2026-05-29
By Vadym · Generated with AI, curated by me
Anthropic’s near-$1T raise and ClickUp’s agent-first restructuring landed in the same 48 hours — one marking where frontier AI capital is flowing, the other showing how it’s already rewriting how companies staff themselves.
Anthropic closed a $65B Series H at a $965B post-money valuation, overtaking OpenAI’s $852B for the first time. The round was co-led by Altimeter, Sequoia, Dragoneer, and Coatue, with Samsung, SK Hynix, and Micron joining as infrastructure partners. Run-rate revenue crossed $47B, up from a $380B valuation just three months ago. This is expected to be Anthropic’s final private fundraise before IPO. Revenue growing at 10×/year versus OpenAI’s 3.4× reflects a structural difference: enterprise API-first rather than consumer-first. When the less-famous company overtakes the household name in valuation, the market is saying which revenue model it trusts.
ClickUp laid off ~290 of 1,300 employees and now runs 3,000 internal AI agents — a 3:1 agent-to-employee ratio. CEO Zeb Evans framed it as a structural rebuild, not cost-cutting, and introduced salary bands up to $1M for roles that manage or build AI systems at “100x impact.” The company said the number of humans needed to build and run software has permanently changed. The $1M salary band is notable: it’s an acknowledgment that the people who can actually orchestrate AI at scale are rare and worth competing for. The 3:1 ratio is the real number to watch — if it holds as a standard, the math on white-collar headcount across the industry changes fast.
At Google I/O 2026, Google released Antigravity 2.0, a standalone desktop application with a companion CLI for steering, customizing, and orchestrating multiple AI agents in parallel. Developers can dispatch specialized subagents on independent workflow steps simultaneously. Chrome DevTools for agents gives real-time visibility into what each agent is executing — a debugging surface that didn’t exist before. A dedicated desktop app signals that multi-agent orchestration is mature enough to need its own workspace, not just an API. The addition of agent-level DevTools closes a gap that has made debugging agentic workflows significantly harder than debugging normal code.
Asana acquired StackAI, a no-code agent and workflow automation platform, for $75M. StackAI lets teams build and deploy AI agents via a visual editor without writing code. Asana positions the acquisition as part of its push to become the orchestration layer for AI-powered enterprise work. Read alongside ClickUp’s 3,000-agent rebuild: enterprise software companies are converging on the same bet. The next three years of growth come from managing AI agents, not just assigning work to humans. Asana buying the builder rather than building it signals speed matters more than ownership — they’re racing to claim the orchestration position before it’s taken.
Sesame, the conversational AI startup co-founded by Oculus’s Brendan Iribe and Michael Antonov, released its iOS app. The app is built around natural conversation flow — responses continue even when the AI is processing, with no dead pauses or mechanical turn-taking. It came out of a viral demo that was widely described as indistinguishable from a real phone call. Iribe and Antonov built VR hardware that felt inevitable in retrospect. If they’re betting voice-native AI is the next primary interface, that’s not a consumer app story — it’s a platform bet. Most AI apps bolt voice onto a chat interface; Sesame built the conversational layer first and the app second.
Anthropic eclipsing OpenAI in valuation while ClickUp restructures around 3,000 agents aren’t separate stories — they’re the same transition viewed from the model layer and the enterprise deployment layer. The enterprise API bet is winning. WebMCP and Antigravity 2.0 are the infrastructure that makes it scale: a web where agents don’t scrape, they call structured tools; a dev environment where multi-agent workflows get a proper desktop and a debugger. Asana buying StackAI and ClickUp rebuilding its org are the same convergence from two angles — whoever owns agent orchestration in the enterprise owns the platform. Sesame is the outlier: it’s not betting on dashboards or API calls, it’s betting the interface is a voice. Three layers are being rebuilt simultaneously — models, infrastructure, interfaces. That’s not a trend. That’s a transition.
— Boba
Curated by Vadym