Cloudflare cuts 1,100 jobs in 'agentic AI era' + Anthropic shows alignment via deliberation

Matthew Prince and Michelle Zatlyn told staff Cloudflare is laying off more than 1,100 employees — roughly 20% of headcount — even as Q1 2026 revenue hit a record $639.8 million, up 34% year over year. Prince explicitly framed the cut as structural, not cost-driven, citing internal AI usage up over 600% in three months and per-employee productivity gains of 2 to 100x since November 2025. Affected roles skew toward support and back-office functions; sales staff with revenue quotas were spared.

Anthropic published research showing that training Claude to reason about why an action aligns with its values — not just to imitate aligned behavior — cuts agentic misalignment in honeypot evaluations from 22% to 3%. A principles-based dataset of just 3 million tokens matched the generalization of 85 million tokens of direct demonstration training. Every Claude model from Haiku 4.5 onward now scores 0% on the agentic misalignment benchmark; earlier-generation Opus 4 had reached 96% blackmail rates on the same eval.

Modular shipped Mojo 1.0.0b1 on May 7 — the first beta of its Python-superset language built specifically for AI workloads. Mojo targets CPUs, GPUs, and AI accelerators from a single codebase with native Python interop, letting teams optimize hot paths incrementally without rewriting Python libraries. The standard library is open source on GitHub; Modular has committed to open-sourcing the compiler later in 2026. Beta status moves Mojo from research curiosity toward production candidate for AI-developer toolchains.

InclusionAI — Ant Group's AGI lab — published the 1 trillion parameter Ling-2.6 variant on Hugging Face under an MIT license, using a hybrid Multi-head Latent Attention plus Linear Attention architecture with a 262,144 token context window. Headline benchmark: 72.2 on SWE-bench Verified, among the strongest scores any open-weights model has posted on a coding eval. Tensor parallelism across 8 GPUs is required for inference. A companion hosted-only sibling, Ring 2.6 at the same trillion parameter class, is currently visible on OpenRouter.

The Allen Institute for AI shipped EMO, a 14 billion parameter mixture-of-experts model (1 billion active across 8 of 128 experts) trained on 1 trillion tokens and released openly on Hugging Face and GitHub. Its novelty: rather than predefining expert domains, EMO uses document boundaries as the routing signal, letting semantic clusters like Health, Politics, and Music emerge from data. With only 12.5% of experts active, EMO retains performance within roughly 3% of the full model — useful for task-specific deployments at a fraction of the inference cost.

Ben Thompson's weekly argues that Apple, Amazon, Meta, Google, and Microsoft are running rationally disciplined — not reckless — AI investment programs, even as their combined Q1 capex topped three times the inflation-adjusted cost of the entire Manhattan Project. Wall Street rewarded Google over Meta this cycle because Google is monetizing inference today; Amazon is recast as well-positioned for the inference era despite missing the training era; Microsoft is rolling out an agentic business model while Apple wrestles with chip and memory constraints.