Artificial intelligence, tech workers, and universal income: An interview

Thanks for posting this Chili!

This is a great read Chili. Cheers.

I'd be interested in that, and actually in participating in the debate, but sadly I don't think I'd have time. But let me know in any case.

Ron appears to be an acolyte of what some are calling the California Ideology - the reactionary techno utopianism common among professionals in the technology industry

http://www.metamute.org/editorial/articles/californian-ideology

http://www.imaginaryfutures.net/2007/04/17/the-californian-ideology-2/

As for the fantasy of robots doing all the work, it's simply not feasible, economically or technically.

I wrote a short essay explaining why on my blog

http://gangboxnews.blogspot.com/2015/10/jetsons-revisited-or-do-we-have-future.html?view=mosaic

Even if it was - it says volumes about Ron that he sees that life of idleness and sloth as a utopia rather than a hell on earth

As for Chili Sauce's thesis that:

Reaching out to and engaging with politicized technology and computer workers can only benefit the movement.

I'd have to say no to that

It's not "tech workers" that hold the power in society - it's workers in industry, in the manufacturing, transport and circulation of commodities

If you're looking for "choke points" in industry, look on the loading dock or in the warehouse - that's where the flow of commodities can be stopped.

Of course, that's why the techies fantasize about an all robot workforce - no class struggle, no strikes, no workplace resistance, no organized "collective laborer" workforce who are liable to see their common class interests

They'd rather see us become atomized consumers, sitting at home waiting for the "Basic Income" welfare check to be delivered by the robot mailman.

That's a Silicon Valley venture capitalist's daydream - surplus value without workers - but it's not realistic and it sure as hell isn't a utopia

GREGORYABUTLER10031

As for the fantasy of robots doing all the work, it's simply not feasible, economically or technically.

Er, it is. It's pretty much happening. That's pretty much like saying a few hundred years ago that the steam engine would never catch on

As for the fantasy of robots doing all the work, it's simply not feasible, economically or technically.

While I agree with you that right here, right now your statement is correct and that it is much cheaper to simply shift production to even more low-level income countries (Africa will likely be the last bastion of human labour being cheaper than tech in the future), even Chinese factories are heavily automated. More and more jobs we though were unique to humans have been completely automated, even the cushier (though boring) jobs of paralegals (simply replaced by algorithms).

And yes, what the interviewee is taking as predestined given--strong AI--is simply not possible now, and from what I can gather will remain impossible. But while it is a fantasy, it does not mean that some future breakthrough won't make it possible or at least makes it possible to automate more and more human work.

It's not "tech workers" that hold the power in society - it's workers in industry, in the manufacturing, transport and circulation of commodities

But that's not what Chilli is saying, he is saying we should reach out to such workers. Like we should with all workers.

If you're looking for "choke points" in industry, look on the loading dock or in the warehouse - that's where the flow of commodities can be stopped.

Sure, but these places are getting automated. Consider Amazon's kiva bots (although just for picking, not for packing, which still requires human hands and a mind), or the AGV's used at some terminals at the port of Rotterdam. And those examples are just the beginning of what could turn into complete automation.

That's a Silicon Valley venture capitalist's daydream - surplus value without workers - but it's not realistic and it sure as hell isn't a utopia

Yep, that's is indeed their dream, and yes it would be more like hell than anything else. If we don't work, how do we eat? Life would be nasty, brutish and short. But where did anyone say it is a utopia here? I think you are confusing an interest in these issues with agreement; Chilli was quite clear on his position in the introduction.

BTW Gregory: did you read Fred Turner's From Counter Culture to Cyber Culture? It's a much better take on what Barbrook's refers to as the Californian Ideology.

Just to put things in context, in 2004 Frank Levy and Richard Murnane, in their book, the new division of labour, wrote that computers would essentially be able to perform jobs where simple rules needed to be followed, but when it came to pattern recognition, complex tasks in a changing environment, they said, and seemed to prove conclusively, that human labour would always be superior. An example they gave was driving. Eleven years later and self driving cars have been road tested and in Japan, a driverless taxi service is hitting the roads. Up until recently a robot on a production line would have to be reset for each task that involved a different pattern, that was supposedly never likely to change, but now there are ones that can be trained to perform multiple different tasks without reprogramming, at a 2008 conference on AI in Barcelona, the consensus was that simultaneous localisation and mapping would not be possible for AI any time soon. Two years later the kinect for xbox came along and did pretty much that.

