I'm a child of the Carter administration, and haven't noticed anything like that.

It would be nonsense if he were saying no one over 15 can understand the internet. But obviously, he isn't.

He's saying that kids under 14 have spent so many of their formative years in a networked life that their view of the world is completely alien to you. And he's right.

Science explains everything, and Neil deGrasse Tyson is its prophet.

But seriously, it would be interesting to compare the ratings of Tyson's TV shows with those of Yusuf al-Qaradawi.

Pat Robertson trumps them all. He frequently talks about science. And why it is wrong.

As Al Gore says, "Science is whatever we want it to be."

Al is a big reason that we can pretty safely say that Science Is Over.

Is it really a compliment to be told that "lack of intelligibility" runs through one's book?

By the way, Tyler might like that review on substance, but in terms of writing style it is one of the worst I've tried to slog through in a long while. Once again, "lack of intelligibility" rears its ugly head.

Average is over - book review reader edition

To be fair, it was written by a robot.

From the Searle review:
"For Cowen much of science in the 21st century will be driven by coming up with theories and correlations from the massive amount of data we are collecting, a task more suited to a computer than a man (or woman) in a lab coat. Eventually machine derived theories will become so complex that no human being will be able to understand them. Progress in science will be given over to intelligent machines even as non-scientists find increasing opportunities to engage in “citizen science”" [snip]

I haven't read Cowen's book. But, 'machine derived' correlations of data are certainly a reasonable forecast, and even exist today: http://www.huffingtonpost.com/2014/05/15/spurious-correlations-graphs-make-no-sense-video_n_5325835.html

But to imagine that software will generate 'theories' about how and why these correlations exist, and whether they're indeed valid, is an extremely long bow to draw.

Correlations are interesting, but almost always meaningless [have a browse @ the 'Spurious Correlations' blog]

The critical part that both Cowen & Searle miss in this thought experiment is articulating & 'proving' the mechanism by which this correlation/phenomenon occurs. A putative mechanism can then drive experimental design to disprove it. But even articulating potential mechanisms in nature is fraught with difficulty, and involves awareness/understanding of past work, context, creativity, 'order-of-magnitude reasonableness', adjacency of similar research, trial & error & serendipity [luck].

Whether or not software will ever be able to achieve this, its certainly not going to do so in our lifetimes. Imagining that it will betrays a touching naivety about how science is really done:

Derek Lowe, and others on this notion, in the realm of target based drug discovery: http://www.nytimes.com/2014/11/16/magazine/why-are-there-so-few-new-drugs-invented-today.html?gwh=2F536543830FAFD960C04B282192D81B&gwt=pay&assetType=nyt_now

& Ash Jogulekar: http://blogs.scientificamerican.com/the-curious-wavefunction/2014/02/05/why-drug-discovery-is-hard-part-4-taking-the-fight-to-the-enemy/

UPSHOT: As the NYT writer pithily summarises: 'Biology is just too complicated'.

Re: "The review is excellent and interesting throughout"

I say the same when someone gives my comments a +1.

I have to say so myself, in all modesty, the review of my review of my comment is excellent and interesting.

+1

My review of this reviewers review of his review is that his review of his review would be much better if his review of his review were less repetitive. That is my review of his review of his review of his review and I will not review my review of his review of his review. To review my review of his review of his review would be a repetitive review of his review of his review.

Thank You

Self-recommending.

The post? "The review is excellent...."

'My review of this reviewers review of his review is that his review of his review ....'

“One only wishes that his vision had room for more politics, for if Cowen is right, and absent us taking collective responsibility for the type of future we want to live in, 2040’s America might look like the Britain found in Dickens, only we’ll be surrounded by robots.”

Cowen, as a libertarian, sees only his idea of “markets” like Newton saw gravity. If the markets take us to an “Average is Over” world, so be it! His type will do fine! But actually “markets” are an invention of man and politics. Man and politics came before markets! If we wanted to, we could organize a social system of greater inequality. “Average is Over” is not a fait accompli!

You said "we could organize a social system of greater inequality." Clearly true! But I think you meant "greater equality." Also clearly true: we could for example engage in worldwide war and drive total human wealth down a great deal.

Something bothers me about this. It seems Tyler (and most here) are agreed that we must accept what "science" tell us without detailed inquiry since we are incompetent to assess those things without the expert knowledge which only scientists possess. But we all understand favorable or unfavorable outcomes and accurate or inaccurate predictions. Without knowing anything about how a drug works, I can determine whether it saves lives. Demanding evidence before accepting what "science" recommends may delay the favorable outcomes which "science" has promised but that seems like a small price to pay whether the outcomes are not all that obvious.

Real life is not all that much like chess. Among other things, the global warming "pause" has now lasted almost as long as the period of rising tempratures which brought the theory into favor in the first place. Before the rise in tempratures which first became evident in the mid-1980s, "scientists" (including some of the same individuals) were confidently predicting that human activity was about to cause a new ice age. Within living memory, moreover, it was possible to meet and receive instructions from full professors of the "science" of phrenology.

It seems to me that we should accept or reject "science" or "technology" to the extent that "science" produces useful products or, at a minimum, verifiably accurate predictions. Where "science" and "technology" have not already demonstrated favorable result in terms that we can understand, we should not meekly accept tuition just because our instructor has a lab coat.

Comments for this post are closed