No fear of the future

It is surprising that our lives are not already suffused by worries about the long-term effects of artificial intelligence (AI). AI is unlike any technology that we have ever seen before, and yet there is a remarkable complacency about its effects.

 

One relatively well understood reason for this difference from what has gone before is that it will likely have a different impact on the labour force than any of its transformative predecessors. All of those — from the steam engine to electricity to information technology — wound up increasing the returns to human capital enjoyed by the moderately skilled white-collar worker.

 

AI, by comparison, threatens to turn these individuals out of the workforce completely. This would entirely rework social hierarchies.

 

It was commonplace a few years ago for people to worry about the impact that robotics would have on the prospects of skilled workers on the shopfloor. And certainly, fewer such workers are required these days, when normalised by the value of output created. But it is hard to make the case, either, that the overall number of factory workers has decreased in the world once the People’s Republic of China is taken into account. In fact figures from the United Nations Industrial Development Organization suggest that manufacturing employment went up from 448 million in 2015 to 477 million persons in 2022. And, indeed, it seems that for certain workers — skilled factory workers, supervisors and foremen in specialised fields like semiconductors, metallurgy and pharmaceuticals — there is a severe shortage. The widely reported news that the South Korean conglomerate SK Hynix had paid bonuses of more than $500,000 to its workers certainly lends some credence to this belief.

 

Currently, certainly, it looks like the white-collar worker is more endangered by technology than her blue-collar colleague. Again, this is not the usual case for technical change, which economists have shown is usually skill-biased.

 

Nor has the possible impact on another influential, if much smaller, class of individuals been seen before: Creative workers. Other transformations have increased their reach. The printing press made it possible to make a living from writing; electricity and the mass media allowed them unprecedented control over the national and global narrative; the communications revolution allowed influencers to emerge from anywhere in the world.

 

What happens when AI begins to compose music, or direct short films, better than these individuals? For those without any particular aesthetic sensibility whatsoever — a depressingly large proportion of your Facebook friends, for example — AI-created content is already supplanting the tired forwards and recycled memes that provided them with entertainment and edification. It will shortly start climbing up the accomplishment ladder.

 

On Spotify and other music-streaming platforms, it is already hard to tell apart AI music from the regular overproduced slop churned out by the music industry. AI movies are still some distance away — though Elon Musk has promised us an all-AI, non-woke Odyssey before the end of the year — but you can go on YouTube and watch any number of AI-constructed trailers today. Almost all of them are appalling and unwatchable. But there are already some of genuine quality, including some superb parodies. One such AI-using creator, who skilfully reconstructs trailers for such staples of pop culture as the Harry Potter and Lord of the Rings movies in the style of, say, North Korean propaganda or Balenciaga advertisements, has tens of millions of views.

 

The implications of the very real possibility that AI will put the entire creative industry out of work are worth considering. On the one hand, it means that future iterations of AI will be forced to recreate new content by regurgitating the old — including all the current AI-generated, low-quality slop that gets thousands of views. We should fear that this will cause a spiral to the bottom in quality, rather than a race to the top.

 

But also, have we considered what happens to the narrative around a technological revolution that alienates the entire creative class — the people responsible for shaping those stories? It is remarkable that we are not already surrounded by an atmosphere of deep alarm about AI, that this existential fear is not being built into the music we consume, the books we read, and the shows we watch. Still, even if we are not now, we likely will be once the creative class discovers the level of the threat.

 

The final reason why one would imagine that the public sphere would be far more concerned about AI than it is emerges from the final unusual thing about AI: Those responsible for developing and profiting from it — its owners, its creators, its parents in a sense — are the most worried about it. Regardless of their specific attitude to regulation, almost to a man (and they are all men), they have expressed concern about what AI, including truly self-generating and self-replicating AI, could do to human civilisation. This is not how James Watt sold the steam engine, Edison electricity, or Bill Gates the personal computer.

 

Those who know AI the best fear it the most; but somehow the rest of us don’t.