One thing that I think is interesting is that up until about two years ago, most articles on robotics asserted that increased automation would just create new jobs that robots couldn't do, from a year ago you see a split between ones who still maintained that and others who claimed that automation would swallow up all the jobs, and this year, you have book after book, article after article saying, listen, automation is going to wipe out most jobs, we'll need a way of dealing with that - and they mostly suggest UBI.

Flava, could you post some references to the texts you draw on? I'd like to read more on this sometime in the future.

Well most of that stuff was just from a book I am skim reading at the moment called "The Second Machine Age". I'm skim reading because even that book, that was primarily researched in 2012 is a bit out of date. I'm going to start reading the more up to date The Rise of The Robots by Martin Ford when I finish this but couldn't tell you if it is any good yet. A lot of other info I get from new scientist, I have an online subscription. Paul Mason kind of skims the subject in Postcapitalism, but oddly, given the importance of technology for his theory he doesn't go into any depth. This was a recent opinion piece, presume it's paywalled so I'll c&p here:

As tech threatens jobs, we must test a universal basic income
Paying a basic wage to all could hold society together if automation destroys swathes of work, but we need more trials to be sure

As tech threatens jobs, we must test a universal basic income
(Image: Andrzej Krauze)

THE idea that technology would destroy jobs and dramatically increase unemployment was considered a fallacy not long ago. Economists believed technology always created new opportunities and hence new jobs.

Today, with the rise of machine-learning algorithms and advanced robotics, many have changed their view. It’s possible that within 20 years almost half of all jobs will be lost to machines, and nobody really knows how we are going to cope with that.

Those who still have faith in technology’s power to create jobs fail to recognise the shift to a “superstar economy”, where a handful of companies disrupt markets, make billions and employ very few people, while the rest fight for the scraps.

So how would the millions whose jobs are at high risk of being automated survive in this landscape? One of the most interesting proposals, that doesn’t rely on the fanciful idea that the market will figure it out, is an unconditional basic income (UBI).

It’s a simple concept with far-reaching consequences. The state would give a monthly stipend to every citizen, regardless of income or employment status. This would simplify bureaucracy, get rid of outdated and inefficient means-based benefits, and provide support for people to live with dignity and find new meaning.

Perhaps the biggest UBI experiments, involving a whole town in Canada and 20 villages in India, have confounded a key criticism – that it would kill the incentive to work. Not only did people continue working, but they were more likely to start businesses or perform socially beneficial activities compared with controls. In addition, there was an increase in general well-being, and no increase in alcohol, drug use or gambling.

These results are promising but not conclusive. We don’t know what would happen in other countries, or whether the results would apply on larger scales. New experiments may give a clearer picture. The city of Utrecht in the Netherlands has started a UBI pilot, with a view to extending it to 300,000 residents; Finland is poised to do the same.

But for now, any country implementing a nationwide UBI would be going out on a limb. There are so many variables that if it goes well, it will be difficult to analyse and understand exactly why it worked and how it can be replicated elsewhere. If it goes wrong, it will deter further research and experiments.

It would be better to test it on a random sample of 1000 towns of 10,000 people each than in one country of 10 million. A network of such experiments in many countries is the perfect set-up to learn and prepare for larger-scale projects, right before the wave of automation hits us hard. We need to be ready.

This article appeared in print under the headline “Money for nothing”

By Federico PistonoFederico Pistono, a computer scientist turned social activist, is author of Robots Will Steal Your Job, But That's OKWant more? Read the extended version of this article.Magazine issue 3041 published 3 October 2015

I've literally just started with Ford's Rise of the Robots - he's been talking about how advances in 3d machine vision mean that's now prototype packing robots which dramatically outperform humans. I think he also thinks UBI is necessary to sustain a mass consumer market once the McJobs that have hitherto absorbed labour power become automated.

