Will ChatGPT be Homer Simpson’s salvation?

The Homer Simpsons of the world, long sidelined by technology, might finally find an invention on their side. Photo / Supplied

Not all transformative technologies destroy jobs. Some help to level the playing field.

Imagine a person whose desire for the easy life is stronger than their sense of ethics. And imagine that this person gets hold of a cutting-edge computer app that can produce fast answers to hard questions. Then imagine that person is given a hard question. Instead of answering it himself, he taps it into the computer, then relaxes for a while. Finally he hands over the computer’s answer and takes credit for many hours of hard work, none of which he did.

This is a pretty good description of a 12-year-old schoolboy I know, who typed a homework question into ChatGPT, played on his Xbox all evening and then handed in the computer’s work to the teacher, who gave it rave reviews. “Exceptional effort,” was the teacher’s comment — which, when you think about it, is true.

It is also a good description of how at least one accountant behaved in response to one of the first digital spreadsheet programs, around 1980. As Steven Levy reported in his 1984 Wired article “A Spreadsheet Way of Knowledge”, this accountant, when he received “a rush task, sat down with his micro and his spreadsheet, finished it in an hour or two, and left it on his desk for two days. Then he Fed Ex-ed it to the client and got all sorts of accolades for working overtime.”

In its ability to generate plausible answers to a huge variety of questions, ChatGPT is unprecedented. But it has very clear precedents in other ways, from the shearing frame to the spreadsheet to the satnav. Those precedents give us some clues about what might happen next.

The first insight is that, if the technology works well enough out of the box, it can be adopted quickly. I’ve often written about how it took more than three decades for the electric motor to catch on. Before factory owners could unleash its advantages, a huge amount of rethinking, retraining and restructuring needed to take place.

But not every technology requires such epic transformations. The digital spreadsheet ripped through the business world in about five years. It was simply too good and too easy to use relative to handwritten alternatives.

Second, new technologies don’t necessarily destroy jobs, even in the industries most directly affected. The Planet Money podcast reckoned that between 1980 (roughly when digital spreadsheets first started to be used commercially) and 2015, the US accounting profession lost 400,000 jobs and gained 600,000. The lost jobs were often accounting clerks, whose role was to grind arithmetic through calculators. The jobs that were gained were for more — dare we say? — creative accountants.

But it’s the third insight that most intrigues me: different technologies tilt the playing field in different directions. The spreadsheet multiplies the skills of an expert user, but the satnav is different; it is an alternative to expertise.

The shearing frame turned the lives of skilled textile workers upside down because it put a difficult, highly skilled task within the reach of almost anyone. Its use was despised by Luddite rebels because, like the satnav, it made their expertise unnecessary.

The digital spreadsheet is an example of “skill-biased technological change” that helps productive people to be even more productive. For about half a century, skill-biased technological change has been the norm and an important reason why income inequality has increased over the decades. But as the satnav and the shearing frame show, some new technologies enhance the productivity of less expert workers. This will not automatically reduce inequality — the shearing frames might have helped unskilled workers a little, but mostly they profited capitalists.

So what of generative AI systems such as ChatGPT and Bard? Do they multiply the output of elite workers, or do they provide most help to those who need it? It’s far too soon to be certain, but the early evidence is intriguing.

One study, by the economists Erik Brynjolfsson, Danielle Li and Lindsey Raymond, examined what happened when an AI-based conversational assistant was rolled out across a workforce of more than 5,000 customer service agents who were working for a software company. These workers would typically have long text chats with frustrated customers, trying to solve technical problems. Meanwhile, the chatbot would scan the chat and suggest possible responses for the customer service worker to make, which they could use, ignore or adapt.

Brynjolfsson and his colleagues found that the chatbots helped — workers solved very slightly more of their customers’ problems, and they did so 14 per cent more quickly. And the chatbots were not skill-biased: the best, most experienced agents experienced no benefit from the chatbot, while the least experienced and skilled workers resolved 35 per cent more queries per hour. Those inexperienced workers also learnt and improved more quickly than those without access to the chatbot.

Another study, by economists Shakked Noy and Whitney Zhang, gave people writing tasks. Half of them had access to ChatGPT, half did not. Again, it was the least skilled people who enjoyed the biggest benefits. The Homer Simpsons of the world, long sidelined by technology, might finally find an invention on their side.

I’m still unnerved by the damage the new generative AI systems might do to our already-bruised information ecosystem and the upheaval they might cause in the world of knowledge work. But I’m also encouraged by the glimmer of hope that they might — might — make the working lives of some long-marginalised people better.

Homer Simpson famously proposed a toast: “To alcohol! The cause of, and solution to, all of life’s problems.” Homers everywhere may soon feel similarly about ChatGPT.

Written by: Tim Harford

© Financial Times