AI Could Spell the End of Big Business

Big business has been a fixture of American life since the late 19th century, and today more Americans work for big companies than for small ones. That could be about to change — in part because of the rise of artificial intelligence.

Consider the most prestigious service that generates images using AI, a company called Midjourney. It has a total of 11 full-time employees. Perhaps more are on the way, but that is remarkably few workers for a company that is becoming widely known in its field.

Part of the trick, of course, is that a lot of the work is done by computers and artificial intelligence. I don’t think this will lead to mass unemployment, because history shows that workers have typically managed to move from automating sectors into new and growing ones. But if some of the new job-creating sectors are personal services such as elder care, those jobs are typically in smaller and more local firms. That means fewer Americans working for big business.

Or consider ChatGPT, which has been described as the most rapidly growing consumer technology product in history. It is produced by OpenAI, headquartered in San Francisco. By one recent estimate the company has about 375 employees. By contrast, Meta, even after some layoffs, currently has more than 60,000.

OpenAI will make more hires. Still, it may be time to reset expectations for what a major tech company looks like.

To the extent other companies use AI to shrink their headcount, the political power of big business may decrease. Companies may not have the staff or bureaucracy to support their lobbying efforts, the way that Exxon or General Motors used to. Nor will these new smaller companies be able to go to their two senators and ask for favors because they employ tens of thousands of people in the state.

What will be the effect on public policy? Overall, there might be less political support for a broad increase in immigration, which big business has long favored. Instead, immigration policy could become narrower and more targeted, to identify people who could contribute to tight, highly productive teams in these new smaller businesses.

Important businesses with a small number of employees will need to hire exactly the right people to oversee the larger automated and AI-managed superstructures. Talent selection will rise in importance as a skill, and the people in these smaller units will command very high compensation. Even if those companies do not become formal partnerships, their internal cultures will adhere more closely to those of partnerships or sports teams, with a good deal of mutual internal monitoring. That too will represent a workplace evolution.

Investing will evolve as well, with Americans less likely to own businesses in their stock portfolios. If businesses become smaller overall, they will not have the same need to go public to raise capital for expansion. They will be privately held. Venture capital will rise in importance, but since most solo investors face SEC restrictions on making venture investments, they will have to put their money elsewhere.

Of course not all large companies will shrink. Some customer-service companies, such as Starbucks, may not be able to automate their core activities, and they rely on national branches, which means they still will have a large number of employees (the current estimate is 400,000). But many Starbucks employees age out of their jobs and enter other segments of the labor force. Working for big business might become something people do when they are young and then leave, never to return.

Imagine yourself at a dinner party in the future talking to someone who seems so interesting and unusual, even exotic, because they work for big business. Maybe it would be the equivalent of talking today to someone who is a spy or works on an oil derrick.

It may be hard to imagine now — especially if you happen to have written a book all about how important big business is to US productivity and culture. But in the not-too-distant future, big business may become more mysterious to many Americans. It is yet another unexpected consequence of the rise of AI.