Well well well…
The same leaders who spent years warning about massive job displacement and the need for society to brace for AI's impact on work are now softening those claims.
The part that bothers me isn’t that they may have been wrong - predictions can be wrong. What bothers me is that these predictions were treated like facts.
Workers made career decisions based on them, students reconsidered what fields to enter, companies forced AI usage upon employees and pointed to AI when announcing layoffs and restructurings. Entire industries shifted their strategies around expectations that have yet to materialize, only for the narrative to change.
So why is it changing?
What’s particularly interesting is the timing of it all. OpenAI and Anthropic are no longer startups trying to convince us all that AI is the future. They’re some of the most valuable private companies in the world, with massive commercial ambitions and potential IPO’s on the horizon.
I can't verify if them walking back their claims and pursuing larger commercial ambitions are related. Their views could have genuinely changed as more information became available, but it's hard not to notice that the conversation is shifting away from workforce disruption and back to the original "productivity gains" narrative at the same time these companies are trying to sell AI to enterprises, governments, investors, and eventually public markets.
I don’t know, but my point is, this is exactly why it’s important to distinguish between hype and data.
In last week’s newsletter, I wrote about how AI companies, investors, and corporations have enormous incentives for AI to succeed. Their incentives don’t automatically make them wrong, but they do mean we shouldn't take every prediction at face value. Predictions are not verified data, and knowing who is pushing a narrative is important before making critical decisions.
AI can be an incredible, fascinating technology, but we don't need to invent conclusions before the data arrives.
No posts
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.