The SaaS sell-off accelerates, the new space race is for data centers, and AI is intensifying work, not reducing it

Software stocks take more damage, the UAE positions as a third AI superpower, and a Berkeley study shows AI is making us work more, not less.

The market is starting to price in what AI actually means for software incumbents, the geopolitics of compute is getting weirder, and we now have hard data on how AI changes day-to-day work.

The 'SaaS-pocalypse' wipes out hundreds of billions in value

My take: Monday.com down 21%. Thomson Reuters down 20% despite strong earnings. Schwab and Raymond James both down 7%+ as an AI-powered fintech called Altruist eats into wealth management. The disruption is no longer limited to software. Any business with seat-based pricing or repetitive professional work is now in the blast radius. Databricks is the counter-example: AI-native companies are seeing huge tailwinds.

ByteDance releases Seed Dance 2.0, a new state-of-the-art video model

My take: Native audio-visual co-generation, 15-second 2K clips with proper lip-sync. The quality has crossed a real threshold. For anyone running marketing, sales enablement, or training content, the cost of producing video is about to collapse. Plan to do 10x more video this year, not the same amount cheaper.

The new space race is for data centers in orbit

My take: Both Musk and China's state space contractor are now planning gigawatt-class orbital data centers. Sounds far-fetched, but the underlying constraint is real: we can't build enough power on Earth fast enough. Watch this for what it tells you about how seriously the major players are taking the compute crunch.

UAE positions as a 'third AI power'

My take: Abu Dhabi's G42 announced the world's largest AI chip (with Cerebras) and a billion-dollar data center in Vietnam. OpenAI is reportedly building a UAE-specific ChatGPT. If you do business in the Middle East or Southeast Asia, the AI infrastructure landscape there is no longer just a US-vs-China story.

AI is making employees work more, not less

My take: A Berkeley Haas study published in HBR found that early AI adopters experience 'work intensification.' People take on new roles (designers coding, marketers doing analytics), blur work/personal boundaries, and multitask with agents running in the background. The promise that AI would give people their time back hasn't materialized. If you're rolling out AI tools, factor in workload, not just productivity.