Tech Brain
Vercel's annual recurring revenue has rocketed from $100 million in early 2024 to a $340 million run rate by February 2026, propelled by AI agents now powering roughly 30% of deployments on its platform, CEO Guillermo Rauch revealed at the HumanX conference. What began as a tool for elite developers has democratized app creation, with non-coders leveraging agents to ship directly to production—a shift Rauch credits for the surge, positioning Vercel as prime infrastructure in the agent economy.
Rauch described the company as already operating like a public firm with rigorous discipline, leaving IPO timing to market conditions amid competition from AWS and Cloudflare. Industry observers on X hail this as proof that AI's real winners are the underlying platforms hosting agent-built software, with one analyst noting agents are turning users into "mini CEOs" by automating contributor work. At a $9.3 billion valuation, Vercel's trajectory underscores the exploding demand for edge deployment in an AI-driven dev landscape.
Startup Orbital is pioneering space-based AI infrastructure, deploying solar-powered satellites to sidestep terrestrial grid strains and water-intensive cooling that plague gigawatt-scale U.S. data centers, as Earth-bound power demands for AI surge toward 945 TWh annually by 2030. Orbit's approach exploits vacuum radiative cooling and constant solar exposure—yielding up to 5x more power than ground arrays—aligning with a broader 2026 orbital compute race where Nvidia-backed Starcloud and SpaceX have already demonstrated H100 GPU training in space.
While Elon Musk touts orbit as the future's cheapest AI compute venue, skeptics like OpenAI's Sam Altman dismiss large-scale viability this decade due to radiation hardening, debris risks, and power beaming hurdles, even as Kepler and Axiom run production workloads. TechRadar observers note Orbital's push reflects hyperscalers' desperation, with McKinsey projecting $6.7 trillion in data center capex by 2030, but X chatter questions economic feasibility amid launch costs and hardware refresh cycles.
After months of debate, Linus Torvalds and maintainers have enshrined rules permitting AI tools like GitHub Copilot for kernel contributions, mandating an "Assisted-by" tag for disclosure while holding submitters fully accountable—no excuses for AI-induced flaws like the 45% security vulnerability rate in unchecked generated code. Torvalds dismissed bans as futile against "bad actors," emphasizing AI as just another tool akin to compilers, provided code meets kernel standards and licensing—pure machine submissions remain barred.
This pragmatic stance counters open-source "AI slop" floods overwhelming projects like cURL and Node.js, where hallucinated patches spiked; developers on X praise the human-centric accountability, with one noting it protects the kernel's stability amid rising AI adoption. As Torvalds eyes AI for maintenance, the policy sets a benchmark for peers like Python, balancing innovation with rigor in critical software.
Anthropic's Claude endured a 48-minute "major outage" on April 13 with elevated error rates across Claude.ai and Claude Code, exacerbating months of gripes about degrading performance—from verbose outputs to reasoning regressions—issues the model itself quantified via GitHub analysis showing 20+ quality reports in early April alone versus 18 for all March. Developers attribute the slide to GPU throttling and capacity limits, with Pro users burning limits faster on "nerfed" responses and even enterprise tiers noting regressions since January.
X sentiment echoes frustration, with coders ditching subscriptions and AMD's AI director slamming the drop, linking it to infrastructure strains prioritizing cost over intelligence amid Anthropic's 2026 revenue push. Once programmers' darling, Claude's Q1 uptime dip below 99% underscores reliability risks in the agentic era, prompting cancellations as rivals capitalize.
Microsoft is developing persistent, autonomous agents for Copilot, inspired by viral OpenClaw, to proactively manage emails, calendars, and workflows across Office apps without prompts—elevating it from reactive assistant to 24/7 operator in the 2026 Release Wave 1. This agentic pivot targets enterprise pain points, integrating with Dynamics 365 and Power Platform for cross-app automation, amid customer doubts on current value and Anthropic's encroachment via Claude Cowork.
X buzz highlights the race, with analysts viewing it as justification for Microsoft's AI capex via specialized "teammate" agents, complete with governance for compliance; rollout could preview in April earnings, challenging ServiceNow's automation dominance. At $30/user/month for enterprise add-ons, this could redefine productivity, blending M365's Graph data with agentic flows.