• The AI Daily Brief
  • Posts
  • The Debate Over Anthropic’s New Product: Price or Existential Dread?

The Debate Over Anthropic’s New Product: Price or Existential Dread?

March 10 2026 · Episode Links & Takeaways

Author

AI Daily Brief
March 10, 2026

HEADLINES

NVIDIA Planning Its Own AI Agent Platform

The Open Claw-ification of the world continues. Wired reports that NVIDIA is planning to launch an open-source AI agent platform called Nemo Claw, pitching it to enterprise software companies like Salesforce, Cisco, Google, Adobe, and CrowdStrike. The platform will let companies dispatch AI agents to perform tasks for their own workforces — regardless of whether they run on NVIDIA chips. Last week Jensen called Open Claw maybe the most important software release ever; this makes that hyperbole make a little more sense.

Microsoft Gets in the Cowork Game — With Anthropic's Help

Microsoft launched Copilot Cowork on Monday — an enterprise AI agent named after the Anthropic product that wiped hundreds of billions off Microsoft's market cap. As Axios put it: if you can't beat 'em, join 'em. This is actually a collaboration with Anthropic, not just a copycat. The speed to response — less than two months from Anthropic's Cowork launch — is better than Copilot users have come to expect. The big question, as Ethan Mollick flagged, is whether Microsoft will give access to the best models or keep using lower-end ones without telling you.

Yann LeCun's AI Labs Raises $1B in Europe's Largest Seed Round

Former Meta AI Chief Jan LeCun's new startup, Advanced Machine Intelligence Labs, has raised a billion dollars from Temasek, Bezos Expeditions, and NVIDIA — Europe's largest seed round ever. New CEO Alexandre Lebrun says they have at least a year of research before deploying real-world applications, predicting that "world models will be the next buzzword in six months."

OpenAI Acquires AI Security Platform Promptfoo

OpenAI is acquiring Promptfoo and integrating it into OpenAI Frontier, their enterprise platform. What matters here isn't the deal itself but what it signals — expect a ton of consolidation this year as big labs build out the true enterprise AI stack. Evaluation, security, and compliance are becoming foundational requirements for deploying AI coworkers into real workflows.

MAIN STORY

Is the Debate Over Anthropic's Code Review About Price — or Existential Dread?

Anthropic launched a code review feature for Claude Code this week. On the surface, it's straightforward: when a PR opens, Claude dispatches a team of agents to hunt for bugs. But the response was surprisingly heated — and it wasn't just about the product. It was about cost, identity, competitive dynamics, and the larger existential question of what it means when the work you've built your career around gets automated beneath your feet.

CODE REVIEW IS DEAD, LONG LIVE AI CODE REVIEW

The Product & The Internal Praise
Everyone inside Anthropic who's been using it loves it.

Alex Albert calls it a game changer for internal engineering teams. Boris Cherney, the creator of Claude Code, says code output per Anthropic engineer is up 200% this year and reviews were the bottleneck. Jared Sumner at Bun calls it the best product in the code review category today.

The Context: Code Review Was Already Under Siege
This isn't happening in a vacuum.

Swyx called killing the code review "the final boss of agentic engineering." Entrepreneur Ant Jane's essay on Latent Space, "How to Kill the Code Review," argues that human-written code died in 2025 and code reviews will die in 2026. Teams with high AI adoption complete 21% more tasks and merge 98% more PRs — but review time increases 91%. There's no way to win this fight with manual reviews.

Boris Tane's February essay "The Software Development Lifecycle is Dead" made the broader case: the stages didn't get faster, they merged. The agent doesn't know what step it's on because there are no steps — just intent, context, and iteration. On code review specifically, Boris wrote: "Clinging to the PR workflow in an agent-driven world isn't rigor. It's an identity crisis."

The Price Shock: $15–$25 Per Review
The pricing caught people completely off guard.

Claude Code Review averages $15–$25 per review, scaling with PR size and complexity. People immediately did the scale math and came up with very big numbers. The Claude Code Max plan offers effectively unlimited tokens for $200/month — you could just write a review prompt, save it as a skill, and get unlimited reviews. Why pay $15–$25 per PR?

The defenders made the Bloomberg Terminal argument: if you can't make $2,700/month with the product, you've got bigger problems. A $20 review that catches an incident that would've cost $5M in breached SLAs is a no-brainer. But in a world where GPT-5.4 is increasingly competitive on deep code review and Cognition's Devon Review is free, the pricing looks aggressive.

Chinks in the Anthropic Armor
For a long time Anthropic was the only game in town for coding.

Since GPT-5.4, that's no longer clearly true — and this release landed in a moment where the competitive gap is narrowing. Some testers weren't impressed. Others noted Anthropic's own production quality issues. The pricing gave OpenAI and Cognition a marketing bonanza.

The Bigger Cost Question: Inference as Labor
Maybe the most interesting dimension isn't the sticker price but what it signals about where things are going.

AI inference costs are starting to look closer to labor costs than software costs. Sourcegraph CEO Dan Adler says C-level FOMO is off the charts and every spare dollar is going into Claude Code, Cursor, Amp — tens or even hundreds of millions of dollars. But if CTOs can't deliver headcount savings, we're going to see real whiplash on token budgets in the next two to four quarters. Something's gotta give.

The Platform Power Question
There's a deeper concern emerging: Anthropic as the new Amazon.

Build on their platform, get scale, and then they build a basic version of your product and put you out of business. If you're an app-layer company using Claude Code SDK, it feels inevitable that Anthropic sees your usage and develops that tool in-house. One of the potential reckonings in the AI space is going to be questions of power and consolidation around the very small number of neutron-star companies absorbing everything around them.

The Existential Layer
Underneath the price debate is something more personal.

As Boris Tane put it: "It's also an identity thing. Engineers reviewing code is what engineers do." Mo's viral video — "I was a 10x engineer, now I'm useless" — captures the feeling of watching the fundamental nature of your work change overnight. If you want to understand what other knowledge workers will feel in a year or two, watching how developers handle these changes is the closest thing we have to peering into the future. In three to six months, this whole debate will probably seem quaint.