Saturday 4 November 2023

The Bourgeois Politics of AI

Nice to know how, during a quiet political period, Rishi Sunak prioritised interviewing Elon Musk at this week's Artificial Intelligence safety summit. Dubbing it Rishi Sunak X Elon Musk as if the Prime Minister's a veteran DJ/Producer sprinkling career magic on a talented new face, there wasn't much of substance said during their 50 minute conversation. Except Sunak got to show off his superficial knowledge of "tech", undoubtedly to position himself as a future silicon valley/roundabout angel investor after the curtain has fallen.

As inexperienced an interviewer as he is a politician, Sunak's gushing indulgence of Musk allowed him to go full hype with the potentials of AI. The throttler of Twitter talked about the exponential development of the technology, estimating its capacity grows between five/ten-fold per year. We don't really need to worry too much about economic impacts because the time will come when no one needs a job. The productivity and automation AI promises means we can look forward to a pampered future with our needs met through a 'universal high income'. Technology will be the great leveller and an age of abundance beckons. "Read Iain Banks!" enthused Musk, showing absolutely no awareness of the political edge that suffuses the Culture novels.

This techtopianism was too much for Sunak, who has spent nearly four years clamping down on any sign of political hope. At the risk of challenging his future boss/client he gently pushed back saying work is important because it's the source of people's drives. AI should "assist" people at work, not replace it. This philosophy, of course, has nothing to do with senses of self-fulfilment. Like all Tory governments the primary concern is the preservation of the wage relation, hence their hostility to four-day week experiments and anything that loosens the tie between income and employment, and with it the dependence of the worker on their boss. Not that Musk takes his own communism-lite vision mongering seriously. A look at labour practices in the firms he's owned over the years is enough to tell you that.

The only other couple of nuggets worth commenting on was Musk's attempted justification of introducing charges for using X/Twitter. Because generative AI makes faking identities or putting out false information much easier, charging users a nominal fee of around $1/year would get around this authentication problem because no bot farm would dole out anything to run hundreds of thousands or millions of fake accounts. If the initial pilot in the Philippines and New Zealand is rolled out globally, surely this would sound the death-knell of the platform for many of its users (see you on Bluesky). The second was safety and regulation. Musk is a proponent of 'longtermism', the fad interior decor many silicon valley moguls have adopted for their mind palaces. It worries about possible future apocalypses, such as Terminator-style rogue AI scenarios. Though, tellingly, not environmental crises and climate collapse. And so he praised efforts at regulating AI, including congratulating Sunak on making sure China was present at his summit. He noted they had signed the official communique and were very interested in safety - a welcome alternative to the reflex Sinophobia fast becoming de rigeur in bourgeois circles.

All told, what was the point of the summit? Was it just an effort by Sunak to bask in reflected glory, rub shoulders with people even richer than him and set him up for a Nick Clegg-style exit? That's too simple an explanation. As Musk rightly noted, London is second only to Silicon Valley as a centre of AI development. True, but this is hardly an "indigenous" development seeing how Deep Mind - a world leader - was acquired by Google a decade ago. But, as far as British statecraft and capital are concerned, there are good reasons to keep their seats at AI's top table with the US and China. As noted here and elsewhere many times, the management of the UK's decline as a world power has seen it assume as close an alliance with the US as possible in a sort of elder statesman and, occasionally, backseat driver role. This is the imbibed foreign policy common sense as far as both parties of government go. The development of AI is another opportunity to enmesh those interests more closely and ensure the Americans don't have a complete monopoly on AI-enabled military hardware and intelligence technologies. The second is the City. The automation of trades and the increasingly complex algorithms and modelling that goes with them has and remains a key driver of IT innovation. If the City is to remain the global hub of commercial and finance capital it has a clear interest in keeping abreast of, driving, and integrating the development of generative AI. Something Sunak is keenly aware of and helps explains why he takes this seriously. He is from that world, after all.

For all the hype around AI, we have to temper the official optimism with more earthy realities. Musk's flights of fancy are a crude progressivism: AI offers the possibility of a life of ease, which means everything in the meantime can be justified with that end in mind. Sunak's aversion to this is more honest in the sense that positing AI as a workplace "co-pilot" asserts its unambiguous articulation with existing patterns of class power and exploitation. Our position is the opposite, separating out the cover for capital the utopian impulse provides and marrying it resolutely to social critique. All previous machinery have reconfigured and reinforced exploitation, but it's the opportunities in the here and now AI and further automation offers that we should grasp. First, as an internal critique of the narrow purposes of accumulation to which it will be put, but also as escape. I.e. A forceful assertion of our politics of work and what it should look like. The better tomorrow won't come with the Sunaks, the Musks, and the AI they're getting excited about. Struggle is our only means of deliverance.


Anonymous said...

Lol, make up your mind Elon, eh? Is AI a bringer of Banks-esque utopias, or an existential risk (to your personal position at the top of the social tree, that is)?

Notice that the routines of both the clowns on stage here are essentially science fiction - necessarily so, because the prediction horizon for anything AI-related is so close, that basically anything beyond the nitty-gritty of implementing some piece of software right now is science fiction by default. So, this is a good time to remember what Harari has to say about science fiction. Science fiction is never about predicting the future accurately, because a realistic assessment finds that to be a mug's game. Science fiction doesn't predict the future, it creates the future. Its pretend pictures of the future are always about capturing important nuggets of useful information about the world that we live in, which can influence our thinking to be less behind the curve in the world of the now. Examined in that light, the routines presented by Sunak and Musk here both sound reasonably on the nose, as you tell it, since they both capture something important about what our relationship with this tech (and the anachronistic social structures that have to change to fit it in, now that it's here) is going to have to be.

Graham said...

The idea that AI will mean nobody has to work is the same nonsence that was popular in the 60's when automation was promised to be the gateway to a life a leisure. It turned out to be the gateway to unempolyment and low skiled / low paid jobs.

AI is mainly just the extension of the current trends in automation and computerisation with more processing power and vastly more data.

The only jonbs that will be safe are those too cheap to automate.