Are AI companies building hell and calling it success?

We spent an afternoon at the Rodin Museum and stood before Rodin’s enormous Gates of Hell. More than six metres high, the doors are covered with human figures twisting, clinging to one another and falling towards Hell. Above them sits The Thinker, originally conceived as Dante himself, looking down upon the damned. I found it disturbing. It appeared ugly to me. But it was impossible to ignore.

It made me think about the seven deadly sins. Pride. Greed. Lust. Envy. Gluttony. Wrath. Sloth. They sound old-fashioned today. We have different words for them. Greed becomes incentive. Envy becomes aspiration. Pride becomes self-belief. Gluttony becomes consumption. Wrath becomes engagement. Sloth becomes somebody else’s responsibility. We have not defeated the seven deadly sins. We have rebranded them.

I have spent nearly fifty years in business and benefited enormously from capitalism. At my Goldman Sachs interview in 1982, a senior trader told me: “To be successful, you need to be greedy. To be very, very successful, you need to be very, very greedy.” Later, the partner sponsoring me told me I would never be satisfied, no matter how successful I became, and that this determination would make me and Goldman an enormous amount of money.

There was truth in that. Markets are extraordinarily good at turning the human appetite for success into economic activity. But markets are much better at answering what people will pay for than whether something should exist. That second question belongs to us.

I learned something about this through William Hill. In 1997, Nomura, the bank I worked for, acquired the bookmaker and I led the acquisition. We pushed William Hill into internet betting because it was the obvious path to reach more customers and make the company more valuable. The internet was faster, more convenient and more efficient. We talked about creating “sticky customers”. It was normal commercial language.

We did not ask when customer loyalty became dependency. What we thought was a change in distribution was more profound. We were removing the friction between an impulse and a bet. Today the betting shop is in your pocket. There is no walk to the high street. No closing time. Nobody watching as you place another bet at three in the morning.

We could not have foreseen the smartphone. But looking back, we were asking the wrong question. We asked 'how do we succeed,' but not, 'what happens if we succeed?'

That is one reason I ask the same question about AI. AI is removing friction again. Between a question and an answer. Between an intention and an action. Increasingly, between a human decision and its execution by a machine. I use AI every day. It enables me to translate my thoughts into understandable language. I use it to challenge my thinking, analyse information and improve businesses. I think it is extraordinary, and it is becoming more powerful every day. That is precisely why I worry about it.

AI could transform medicine, education and science. It is also likely to destroy millions of existing jobs. But technology has always destroyed jobs. Mechanisation removed millions of agricultural jobs. Industrialisation destroyed much traditional craft work. Later technologies eliminated whole categories of employment. Yet despite this, we as human beings repeatedly found fresh ways to use our abilities. New industries, occupations and forms of work emerged. We became enormously more prosperous.

So job destruction itself is not my greatest fear about AI. Speed is. What if this transition happens faster than human beings, education systems, businesses and governments can adapt? Will we find new productive roles quickly enough for millions of people whose existing skills may suddenly be worth much less? What happens to them while we do?

AI can also facilitate fraud, manipulate voters, transform warfare and concentrate extraordinary power in a very small number of companies and individuals. Nobody knows where this ends. And almost everyone has a reason to keep it going. If one company slows down, another wins. If Britain regulates, perhaps America, India or China does not. If a chief executive says the risks are too great, shareholders may find another chief executive.

Every individual decision can be rational. Together they can produce something profoundly irrational. Technology did not invent greed, pride, envy or the desire for power. It industrialised them.

It gives ancient human weaknesses extraordinary speed, scale and reach. And it creates distance between those who benefit and those who suffer the consequences. Distance is useful to commerce. It is even more useful to conscience. Perhaps sloth has a modern form. Not physical laziness. The temptation, which I have felt myself, to leave the difficult moral question to somebody else.

To say it is too complicated. Regulators will sort it out. Everybody else is doing it. If it is legal, it must be right.

I don’t believe ambition is wrong. I don’t believe wealth is wrong. And I certainly don’t believe capitalism is wrong. But success is not proof of virtue. Nor is innovation automatically progress. Perhaps every boardroom, investor and politician should ask two questions before seeking success. How do we make this succeed? And then the harder one: What happens if we succeed? What happens if this reaches 100 million people? What human behaviour will it encourage? Who gets rich? Who pays the price? What will we wish we had done before success made it too late?

Standing before Rodin’s Gates of Hell, I kept looking at the bodies. They were not customers, users or data. They were human beings, consumed by the very appetites that had once promised them satisfaction. That has stayed with me. I have spent my life trying to succeed. I have tried to build bigger businesses, make better investments and achieve more. I do not regret that ambition. But I increasingly wonder whether we too easily mistake commercial success for proof that what we are doing is right. If it grows, we call it progress. If it makes money, we call it valuable. If technology makes it possible, we call it innovation. And when the human consequences become impossible to ignore, we tell ourselves that nobody could have known, somebody else would have done it anyway or it was perfectly legal.

I no longer think those answers are good enough. The road to Rodin’s Hell is not filled only with evil people making evil choices. The more frightening possibility is that it was built by people just like us, making decisions that seemed sensible, rational and profitable at the time. Maybe the road to Hell isn’t built by people trying to reach Hell. Maybe it’s built by people trying to succeed.

____________________

Guy Hands is the Founder of private equity firm Terra Firma.

LBC Opinion provides a platform for diverse opinions on current affairs and matters of public interest.

The views expressed are those of the authors and do not necessarily reflect the official LBC position.

To contact us email opinion@lbc.co.uk