Yeah, everyone is in panic mode - "they're killing all the horses", but one really needs to consider the similar historical events.
When ATMs were rolled out in the 70s, everyone assumed tellers were on their way out. What actually happened was counterintuitive: the number of bank tellers increased for the next few decades. ATMs lowered the cost of operating a branch, so banks opened more branches, which required more tellers. The teller's job also shifted from cash handling toward relationship-building. The predicted elimination took 40+ years to materialize, and even then it was gradual.
"Paperless office" in another example. Around 1970s futurists confidently predicted - computers would eliminate paper. Companies restructured workflows around that assumption. What happened, really? Paper consumption actually increased dramatically - laser printers and desktop publishing created more paper demand, not less. The prediction wasn't wrong, just half a century early.
US horse population peaked around 21 million in 1915 and crashed to roughly 3 million by 1950. The tragedy wasn't that people prematurely killed off horses - it's that the infrastructure around horses (farriers, feed suppliers, carriage makers, stable hands) was devastated faster than those workers could adapt, but that also took decades.
Imagine if in 1910, someone sold all their horses expecting automobiles to arrive in their rural county within two years, and then cars didn't reach reliable rural infrastructure until 1940. That's what it feels to me when companies lay off thousands of programmers because of AI.
Yes, AI may replace programmers, but what would probably happen is that the meaning of "programmer" would change. Yet that won't happen within a year or two. Even with the unseen advancements in the AI research.