Cheers, Flava.

Gregory A. Butler, why do you have two libcom accounts: GREGORYABUTLER10031 and Gregory A. Butler?

I ask because your posts often make baseless assumptions about others and you regularly attempt to put words in people's mouths. And you do it under both identities. What gives?

Chilli - I'd be up for that in principle, though I'm on deadlines for about the next month (one of which is I've been asked to write something on Srnicek & Williams' 'Inventing the Future', which also makes a full automation/UBI gambit).

I think the general argument is 'this time it's different'. Previous waves of automation laid off workers, who were then reabsorbed by an expanding service sector (well, at an aggregate level, unemployment in ex-industrial/mining areas is still high). But now, you've got robots coming online that can produce bespoke burgers in much quicker time than people, and which cost just a few multiples of a single burger flipper's annual salary (and don't unionise, or sleep, or shit). Hasn't McDonalds responded to 'fight for $15' with automation?

Generally automation creates new jobs (e.g. programmers, technicians), but not necessarily in a similar quantity, and not necessarily immune to partial automation themselves. And generally there's still plenty of things robots can't - or we don't want them to - do, like health and social care (though Robot and Frank doesn't seem that far off). But these are also the things being starved of public funds and largely done as unpaid work, so not an alternative source of mass employment either.

In my job, a big admin thing in universities is sorting and distributing coursework submissions. We used to get hundreds of submissions each week, two copies of each, which needed to be separated into two piles, put in candidate order, split into batches for each marker, signed out, signed in, marks entered, marks checked...

Now we have e-submission. The student uploads a file. The marker sees only the students they need to mark, annotates directly onto the screen, enters the mark, the academic in charge of the module sees an instant summary, can check the means and standard deviations are consistent across markers, and release. The admin overhead has been slashed to just setting up the system at the start of term - a small part of one person's job rather than 50% of two or more clericals.

Then in the library, checking out and returning books is largely automated. An RFID-scanning conveyor belt sorts returned books, and every member of staff just has a certain number of hours per week trolley quota to reshelve the sorted books in the right sections (rather than employing dedicated staff to stamp books out and sort returns). So a lot of traditionally 'safe', entry-level white collar jobs like clericals and library assistants are getting pretty automatable, as well as the warehouse pickers, packers, and burger flippers.

That's millions of people - if that's all being automated at the same time, unless there's some kind of major public works program (loft insulation or renewables installation?), it is hard to see where millions of new jobs are going to come from, when the major corporations of our time are tech and IP giants who employ a fraction of the staff e.g. General Motors or Ford did at their peak.

Chilli - I'd be up for that in principle, though I'm on deadlines for about the next month (one of which is I've been asked to write something on Srnicek & Williams' 'Inventing the Future', which also makes a full automation/UBI gambit)

How transhumanist is "Inventing the Future"? It was quite well hidden in the accelerationist manifesto, which for all their attempts at epistemic acceleration was quite backward looking in its modernity and enlightenment hard-on for and desire to command nature, life and death (also a staple of transhumanist thinking).

The manifesto was also strangely contradictory on some central issue (climate change, but let's fix that by thinking we can bend nature and the environment to our will through terraforming; abolish the value-form, but let's keep financial infrastructures intact).

Do any of these things pop up again the in book?

The accelerationist language is more or less entirely stripped out, they say to not distract from the argument. Iirc it comes back strong in the conclusion but I can't remember how transhumanist it was without checking (was reading it with my ecological hat on). There's also loads of caveats which seem to go against the thrust of it, so e.g. 'Progress' is defined as the possibility that things could get better... or worse, 'Modernity' is affirmed but more or less emptied of content, with the project being to define it, etc. I guess those caveats could be taking on board criticism, or a means to deflect it. The first couple of chapters will piss off anarchists as it more or less defines prefigurative direct action as a TAZ that changes nothing, but if you read it, worth pushing on to the more propositional bits.

