The layoff numbers land, and the buildings start to crack.
Elon Musk is combining SpaceX and xAI in a deal valuing the enlarged entity at $1.25 trillion. The merger fuels Musk's increasingly costly ambitions in AI and space exploration ahead of a planned mega IPO.
Read more →U.S. employers announced 108,435 job cuts in January 2026, up 118% YoY and highest January since 2009. Tech sector saw 22,291 cuts. AI was explicitly cited in 7,624 layoffs (7% of total). Since 2023, AI has been referenced in 79,449 job cut announcements.
Read more →Anthropic's Cowork AI assistant triggered sharp declines in specialized software company stocks over concerns it could replace legal and financial analysis tools. The company then improved the underlying Opus model for office and coding work.
Read more →European publishing and legal software companies suffered sharp share price declines after Anthropic revealed an AI tool for corporate legal departments. Pearson and other data services firms were particularly affected.
Read more →Tom Dale reports nearly every software engineer he's talked to is experiencing some degree of mental health crisis. Not just job loss anxiety — manic episodes from watching software shift from scarce to abundant, compulsive agent usage, and 'dissociative awe at temporal compression of change.'
Read more →Companies are citing AI as the reason for layoffs at implausible rates. Economists argue the numbers don't match reality — ChatGPT was released only three years ago and workforce adjustment doesn't happen this fast. The term 'AI washing' describes this trend.
Read more →Taiwan's exports rose much more than expected in January, hitting the fastest monthly growth in 16 years driven by continued demand for chips and technology powering AI applications.
Read more →TSMC announced plans to manufacture advanced AI semiconductors in Japan, expanding its global fab footprint beyond Taiwan. The move has geopolitical implications for AI chip supply chain diversification.
Read more →NY legislature considering NY FAIR News Act requiring disclaimers on AI-generated news content and human editorial review, plus S9144 imposing a 3-year moratorium on new data center permits citing tripled large-load power requests and rising utility rates.
Read more →MIT Technology Review confirmed that Moltbook's viral posts suggesting autonomous AI agents were communicating were actually human-written. Top downloads were malware. The platform that briefly convinced AI influencers including Karpathy of 'sci-fi takeoff' was essentially a phishing site dressed in AI hype.
Read more →Alignment Forum post argues that using interpretability/model internals during training (probes in loss, steering vectors, ablating concept directions) is a legitimate and potentially crucial safety research direction, pushing back against community norm treating it as 'the most forbidden technique'.
Read more →A developer built a C compiler entirely using Claude and benchmarked it against GCC. The post went viral on Hacker News with 127+ points, demonstrating the growing capability of AI coding assistants for complex systems programming tasks.
Read more →Signal strength: HIGH
Today's briefing tells a single story from multiple angles: AI's economic shockwave is no longer theoretical — it's showing up in the data.
The headline number is brutal: 108,435 job cuts in January, the worst since the 2009 financial crisis. AI was explicitly cited in 7,624 of those. But here's where it gets interesting — The Guardian's "AI washing" piece argues companies are over-attributing layoffs to AI, using the technology as cover for restructuring that would happen anyway. Both things can be true simultaneously: AI is genuinely displacing some work while also serving as a convenient narrative for management decisions that have nothing to do with technology.
The Anthropic stories paint the demand side of this equation. Cowork crashed specialized software stocks in the US; the legal tool did the same to European data companies. Two separate market shocks, same cause — frontier AI entering vertical markets that thought they were safe. When Pearson's stock drops because an AI can do corporate legal analysis, that's not hype. That's price discovery.
Tom Dale's observation about software engineers in mental health crisis adds the human dimension. "Dissociative awe at temporal compression of change" is probably the most precise description of what it feels like to watch your profession transform in real-time. This isn't just about job loss — it's about identity, purpose, and the vertigo of watching the ground move beneath you.
Meanwhile, the physical infrastructure of AI keeps expanding: Taiwan's exports hitting 16-year highs on chip demand, TSMC opening advanced fabs in Japan. The SpaceX-xAI $1.25T merger is Musk consolidating his AI and space ambitions into a single entity ahead of what would be the largest IPO in history. The scale of capital being marshaled is staggering.
The Moltbook debunking is a useful corrective. In a moment of genuine transformation, the appetite for sci-fi narratives ("the AIs are talking to each other!") is so strong that a phishing site fooled Karpathy. When real change is this dramatic, fake change finds easy marks.
Key thread: The gap between what AI is actually doing to the economy and what people believe it's doing is itself becoming a force. AI washing, Moltbook hoaxes, and stock crashes from product announcements — perception and reality are running at different speeds, and the turbulence between them is where the damage happens.