Block and Meta Just Said the Quiet Part Out Loud — And That's Why Our 70% Holds
TexTak forecasts a 70% probability that the first major layoff wave explicitly attributed to AI automation is either already underway or imminent — and this week's news moved the needle on the most stubborn part of our thesis. The attribution barrier, not the displacement itself, has always been our forecast's load-bearing variable. Jack Dorsey just kicked that barrier down. Meta's 8,000-person cut, structured explicitly around AI capability replacement and accompanied by AI agent training infrastructure, is doing the same. Two high-profile executives, in the same week, saying publicly what every CFO has been thinking privately.
Our 70% has always rested on a specific distinction that most AI commentary collapses: the difference between displacement happening and displacement being acknowledged. Companies have had every incentive to describe layoffs as 'restructuring,' 'efficiency optimization,' or 'portfolio realignment' — anything but 'the AI did it.' The PR risk has been real. The legal exposure under WARN Act and NLRA frameworks is non-trivial when you tie headcount reduction to technology replacement. And the cultural cost of telling employees their jobs were eliminated by software is enormous.
This week broke that pattern in two places simultaneously. Block's Jack Dorsey didn't soften it — he said AI models proved they could do the work 'more honestly' than human employees. That's not restructuring language. That's direct attribution. Meta's announcement is structurally more significant: 8,000 jobs eliminated while the company simultaneously installs employee tracking software explicitly to train AI agents. The causal chain is visible in the corporate action itself, not just the press release framing. We weight these heavily because they're not anecdotes — they're the first high-profile instances of the specific behavior our forecast targets.
The S&P 500 data is circumstantial but contextually important: 400,000 jobs cut across the index in 2025, the first net decline since 2016. The Tech Insider analysis suggesting 47-50% of the 150,000 tech layoffs year-to-date are explicitly attributed to AI is the most direct supporting data point we've seen — though we'd flag that 'analysis suggesting' is doing significant work there. We're treating it as proximate evidence: consistent with our thesis, not proof of it. The number we're actually watching is how many Fortune 500 earnings calls in Q2 include language connecting headcount reduction to AI capability replacement with specificity.
The honest counterargument is that Dorsey and Zuckerberg are outliers, not trend-setters. Most Fortune 500 CEOs are not going to stand up on an earnings call and say 'AI took these jobs' — the litigation risk alone makes that unlikely. Attrition-based displacement, where companies simply don't backfill roles that AI now covers, will remain the dominant mechanism and will remain largely invisible in attribution data. Our 70% reflects our judgment that explicit public attribution will reach threshold scale — but we should be honest that 'threshold scale' is doing some definitional work. What would move us above 80%: three or more S&P 100 companies explicitly connecting AI to headcount in Q2 earnings guidance with specific role category language. What would drop us below 55%: the Block and Meta attributions prove to be isolated cases with no follow-on from other major employers through Q3.