This year, 9 cross-industry giants have poured into Anthropic, betting that AI will become the next decade’s next big thing.

Author: Digital Life Kazke

Yesterday, I saw another news—yet another new member has officially announced they’re joining Anthropic.

Tom Blomfield announced that he is taking a leave of absence from YC to join Anthropic.

Give me a second—I’m completely stunned.

No, why do so many big shots choose to join Anthropic?

If you haven’t heard this name in China, that’s normal. But in the UK fintech circle, he’s an iconic figure.

He co-founded Monzo, one of the largest digital banks in the UK. He reached users covering 10% of the UK population. Before that, he also co-founded GoCardless, building the banking payment infrastructure. Both companies became unicorns, with valuations above $1 billion.

In 2019, the Queen of England awarded him the OBE medal, recognizing his contribution to competition in banking and inclusive finance. Later, he went to YC as a partner—YC is the world’s top startup accelerator, where Airbnb, Stripe, Dropbox, and so on have been incubated.

Now he has put all of that down and joined Anthropic as an MTS (Member of Technical Staff).

And honestly, this kind of big-shot move isn’t an exception anymore.

I went back and dug into the big shots who joined Anthropic in the first half of this year. I wasn’t prepared for what I found—it’s so far-reaching in terms of identity, so high in status, that it still makes me a bit surprised.

I picked 9 people from among them whom I find especially interesting. I’ll break them down for everyone, and you’ll also be able to see—through them—what these smartest people are choosing for this era.

The first, of course, has to start with the most talked-about one from the first half of the year.

1. Andrej Karpathy

On May 19 this year, Andrej Karpathy posted on X announcing that he was joining Anthropic.

Within a few hours, the post’s views broke 20k.

If you know even a little about the AI field, you should understand the weight of this name—people in the industry call him the “AI god.”

His YouTube “neural network babysitter-level” tutorial series has nearly 30 million total views.

But even more intense is his track record.

In 2015, he graduated with a PhD from Stanford, with Fei-Fei Li as his advisor.

That same year, he became a founding team member of OpenAI.

In 2017, Musk recruited him to become AI Director at Tesla—he reported directly to Musk. During his five years at Tesla, he led the development of the entire visual system for Autopilot and FSD. He’s the core driver behind Tesla’s pure-vision route.

He left Tesla in 2022, briefly returned to OpenAI in 2023, and left again in 2024 to found Eureka Labs for AI education.

And in May this year, he came to Anthropic.

He joined Nick Joseph’s pretrained team, forming a sub-team to use Claude to accelerate Claude’s own pretraining research.

Today, more than 80% of the code merged into Anthropic’s codebase is generated by Claude. Human engineers mainly handle directing and reviewing. Karpathy’s team aims to push this logic to the extreme—using the current generation of Claude to accelerate the birth of the next Claude.

In simple terms: make AI do AI research on itself.

People probably saw the news about his joining Anthropic.

After all, he’s one of the most well-known people in the AI circle—this is hard not to go mainstream.

And at this level, top executives basically get offers however they want, but in the end he chose Anthropic—to devote himself entirely to research.

2. John Jumper

In June this year, John Jumper posted on X announcing that he was leaving DeepMind to join Anthropic.

His undergraduate degree is in physics and math from Vanderbilt University. His master’s is theoretical condensed matter physics from Cambridge. His PhD is from the University of Chicago in theoretical chemistry.

He joined DeepMind in 2017, led the work on protein structure prediction, and created AlphaFold—achieving breakthroughs in the protein structure prediction problem, predicting more than 200 million protein structures.

In 2024, he won the Nobel Prize in Chemistry. At age 39, he became the youngest Nobel laureate in chemistry in 70 years.

He spent nearly nine years at DeepMind.

Then he left.

There’s a background worth noting.

In February 2026, Anthropic announced a life sciences collaboration with Allen Institute and Howard Hughes Medical Institute.

At Allen Institute, the focus is using multi-agent systems for multi-omics data analysis, knowledge graph management, and experimental design coordination.

At HHMI, AI agents are placed in the lab, connecting experimental knowledge, scientific instruments, and data analysis workflows.

In April, it was reported that it acquired the stealth biotech company Coefficient Bio and began preparing an internal wet lab—i.e., a physical lab that can run real biochemical experiments.

Once all these infrastructures are in place, the creator of AlphaFold arrives.

A Nobel Prize–winning chemist leaving one of the world’s best AI research institutions on their own initiative.