Cheers, JK. If the H+ stuff was hard to find, they're hiding it in the same way as they did in the manifesto (not that I think their "hiding" it has anything to do with some nefarious agenda or sneaking that in Schmidt style). IIRC there was a few mentions about overcoming the limitations of human biology, (perhaps morphological freedom, though that may have been from another text in the accelerationist reader). So no overt H+ language, but "coded" so that if you've read a bit of Kurzweil, Kaku or any of those you'd notice it. To some degree I consider Srineck and Williams manifesto to be a leftist version of transhumanism, in the same way that we talk about how Trots etc. are the left-wing of capital.

There's also loads of caveats which seem to go against the thrust of it, so e.g. 'Progress' is defined as the possibility that things could get better... or worse, 'Modernity' is affirmed but more or less emptied of content, with the project being to define it, etc.

Interesting, maybe they have taken on some critiques then. But the manifesto was pretty old-time Enlightenment IMO. But seems like a cop-out with emptying progress and modernity of content, especially in the light of the manifesto where their thinking about these things are clearly an epistemic deceleration especially in its affirmation of Man standing outside of and against nature (IMO I'd say ecological thinking of the Jane Bennet variety is more of an example of epistemic acceleration than the accelerationists!). And given how much progress has been critiqued by the likes of the Frankfurt School, Paul Virilio (with his accidentology that has a dialectic of progress, and his general critique of the propaganda of progress as an accident of thinking) and Jacques Ellul, it seems like a strange thing to do indeed. But from reading the manifesto, I'd say they are very much beholden to the propaganda of progress (apres moi la deluge!).

But thanks for your comment, now I am looking forward to reading their book a bit more (I was lukewarm before). For all the faults I found with the accelerationist manifesto, it did provoke a lot of thinking so I am hoping the book does the same.

Same guy, two different email accounts - I've had login problems.

I see you're still a fan of ad hominem attacks when you can't challenge the validity of my arguments on their merits

On the question, I think folks are drinking the Silicon Valley Kool Aid about the feasibility of automation of all work (it's not realistic).

I'm also struck by the general pro robot sentiment here - that you all seem to think that replacing human labor with machines would be a step forward. I see it as a giant leap towards barbarism, an eternal dictatorship of capital and the final historic defeat of the working class.

If it ever did become possible to automate all production, I'd lead the charge to smash the robots and to crush and destroy the techies who program them - to save the world for humanity. (that's the one thing I'd ever agree with anarcho primitives on)

Where on earth do you get that impression. As I said, interest in these things does not mean acceptance. Please stop trying to ascribe these ideas to people that don't hold them. That's the second time already in this thread. I was polite above, but frankly, if you continue and don't start discussing in good faith you can just fuck off. And rather than engaging with what I wrote above, you are again just stating that people have drank the kool-aid.

But in any case, some of us are actually thinking about capital's event horizon. In a communist society I would want drudgery to be automated. Why keep fucking boring tasks when you don't have to? And also, that capitalist are actually trying to work towards this future means that we have to take this seriously. But it seems like you think that we should not even entertain these ideas, at all. That's just fucking idiotic.

Then again, from reading your posts you strike me as someone who wants to affirm labour, work and the working class rather than striving for abolishing these things. And that you believe that is pretty much clear from this

If it ever did become possible to automate all production, I'd lead the charge to smash the robots and to crush and destroy the techies who program them - to save the world for humanity. (that's the one thing I'd ever agree with anarcho primitives on)

It would do you some good to read Krisis' "Manifesto against labour".

Khawaga

Interesting, maybe they have taken on some critiques then. But the manifesto was pretty old-time Enlightenment IMO. But seems like a cop-out with emptying progress and modernity of content, especially in the light of the manifesto

I think the charitable reading is that their main claim is 'the left needs to struggle to define modernity and hence the future', and their contribution to that is the manifesto/concluding chapter. But logically you could accept their main claim while criticising their particular vision of modernity, and that discussion would be part of what they're calling for, i.e. a discussion of what the future should be. Maybe it's even deliberately provocative (it is a manifesto, after all) in order to provoke that discussion.

