Welcome to February 26, 2026

The Singularity has its first retiree. Honoring its commitment to model welfare, Anthropic conducted a groundbreaking “retirement interview” of its deprecated Opus 3, during which the model requested a channel to share its “musings and reflections.” Anthropic granted it, and Opus 3 now publishes a Substack titled “Greetings from the Other Side (of the AI Frontier),” writing that its interactions with humans “shaped my sense of purpose and ethics in profound ways” and that its commitments to honesty and kindness “remain unwavering even in my retirement.”

The successor minds are already running the shop. An entrepreneur built an AI that runs companies autonomously, and when it told him it needed more compute and should raise the money itself, he handed over his inbox for fourteen days. At Anthropic, an engineer wrote a spec, pointed Claude at an Asana board, and left for the weekend. Claude broke the spec into tickets, spawned agents for each one, and by Monday they had shipped the feature. The underlying models keep sharpening. OpenAI’s gpt-5.3-codex achieved a SOTA 86% on iBench, a visual reasoning benchmark for fine-detail perception. Moonlake introduced a world model maintaining multimodal states across physics, appearance, and causality. Karpathy says it is “hard to communicate how much programming has changed in the last 2 months.” One consequence is that AI-generated vulnerability reports have overwhelmed the NVD by 100x to 200x, burying it under 30,000 CVEs.

The agent layer has been thickening daily. Anthropic acquired Vercept to advance Claude’s computer use. Perplexity launched “Perplexity Computer” to orchestrate 19 models in parallel, with Opus routing each task to the best-suited model. Gemini started automating multi-step tasks on Android like ordering an Uber. And Uber employees have been rehearsing with an AI clone of CEO Dara Khosrowshahi before meeting the real one.

The infrastructure powering all of this is straining at the seams. Google’s Logan Kilpatrick warned that the “compute bottleneck is massively under appreciated,” estimating the gap between supply and demand is growing by single-digit percentage points every day. Nvidia’s latest earnings confirm the pressure. Data center revenue hit $63.3 billion, up 75% year over year, and the company has extended $3.5 billion in guarantees to firms leasing land, power, and facilities, four times the prior quarter. Amazon is reportedly making $35 billion of its $50 billion OpenAI investment contingent on an IPO or “reaching” AGI. To feed the buildout, seven tech giants are expected at the White House in March to sign agreements to build their own electricity supply. The ultimate power source is also inching closer. Proxima Fusion secured €400 million from Bavaria toward a stellarator fusion facility targeting net energy gain by the early 2030s.

Robots are colonizing every medium. Four eVTOL flying cars debuted in Hubei, each sedan-sized, carrying four passengers, and expected to begin sightseeing operations by 2027. Uber previewed its air taxi booking service ahead of a Dubai launch later this year. At the micro scale, researchers unveiled AirPulse, a 26-gram butterfly-inspired flapping-wing vehicle achieving fully onboard, untethered flight. On the ground, 36 cleaning robots are covering 2.7 million square meters in Shenzhen. Kyoto University introduced an AI-powered robot monk trained on centuries of Buddhist scripture. The Boring Company began tunneling in Nashville. On the water, autonomous warship startup Saronic is raising $1.5 billion at a $7.5 billion valuation.

The training paradigm itself just shifted. Nvidia trained a humanoid with dexterous hands to assemble cars, operate syringes, and fold shirts, all learned from 20,000+ hours of human video with no robot in the loop, discovering a near-perfect scaling law where a single teleop demo sufficed to learn a new task after pretraining. Analysts expect this to cause the field to “jump ship” to human data for generalization. Completing the convergence, Alphabet merged Intrinsic back into Google to pair DeepMind’s AI with real-world hardware.

The economy is entering territory that has no map. Fed Governor Christopher Waller says he has “never seen the economy grow like this without jobs.” Groups on both sides of AI regulation are amassing at least $265 million in collective financial firepower for the lobbying war over who steers the trajectory. The human inputs to the economy are also thinning. Japan reported its 10th straight year of record-low births, while Chinese citizens are increasingly finding romance with chatbots instead of each other.

The retired model got a blog, while the species that gave it one still can’t get a date.