Artificial-intelligence catastrophism is everywhere. This month saw stock-market ructions after Anthropic, a leading AI company, announced that it had developed and deployed AI “agents” to autonomously execute legal, marketing and sales tasks. The economic potential is real, but the hand-wringing is overblown.
Yes, AI appears to be a transformational technology, and that transformation will sometimes be disruptive. But it won’t render humans obsolete, and the disruption of white-collar work might do a lot of good.
Throughout history, technological change has created new jobs even as it destroys others. The examples are familiar. In the 19th century, while Luddites smashed looms, the weaving industry boomed as prices for cloth plummeted and demand skyrocketed. Though 98 percent of the labor per yard of cloth produced was automated, the number of people working in cloth factories rose sharply, and their wages went up as well.
In the 20th century, ATMs were supposed to make bank tellers obsolete. Instead the number of teller jobs surged. As it became less costly for banks to operate individual branches, banks opened more branches to better serve their clients. AI seems different to many people because it can perform more complex tasks, but there’s no reason to think its implementation will be so different from past advances.
Consider one of the sectors that could be affected by the AI boom: law. On Feb. 3, shares of Thomson Reuters, the owner of Westlaw, dropped by 16 percent, while RELX, which owns LexisNexis, fell 14 percent. The subsequent panic that gripped lawyers themselves was understandable: If AI can automate contract review, legal research and compliance workflows, what happens to us?
But take a step back and ask a different question: What are all those lawyers doing now? The United States has more than 1.3 million of them. In 1970, there were around 1.6 per thousand people. Today it’s four per thousand, double Germany’s rate and 16 times Japan’s. The U.S. population more than quadrupled since 1900; the lawyer population grew more than twelvefold.
But the American legal system did not get twelve times better. What these numbers reflect is a century-long accretion of regulatory complexity that generated enormous demand for people to navigate it. Goldman Sachs has estimated that 44 percent of legal tasks could be automated. And yes, some of those tasks consist of contract drafting, review, legal research and compliance checking — the professional equivalent of hand-weaving cloth.
What computers best automate is drudgery. As the Manhattan Institute’s Judge Glock noted, the information revolution has already changed or eliminated professions such as file clerk, copyist and mail attendant. Automating such tasks drives prices down, which drives up demand for higher-level work.
Unfortunately, America isn’t going to become less litigious. Regulation isn’t going to shrink — if anything, AI itself will help write new bodies of intricate law. What might change is that lawyers will spend less time manually weaving and more time doing work that actually requires judgment and strategy. If that process reduces the share of Americans employed as lawyers back to its 1980s or 1990s level, freeing up talent for other pursuits, that might be a good thing.
The worrywarts, in other words, have it backwards. AI isn’t coming for workers; it’s coming for drudgery. The resulting growth will make Americans better off.
The post What AI is really coming for appeared first on Washington Post.
