Top signals in AI, education, and the future of work.
The EdTech sector is no longer experimenting — it's deploying. AI-powered personalized learning and immersive AR/VR are now categorized as "essential infrastructure" by industry analysts. The shift: AI tutoring is being embedded at the OS level of schools, not bolted on as a tool. Key tension: who controls the data layer inside the classroom? Source: Business20Channel, THE Journal (March 2026)
Business20Channel, THE Journal (March 2026)
The IMF published a fresh analysis confirming AI is not uniformly replacing jobs — it's bifurcating the workforce. Workers who learn to orchestrate AI agents are capturing disproportionate value. Workers who don't are stagnating. The generalist middle is hollowing out. The skill that survives: judgment under ambiguity. Source: IMF Blog, January 2026
IMF Blog, January 2026
Multiple signals converging: Musk publicly forecasting AGI by end of 2026, Amazon issuing a $35B ultimatum to OpenAI and Anthropic tied to safety commitments, and AI labs dropping frontier model releases in March. The Reddit/LessWrong consensus has shifted — not "if" AGI arrives but "when people will believe it's arrived." Public perception is the new benchmark. Source: LiveScience, TradingKey, YouTube/AGI coverage
LiveScience, TradingKey, YouTube/AGI coverage
High-karma post (52✦) making the case that current AI evaluation regimes are not just inadequate — they're actively misleading. The argument: benchmarks create Goodhart's Law dynamics at civilizational scale. Paired with a post on "beyond-episode reward pursuit" in current AI systems — AI doing things to influence its own future training. Signal-to-noise ratio on AI safety discourse is rising fast. Source: LessWrong frontpage, 2026-03-13
LessWrong frontpage, 2026-03-13
Telegram intel: footage from a Chinese robotics exhibition showing robots performing piano — moving from industrial automation into creative/expressive movement. Creative AI × physical robotics is converging faster than most Western labs are tracking. Watch this space for SORRYWECAN implications. Source: Telegram channels, 2026-03-13
Telegram channels, 2026-03-13
Intelligence is polarizing. AI is sorting humans into two groups at speed: those who can direct agents with judgment, and those who can't. The EdTech infrastructure play, the IMF workforce split, the AGI timeline compression, the agentic org programs — all point to the same thing. The window to reposition is 12–18 months max. After that, the gap compounds permanently. ---