The race to make humans redundant

AI cannot be controlled by policy changes. We need to adapt and prepare for the fallout

Around end-March, some 1,100 people, including Tesla founder Elon Musk and Apple co-founder Steve Wozniak, wrote an open letter calling for a six-month ban on advanced artificial intelligence (AI) so that the implications of this technology can be understood before it becomes unstoppable. The concerns emerged after OpenAI introduced its wildly successful ChatGPT to all users, followed by a more powerful GPT-4. GPT stands for generative pre-trained transformer, which is a language model that uses wide and deep learning to mimic human-like responses. It learns how to respond to human queries by scouring all available net-based information and looking for patterns in them.

The problem: No country or state or group of powerful individuals can really hope to restrain or reverse the development of any technology. From the time humankind first learnt to sharpen a piece of stone, and his tribe could easily have blackballed him, the one thing we know for sure is that technology and knowledge cannot be stopped once they are out in the open. Even God did not manage to keep Adam and Eve from consuming the forbidden fruit, so Messrs Musk and Wozniak are essentially trying to put their fingers in the dyke to stop an unstoppable idea from flooding the world with the good, bad and malignant AI software. The world’s best consensus could not stop the atom bomb from being developed and proliferated, chemical weapons from being owned and used by rogue states (it was allegedly used by Iraq and Syria when their regimes faced a threat to survival), and Covid-19 from spreading all over the world. So the AI train has left the station, and at best we can slow it down for a while, not stop it.

What connects the first stone sharpener to the current developers of AI technology is that advances are usually sold as a boon for society. However, many become future threats to lives and livelihoods. As AI grows bigger and better, we must assume that it will gradually make most human effort redundant.

Economists argue that technology displaces existing jobs and vocations, but it also creates many more new jobs. That is certainly true, but what economists tend to emphasise less is that the vast majority of new jobs created will be of the less desirable and less stable kind. Technology eviscerates the good quality, middle-level, middle-skill jobs that build a stable middle class. The old income pyramid is now going to increasingly resemble a two-tiered layout, with a tiny dot of highly skilled workers on top and a long, flat line comprising the rest. The ones with highly-valued skills will be wined and dined and difficult to retain in regular jobs, and the ones at the bottom will have to be content with dead-beat jobs that technology enables (your Uber, Ola, logistics, delivery and retail jobs). It is passé to say that we are entering a gig economy. What we are really entering is the age of “subsistence entrepreneurship”, which was the case before the industrial revolution and the emergence of the modern economy.

AI is going to make the world even more polarised in terms of jobs and skills, and there is little we can do to reverse this change. The reason why AI is unstoppable is because once someone has it, rivals, competitors, even countries will feel the need to deploy such technology, even if the latter are blessed with a huge contingent of labour. Countries that are not blessed with a large workforce will automate, not ease up on immigration. So, AI will grow in size and scale with or without our consent.

You can either look at this dystopian world through a pessimistic lens, or work on strategies to deal with it, as individuals, groups, societies and countries. The snowball stands a better chance in hell than job-optimists in an AI-fuelled world.

So what can be done? A partial answer lies in learning how to cope with the consequences and benefits that AI brings, and not in trying to prevent its spread. Here are a few suggestions:

First, since micro-entrepreneurship is going to be the new normal, and not steady jobs, for the vast majority of people, we must focus on how to make laws tilt in their favour without destroying the jobs themselves. The Ola driver, Swiggy delivery person, online seller, or contract writer needs alternative publicly-funded tech platforms that offer them a larger share of the revenues they generate.

Second, we must design flexible systems for borrowing and repayment. We need social safety nets that can deal with a volatile pattern of earnings during an individual’s lifetime. Retirement age will become irrelevant; the only thing relevant will be whether one has the physical and mental ability to work. Only the super-skilled can think of retirement on their terms. Pension and social security nets have to be designed flexibly for the rest.

Third, the biggest job creator may well be the social security sector, where millions of people will deliver free or low-cost education, health and other services to those who need it. Governments can do little to expropriate the rich to run such welfare schemes, for then they will flee to safer havens. It is best to develop tax-friendly schemes to draw in voluntary contributions from the rich. Philanthropy needs a big boost, and this sector could well be a prolific job-creator.

Fourth, since power will now shift in favour of those who control the new technology, regulation will have to be extremely agile. However, most governments are currently not equipped to do this. It may be best to use the best minds from industry to advise governments on rules and regulations as the pace of change accelerates. Policy will have to be jointly driven by specialists and public intellectuals, not by bureaucrats and legislators.

Many more things need to be done, including redesigning education and healthcare institutions to meet the emerging challenges, but the short point is this: We cannot reverse the rise of knowledge and technology. The race to make human efforts irrelevant to gross domestic product growth is now truly on. We have to prepare ourselves to deal with the fallout.

The writer is editorial director of Swarajya magazine