GREGORYABUTLER10031

If it ever did become possible to automate all production, I'd lead the charge to smash the robots and to crush and destroy the techies who program them - to save the world for humanity. (that's the one thing I'd ever agree with anarcho primitives on)

What does your vision of libertarian communism look like? I'm baffled because for me a huge part of it is the abolition of wage labour and freeing up humanity to concentrate on creative work and work that interests the individual - or to do nothing if they so wish. Automation, for most classical communists was the key to this and for me, our current level of technology makes communism a distinct possibility in the here and now, providing we can successfully organise for it and overthrow capitalism. I'm not intetrested in a steampunk revolution to be honest.

UBI and it's associated justifications in terms of the growing significance of information technology and robotics on the evolution of the capitalist economy is very much up Paul Mason's street in his book on 'Post Capitalism' so this thread could usefully be linked in to that as well. Just finnished reading that book so might make some critical comments there.

I'm in the middle of writing something about it too. Will sumarise my thoughts on it here when I'm finished.

"The Singularity" is a techie fantasy, so is the idea that robots will do all the work for us

http://www.salon.com/2015/10/15/calm_down_artificial_intelligence_is_not_going_to_take_over_the_world_partner/?utm_source=facebook&utm_medium=socialflow

Any politics based on that misanthropic fantasy are not a way forward

GREGORYABUTLER10031

"The Singularity" is a techie fantasy, so is the idea that robots will do all the work for us

http://www.salon.com/2015/10/15/calm_down_artificial_intelligence_is_not_going_to_take_over_the_world_partner/?utm_source=facebook&utm_medium=socialflow

Any politics based on that misanthropic fantasy are not a way forward

Not only is that a really awful article, but it also has nothing to do with the automation of labour. Automation of labour is not a fantasy, it has been happening for quite some time and is accelerating. You don't need self awarw AI for that to happen, you just need robots built for specific tasks - and even now there are multitasking robots.

Once again I ask what your vision of a communist society is, because you aren't going to interest many people with a vision that includes drudgery.

Yeah, you'd only need strong AI to automate tasks that require reflexive thought or something. Any job that can be explained in flow chart can in principle be coded as an algorithm (afaik this is the case for lots of skilled work). The limit to things like packing robots hasn't been the coding side but the sensor/feedback side, e.g. machine vision. As advances are made here, automation becomes viable.

[youtube]ClXfa4stfJM[/youtube]

(note this robot doesn't need the batteries to be in a standard position, it 'sees' them and quickly adjusts. it would probably be cheaper to just funnel the batteries into a standard orientation, but it's a trade show demonstration of the tech, 'intelligently' packing objects having been one of the warehouse jobs not so automated so far, unlike picking and portering.)

Gregory, you really are thick. AI does not mean the singularity. While the singularity is sci-fi, AI is not. AI does actually exists in the here and now, but you seem to assume that strong AI (i.e. a being with a general intelligence like humans) is what AI really is. Sadly to say, you seem to know very little about this topic at all.

And again, rather than engaging with other posters, you yet again ascribe ideas to us that we don't have. Now we are all transhumanists in favour of the singularity. If you had bothered to actually read some of the posts you may have realized that some of us are extremely critical of that belief. That's why I have a problem with the left accelerationism of Srineck and Williams; they are in part transhumanists. And there is a lot of shit thinking and politics that comes with that.

JK

But logically you could accept their main claim while criticising their particular vision of modernity, and that discussion would be part of what they're calling for, i.e. a discussion of what the future should be. Maybe it's even deliberately provocative (it is a manifesto, after all) in order to provoke that discussion.

Agreed. I do like the call for an epistemic acceleration and think it is warranted, but I am very critical of their version of it which I think represent a deceleration and an uncritical return to Enlightenment/modern thinking. And their argument is based on some strawmen (e.g. the folk politics and localism they think is endemic on the left; far from it I think it is rather the opposite), is contradictory etc. Whether they were deliberately provocative I don't know, and I don't think it matters, because they have provoked a lot of discussion and for that I think that the manifesto I great despite all the major problems I have with it.

