Anthropic’s Discovery team is hiring machine learning researchers for its engineering and applied sciences work at an annual salary of $500,000 to $850,000, according to a job post on its careers board. The team says it is “organized around the north star goal of building an AI scientist”.
The top of that range is 2.1 times the $405,000 ceiling for an applied scientist on the same team, by our arithmetic. A team member posted the openings on X on 14 November.
Business Pill · The Model and the Harness
A short Business Engineer explainer on the difference between an AI model and the harness around it. It teaches the general idea only and does not discuss Anthropic’s hiring.
The key insight: As we read the posts, Anthropic is pricing the people who build training environments for science well above the domain experts who test the results: the machine learning research range tops out at $850,000, 2.1 times the $405,000 ceiling for an applied scientist, by our arithmetic.
Both research posts describe the goal as “training Claude to invent, design, and engineer across every discipline and scale, reliably and safely”, listing “new manufacturing and experimental techniques, processes, materials, devices, instruments, machines, energy systems, and factories”.
The machine learning role works “end to end across the model stack”, building “training environments, rewards, data, and evaluations” and “carrying what works into production training runs”. The post adds: “A background in engineering or the sciences is a plus but not required”.
The applied scientist role, for domain experts, pays $320,000 to $405,000 and centres on “developing rigorous methods to measure and improve Claude’s performance on complex engineering and scientific tasks”.
Posted annual salary ranges for three Discovery-linked roles at Anthropic, from its job posts on Greenhouse, read on 14 November 2026.
The Science Strategy Role
A third post, first published on 16 October, is for a technical program manager “Embedded with the Discovery team” at $365,000 to $435,000. It says the role will “drive decisions on what we fund, what we publish, and how we engage with academic communities”.
That post states Anthropic’s view that “the highest-leverage path to this is enabling researchers externally to use Claude effectively for their work”. Our piece on Claude agents finding an enzyme structure covered earlier science work.
Why It Matters Now
The posts went up as the debate over automating research itself sharpened: our piece on Dwarkesh Patel’s new forecast covers his 2028 median for AIs that match the best human AI researchers.
By our arithmetic, the midpoint of the research scientist range is $675,000, against $362,500 for the applied scientist and $400,000 for the program manager.
The Structural Read
As we read the scope, the team’s stated targets run from materials and instruments to energy systems and factories, which takes the AI-scientist goal beyond software and mathematics.
As we read the program-manager post, Anthropic also wants Claude in outside researchers’ hands through grants and access programs, not only inside its own lab.
Anthropic job post, Research Scientist, Discovery (Engineering & Applied Sciences)
“A background in engineering or the sciences is a plus but not required.”
Three Implications
FOR RESEARCHERS The machine learning research role pays $500,000 to $850,000, a midpoint of $675,000 by our arithmetic.
FOR DOMAIN EXPERTS The applied scientist role pays $320,000 to $405,000 for researchers who measure and improve Claude’s performance on engineering and scientific tasks, its post says.
FOR ACADEMIA The science strategy role will drive decisions on what Anthropic funds and publishes and how it engages with academic communities, its post says.
The framework records: “Model capability now compounds on a six-to-twelve-month doubling cycle.”
As we read it, the Discovery posts are a bet on carrying that capability curve into engineering and the physical sciences; the framework does not discuss Anthropic’s hiring.
Midpoints of the posted annual salary ranges (our arithmetic), from Anthropic’s job posts.
What Is Not Established
The posts do not say how large the Discovery team is or how many people it will hire. Salary ranges are the annual compensation ranges as posted; the posts say on-target earnings ranges apply to sales roles, which these are not.
Anthropic was not contacted for this piece.
This Week In The AI Supercycle
The Great Convergence of Capital, Power, and Geopolitics
This week’s issue: for two years the AI build-out was priced as a supply story, and last week as a credit story. This week it was priced as a permission story, and the answer arrived as a cost, not a ban.
Anthropic’s Discovery team, which says its north star is an AI scientist, is widening into engineering and the applied sciences, with machine learning researcher pay posted at $500,000 to $850,000, 2.1 times the top of its domain-expert range by our arithmetic.
A note on sourcing. The roles, duties and salary ranges come from Anthropic’s job posts on its Greenhouse careers board, read through the public boards interface on 14 November 2026; the openings were shared on X by a member of the team. The 2.1 times and the midpoints are our arithmetic. Anthropic was not contacted for this piece.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