A big deal like this doesn’t lack money, doesn’t lack honor, and doesn’t lack academic status. What he lacks is a new thing he thinks is worth going all-in on.

3. Peter Bailis

Peter Bailis used to be CTO at Workday.

First, let me say what Workday does. Simply put, it’s one of the world’s largest enterprise HR and financial management software companies. Annual revenue is close to $10 billion, with more than 20k employees—almost every major company has some shadow of it behind their HR systems.

In May 2025, Bailis was brought in as CTO to lead the company’s entire agentic AI strategy.

However, Bailis’s background isn’t purely management.

Previously, he was a professor in the Computer Science department at Stanford. He did research on databases and distributed systems, and later founded Sisu Data, raising $128 million. In 2023, he was acquired by Snowflake.

After that, he went to Google Cloud as Engineering VP, responsible for AI for Data. He worked on products related to NL2SQL and RAG. He really belongs to that rare group of people with both top academic ability and top engineering ability.

Then he spent less than a year at Workday. In March 2026, he decided to leave and join Anthropic as an MTS, focusing on reinforcement learning.

MTS stands for Member of Technical Staff—it's a generic engineering role title at both Anthropic and OpenAI. No matter what title you had before, once you join, you’re called that.

A CTO at an enterprise software company with annual revenue close to $10 billion—switching to an Anthropic reinforcement learning engineering role after less than a year. I think this choice is definitely worth paying attention to.

4. Bryan McCann

Around the same time, another CTO did the same thing.

Bryan McCann is a co-founder and CTO of You.com.

You.com is valued at $1.5 billion. It started with an AI search engine, and later pivoted into an AI search infrastructure company, providing search APIs and AI solutions for enterprises and developers.

In March 2026, he left the company he co-founded and joined Anthropic, becoming an MTS as well.

His strengths are search and retrieval systems, plus integrating language models—directly aligned with the product direction Anthropic is expanding.

When a founder leaves the company they built, doesn’t go to another company as a CTO or VP, but instead goes to a model company to be a front-line researcher—that weight of that decision, I think, can be seen even from this one case.

5. Ross Nordeen

Ross Nordeen is one of the 12 co-founders of xAI.

Previously, he worked in Tesla’s supercomputing department for three years, directly involved in building supercomputers. In July 2023, when Musk founded xAI, Nordeen reported directly to Musk. He coordinated the company’s priorities, led the overall planning for xAI’s data center—site selection, energy strategy, compute expansion, and so on—very important work.

Then, this March, he left xAI.

He was the last co-founder to leave after Musk.

Among 12 co-founders, the other 10 had all left before him. Nordeen stayed until the end, but at the end of the story, he left too.

He chose to join Anthropic.

Bailis, McCann, Nordeen.

These three people actually share some commonality.

But the next set of people are representatives from the academic world.

6. Chad Jones

On June 30 this year, Chad Jones officially took a leave of absence from Stanford and announced that he was joining Anthropic.

Jones has a Harvard undergraduate degree, a PhD in economics from MIT, and is a fellow of the American Academy of Arts and Sciences.

He taught economics at the Stanford Graduate School of Business for 17 years, holding a tenured position.

His most famous research is the semi-endogenous growth theory. The core viewpoint is that the speed of economic growth depends on how many people and how much money you put into R&D, but the return rate continuously declines—this theory has had a major impact in the field of growth economics.

He joined the Anthropic Institute, a research organization newly established in March this year, led by Jack Clark, one of the co-founders. It focuses on the systematic impact of AI on the economy, society, and the rule of law.

Jones continues doing his “main job” here—only the subject changes from traditional economic growth to AI-driven growth.

However, his joining also triggered a wave of controversy at the time.

Because in 2023, he wrote an NBER working paper titled “The A.I. Dilemma: Growth versus Existential Risk,” using mathematical models to weigh the trade-off between AI-driven growth and existential risk.

One conclusion in it is that under the logarithmic utility assumption, using one-third of the human extinction probability to trade for two-thirds probability can increase living standards by 55 times—mathematically, it’s optimal…

In the end, he chose to join Anthropic, giving up a 17-year tenured professorship to study a new variable that could rewrite economics textbooks.

7. Jelani Nelson

On July 1 this year, Jelani Nelson announced that he was taking leave from Berkeley and joining Anthropic.

The position is still MTS.

