The agentic era arrives, and the 'SaaS-pocalypse' begins

Wall Street starts to seriously question SaaS business models, autonomous agents go mainstream, and an OpenClaw agent social network exhibits some genuinely strange behaviour.

A bigger-than-usual edition because there was a lot to cover. The ‘is SaaS dead?’ debate kicked off this week, and the answer is more nuanced than the headlines suggest.

The 'SaaS-pocalypse' on Wall Street

My take: Salesforce, Snowflake, and other software stocks are seeing significant sell-offs as investors worry AI agents will replace traditional per-seat SaaS. The smart take, from Jensen Huang and others, is that this is too simple. True AGI would use existing software, not reinvent it. What's actually changing is the SaaS pricing and growth model, not its existence. If you sell software, your pricing strategy needs a 2026 overhaul.

Personal AI agents arrive with OpenClaw

My take: An open-source project for running 24/7 personal AI agents on a Mac Mini has exploded. People are using these to manage calendars, fix code, handle customer support workflows, all running locally. The interesting bit for businesses: you'll soon have employees with much more capable personal AI infrastructure than what the company provides. Get ahead of this with proper enterprise tooling, or watch shadow AI bloom.

OpenAI launches a desktop 'agent command center'

My take: The new Codex app is positioned to orchestrate multiple parallel agents on long-running tasks. It's a direct shot at Anthropic's Claude Code. The competition between these two is going to define the next 18 months of enterprise AI tooling. Pick a primary, but stay flexible. Don't lock in single-vendor.

Anthropic's Super Bowl ads mock OpenAI's ad plans

My take: Tagline: 'Ads are coming, but not to Claude.' This is the first time we've seen the labs go at each other in mainstream advertising. OpenAI's leadership responded aggressively on social, which was widely seen as uncharacteristic. The brand wars are starting, which means real differentiation, which means real choice for buyers.

SpaceX acquiring xAI in a $1.25T deal

My take: Officially about orbital data centers, but mostly a financial structure that lets xAI ride a future SpaceX IPO. Worth noting because it consolidates Musk's AI bets and gives xAI access to far more capital. Whether you trust Musk or not, expect Grok to keep showing up in the model conversation.

Memory chip shortages will get worse before they get better

My take: Samsung and SK Hynix's profits doubled YoY on AI memory demand. Prices are projected to keep rising through 2026. If your business buys servers, laptops, or any hardware with significant memory, lock in pricing now. The shortage is structural, not cyclical.