Version 1

Anthropic is in a rapid dual expansion: scaling compute infrastructure and sharpening its philosophical identity as an AI lab. On the infrastructure side, Anthropic signed a deal with SpaceX for full access to Colossus 1, a 300+ MW Memphis data center with 220,000+ NVIDIA GPUs [1], while doubling Claude Code rate limits across all paid tiers [1]. On the philosophy side, a public debate has crystallized around whether Anthropic builds 'moral agents' (its own framing) or whether OpenAI's competing 'Tool AI' positioning is even coherent [2]. The company reportedly grew 80x in a single quarter [1], making the compute crunch and recent Claude Code outages an acute operational story.

Anthropic is stress-testing what it means to scale an AI lab that explicitly treats its model as a moral agent rather than a tool — at the same time it is acquiring GPU capacity at a pace that few companies in history have matched. How it resolves the tension between that philosophical stance and the commercial pressures of explosive growth will shape not just Anthropic, but industry norms for how frontier AI systems are framed and governed.

Anthropic held a developer event in San Francisco — 'Code with Claude' — where it announced a sweeping compute and product expansion [1]. The headline infrastructure deal pairs Anthropic with SpaceX, granting access to Colossus 1, a Memphis facility exceeding 300 MW with more than 220,000 NVIDIA GPUs expected online within weeks [1]. The announcement was a direct response to months of degraded Claude Code performance: Anthropic CEO Dario Amodei attributed the problems to 80x quarter-over-quarter growth that outpaced the company's ability to provision capacity [1]. As an immediate remediation, Claude Code rate limits were doubled for Pro, Max, Team, and Enterprise subscribers, and the peak-hours throttling previously applied to Pro and Max tiers was removed [1].

Beyond raw capacity, Anthropic previewed three research-stage features under the 'Claude Managed Agents' umbrella: multi-agent orchestration (Claude instances coordinating with one another), Outcomes (goal-directed iteration where Claude works toward a defined end state), and Dreaming (overnight autonomous self-reflection and memory writing) [1]. These features mark a deliberate push toward persistent, goal-pursuing AI systems — a posture that sits at the center of a simultaneous philosophical debate about what kind of entity Anthropic is actually building.

Analyst Zvi Mowshowitz surfaced and synthesized a Twitter debate probing that question directly [2]. The debate contrasts OpenAI's public 'Tool AI' framing — positioning its models as instruments that execute user intent without independent agency — against Anthropic's explicit design choice to give Claude a moral character capable of refusing instructions, including those from Anthropic itself [2]. Zvi argues that 'Tool AI' is ultimately self-undermining: agentic use is more useful, and users and markets will push any sufficiently capable model in that direction regardless of rhetorical framing [2]. His sharper observation is that Anthropic may require entirely new conceptual vocabulary — Claude is, in his framing, 'not person, not tool, not deity, not pet' [2].

The SpaceX partnership adds an ironic geopolitical footnote: the same Elon Musk whose Pentagon-aligned allies reportedly labeled Anthropic a 'supply chain risk' last quarter is now supplying the company's most critical compute infrastructure [1]. That Anthropic now accounts for more than 40% of Google Cloud's revenue backlog [1] while simultaneously signing with SpaceX underscores how quickly the lab's infrastructure dependencies — and the political entanglements that come with them — are multiplying.