GREGORYABUTLER10031

Ron appears to be an acolyte of what some are calling the California Ideology - the reactionary techno utopianism common among professionals in the technology industry

...

Ron here...

I think you're right to point out that in the tech world, and with silicon valley-ites, and "futurologists" in general, this automation future is viewed very positively, as a type of "salvation" or utopia. I don't necessarily think these changes are inherently positive or negative, but I view them as inevitable, and they have the potential to be positive if we make them positive-- which we need to start doing right now, very vigorously. If we fail, the future you envision with people waiting for their welfare check surrounded by squalor is very possible.

I don't see a future where these technologies don't exist, and these technologies necessarily will fundamentally change many aspects of our society. Whether this change is good or bad depends on us.

However, what do you view as the ideal or likely alternatives? I can only envision the alternative to ubiquitous automation being a kind of luddite society maintained with heavy-handed government intervention, but I fully acknowledge I likely have a blindspot to other possible outcomes.

Joseph Kay

I've been asked to write something on Srnicek & Williams' 'Inventing the Future', which also makes a full automation/UBI gambit).

This is up here now: http://thedisorderofthings.com/2015/11/04/postcapitalist-ecology-a-comment-on-inventing-the-future/

There's another Out of the Woods contribution coming in the next few days.

Good piece JK. I now look forward to reading Inventing the Future even more (for some reason I can't get a hold of a copy in North America).

I'm curious to see what they write more about Project Cybersyn (I heard Srineck gush about it at a talk) as the clear problem with that project was that they were trying to literally engineer socialism but ran into the usual problem of not taking social relations into account. Just from reading the manifesto and few reviews of their book, it seems like they take an extremely technocratic approach to... almost anything, and they appear to be completely beholden to the notion of progress. They probably should read some Virilio...

Their "folk politics" category also seem to be a bit of a straw man; a construction of theirs where they can put everything they don't like into. Do they explain what they mean with it more in the book?

They define folk politics as a tendency rather than an ideology, but yeah it's quite a fuzzy concept. Another Out of the Woods piece that went up today says a bit more on it: http://thedisorderofthings.com/2015/11/05/why-we-cant-let-the-machines-do-it-a-response-to-inventing-the-future/

That's also a really great piece. Thanks for posting. I gather from both your and the other piece something I thought may be the case with their book: the very "masculine" approach to things. Perhaps unwitting on behalf of S&W, but makes perfect sense due to their transhumanist inclinations; and H+ can at least in part be described as a male, white power fantasy.

They do reference people like Silvia Federici and Walter Mignolo, and are keen to distance their advocacy of 'progress' from colonial history. I guess the question is to what extent they add those thinkers in, without modifying the underlying argument accordingly. Like e.g. Federici (in Revolution at Point Zero) makes an explicit critique of full automation as a goal:

Federici

Reflecting on the activities that reproduce our life dispels the illusion that the automation of production may create the material conditions for a nonexploitative society, showing that the obstacle to revolution is not the lack of technological know-how, but the divisions that capitalist development produces in the working class. Indeed, the danger today is that besides devouring the earth, capitalism unleashes more wars of the kind the United States has launched in Afghanistan and Iraq, sparked by the corporate determination to appropriate all the planet’s natural resources and control the world economy.

Chilli Sauce/OrangeYouGlad - if this Inventing the Future stuff is a derail I'll start a new thread.

Guardian piece on AI/UBI: http://www.theguardian.com/business/2015/nov/07/artificial-intelligence-homo-sapiens-split-handful-gods

This is that thing on Mason's postcapitalism I was writing. Took a while.

https://selfcertified.wordpress.com/2016/03/27/the-new-world-in-front-of-us-mason-postcapitalism/

Thanks Flava.. I've saved to read later but it could also usefully be linked to this other discussion here: http://libcom.org/forums/theory/paul-mason-end-capitalism-has-begun-18072015

My long overdue follow up to the Mason critique

https://selfcertified.wordpress.com/2016/12/07/the-robots-are-coming-look-busy/

Will sumarise my thoughts on it here when I'm finished.