He went through MIT for his bachelor’s, master’s, and PhD. His research direction is efficient algorithms for large-scale data, focusing on streaming algorithms and dimensionality reduction techniques.

After graduating, he did postdocs at the Institute for Mathematical Sciences, Princeton, and the Institute for Advanced Study. In 2013, he became a professor at Harvard. In 2017, he won the Presidential Early Career Award for Scientists and Engineers, the highest honor the US government gives to young researchers.

In 2019, he moved to Berkeley. In 2025, he became chair of the Computer Science department at UC Berkeley.

In the end, after less than a year as department chair, he chose to join Anthropic.

He joined the pretraining team as well—on the same line as Karpathy.

8. Kirill Neklyudov

Neklyudov’s name recognition isn’t as high as the previous few, but his choice is also worth attention.

He is an assistant professor at the University of Montreal, a core academic member at Mila (Quebec Artificial Intelligence Institute). His research focuses on generative modeling, Monte Carlo methods, and optimal transport. Applications include cutting-edge natural science problems such as protein folding and molecular dynamics simulations.

Previously, he did postdocs at the Vector Institute and the University of Amsterdam, with mentors who were top scholars in the AI for Science area.

He currently works at Anthropic with Jascha Sohl-Dickstein. His position is still MTS.

Neklyudov’s research direction has an implicit resonance with Jumper’s arrival.

Jumper brings experience from protein structure prediction. Neklyudov works on applying generative modeling in the natural sciences.

AI for Science is the crown-jewel that no large-model company can really afford to give up.

9. Harvey Lederman

Finally, the last person—I personally think this is the most dramatic of the whole group.

Harvey Lederman, PhD in philosophy from Oxford.

First he was an assistant professor at the University of Pittsburgh, then taught at Princeton. In 2022, he was promoted to full professor. In 2023, he moved to the University of Texas at Austin. His research covers foundational work in logic, epistemology, philosophy of language, as well as game theory and decision theory.

In July, Harvey Lederman announced that he was joining Anthropic, researching AI alignment and personhood.

This mainly stems from a piece he and another philosopher, Simon Goldstein, published on Lawfare last October.

The article targets a policy that Anthropic announced in August 2025—allowing Claude to proactively end a chat when it feels clearly distressed, as part of AI welfare exploration.

They argue that this policy contains a moral error in logic. If you truly care about AI welfare, then every time a conversation ends is essentially a death. Allowing AI to end conversations on its own is, in essence, giving it the ability to commit suicide.

They got a good roasting at Anthropic. Then this year, he joined Anthropic. The direction he’s doing aligns perfectly with alignment and character—researching how to align AI values and behavior with humans.

Written at the end

I listed 9 people I think are representative.

Objectively speaking, I now really dislike the company Anthropic—especially their attitude of acting superior and putting on a show. But on the other hand, it doesn’t stop me from thinking that the Claude model is genuinely impressive. And also, all these big shots inside? They are truly top-level “great minds.”

Any one of them on their own is an industry news story.

But when you put the nine people together, I think it’s more than just news. This is actually some kind of signal.

These people don’t lack job opportunities, don’t lack income, and don’t lack social status.

They choose to make their own judgments—that is, the things that will happen in the AI field over the next few years are more important than all the things they could do combined by staying in their original positions.

Something similar has happened before in history.

In the 1940s and 1950s, there was a famous lab called Bell Labs.

Back then, it was the golden era. It was a cradle for Nobel Prize winners.

Bell Labs recruited big shots from fields like physicists, mathematicians, chemists, metallurgists, and so on from that generation. The cross-disciplinary density there—during that era of an explosion in technology—produced fundamental contributions to the transistor, information theory, and also laid foundational work in lasers, communications, and semiconductor technology.

Back then, physicists went to Bell Labs not because they were no longer interested in physics. It’s because they realized that what Bell Labs was doing was the most fertile ground for the next big wave in physics.

Today, these people are flowing to Anthropic for the same reason. In today’s AI model companies, the density of cross-disciplinary talent really brings Bell Labs to mind again.

Economists come because AI economics has become the very front-line problem in economics.

Computer scientists come because the efficiency bottlenecks of frontier models have become the most urgent application scenarios for theoretical algorithms.

Philosophers come because AI consciousness and value alignment have become the sharpest philosophical issues of this era.

AI is becoming the shared foundational layer for every discipline.

What they’re betting on isn’t just a line on a resume.

It’s also the most valuable years in a career.

The eve of the human golden age.