Generative AI Is a Net Loss for Workers
Demystifying generative AI #3
People are losing their jobs to generative AI. That’s uncontroversial. What’s an open question is whether the jobs generative AI creates or improves outnumber those it destroys or deteriorates, i.e., whether generative AI is a net loss or win for workers all around the world.
I focus here on the three main groups of affected workers—programmers, writers, and artists—but my analysis can be extrapolated to any knowledge-based job that’s done, more or less creatively, in front of a computer.
Generative AI is everywhere—for better or worse
First, the bad news—generative AI is killing many workers’ prospects of a stable, secure future. Copywriters at average medium-size companies are no longer needed; freelance writers’ clients may prefer ChatGPT or Google AI; news outlets like CNET, Insider, and BuzzFeed are doing layoffs; and even a tech juggernaut like IBM intends to replace staff with AI over time.
In some extreme cases, the replacement isn’t partial but complete—with the subsequent inevitable regret. Artists and writers are fighting back, but the determination of employers to tighten the yoke around their necks is unrelenting. And it’s not just a US thing. “Chinese anime and game artists” and “video game illustrators” are being replaced, too.
These are anecdotal accounts that will soon amount to statistical evidence. No need for predictive studies of the number of jobs at risk, or estimations of how AI will affect tasks across sectors. The threat to the workforce is no longer a threat but a tangible, measurable impact assessable with available data. Studies are likely underway.
(I want to take this chance to thank you all for supporting The Algorithmic Bridge. Without you, I know this wouldn’t be possible given the overwhelming competence ChatGPT and other tools entail for independent writers like me.)
Now, the bright side of the story. Both expert and noob coders talk wonders about Copilot and GPT-4. Fiction authors use tools like Sudowrite or plain old GPT-3 to write stories. “AI art” and “AI artist” are labels for people who, like digital artists before, are embracing new tech (e.g., Stable Diffusion and even Disco Diffusion or VQGAN-CLIP, its predecessors) to push the frontiers. People are finding ways for generative AI to increase their productivity and explore new forms of creativity. This is also happening. It’s not a black-or-white matter.