👋 Howdy to the 1,794 new legends who joined this week! You are now part of a 204,887 strong tribe outperforming the competition together.

LATEST POSTS 📚

If you’re new, not yet a subscriber, or just plain missed it, here are some of our recent editions.

💥 Strategism: The AI Revenue Per Employee War. How companies are getting leaner, and how you can too.
🏜️ Unfiltered: The Road To Figure’s IPO. A behind the scenes with multi-IPO founder, Mike Cagney.
📝 Strategism: Mastering The Founder Letter. Standing out made easy. How founder letters are shaping company narratives today.

PARTNERS 💫

Visa costs are up. But growth can’t wait.

Now, new H-1B petitions come with a $100K price tag. That’s pushing enterprises to rethink how they hire.

The H-1B Talent Crunch report explores how U.S. companies are turning to Latin America for elite tech talent—AI, data, and engineering pros ready to work in sync with your HQ.

Discover the future of hiring beyond borders.

_{Interested in sponsoring these emails? See our partnership options here.}

HOUSEKEEPING 📨

For anyone who has been following along for a while, I am still slowly eeking my way towards being the technology ecosystem’s strongest old man weightlifter. I am less than two months away from forty, so being able to move any serious weight is good, but my career goal is a 120kg snatch and a 150kg clean and jerk.

Yesterday, after loads of travel, sickness, and a general lack of training, I managed to build up to 100kgs from the hang (knees) and failed. But everything up to that point, 95kg being the heaviest, was perfect. I do feel if I get a run at this—little injuries, consistent training—for a solid 24 months, I can hit those big numbers I mentioned earlier. And take the crown of the tech ecosystem’s strongest old man. Wish me luck!

INTERVIEW 🎙️

Joseph Krause, Co-Founder & CEO, Radical AI

Joseph Krause is the co-founder and CEO of Radical AI, a company pioneering the use of artificial intelligence and robotics to accelerate materials discovery and scientific research. With a background in materials science and engineering, Krause has built Radical AI around the idea of ‘self-driving labs’ — fully automated research environments that can test, learn, and optimize new materials at unprecedented speed.

Before founding Radical AI, Krause worked across both the public and private sectors, including time in the U.S. military, where he developed a strong focus on systems thinking and operational efficiency. Today, he leads Radical AI’s mission to bridge science and automation, empowering industries like energy, defense, and manufacturing to innovate faster and more sustainably.

Tell us the problem you’re trying to solve?

At Radical AI, we’re building a new form of scientific discovery. When we chose where to make an impact, we looked at the industries that matter most—automotive, aerospace, manufacturing, defense, climate, energy, semiconductors, and electronics. The most important industries are all a direct result of materials science and materials R&D.

That’s the hard problem Radical AI is solving with our materials flywheel.

Explain how materials discovery works?

I was a materials scientist in grad school on a fellowship at the Army Research Lab, so I’ve done this. Say I want to make a new material for smartphones. First, I read a bunch of scientific publications on the relevant material system: what’s used today, how it’s made, target properties, and the specs the OEMs require—conductivity, bendability, toughness.

Then I form a hypothesis: “The new material will be X.” I run quantum chemistry and other theories to simulate whether X is viable. If it looks good, I make the material in the lab, fabricate a sample, and test its properties. I characterize what I made and measure performance: current flow, bendiness, toughness—whatever I was optimizing for.

I analyze the data, update my thinking, and form a new hypothesis. Maybe it isn’t as bendy as expected, so I add more copper to the glass and repeat. This process is very serial—one step after another—and constrained by human bandwidth. I can only read so many papers, run so many experiments, and analyze so many results at once.

AI and self-driving labs change that. We can do it all in parallel—read millions of publications, test thousands of images and samples, and consider millions of candidate materials simultaneously. We move from a serial approach to a parallel one. That’s our focus.

What’s the impact of that? Paint the future.

First, data. Ninety percent of what materials scientists do doesn’t work, and those negative results are rarely captured—even though they’re essential to understanding the 10% that does work. Our self-driving labs capture everything—temperature, pressure, oxidation, concentration, even room humidity—and feed it into our models to build the most proprietary experimental dataset in the world.

Run tens or hundreds of thousands of experiments per year, and you amass the largest dataset and can start predicting truly exotic materials—like a room-temperature superconductor. That would enable maglev trains, lossless power transmission, and meaningful progress toward practical nuclear fusion. It has eluded scientists for decades.

If we build the dataset, the best models, and the best self-driving lab, we believe we can experimentally find and confirm breakthroughs like that. The unlocks are so large we struggle to fully picture the future: civilizations on Mars, hypersonic travel from New York to LA in 11 minutes, and floating trains across the country. Novel materials power the world. That’s the future we want.

Walk us through your materials flywheel and why it matters?

The flywheel has two parts. On one end is the AI engine. Its job is to recommend new materials, generate new structures, and orchestrate the flywheel. We do that via an agentic, mixture-of-experts approach: atomistic modeling, inverse design/generative modeling from end applications, and LLM-driven agent frameworks that use these tools to propose candidates and send them to the lab.

On the other end is a fully robotic self-driving lab with no humans in the loop. Two things matter here. First, active learning: after every experiment, we ingest the results and update what to do next—like my manual process, but fully parallel. Second, comprehensive data capture: every single run is logged and fed back into the AI engine.

This creates a compounding flywheel that builds the most proprietary dataset across computation and experimentation. It’s a new paradigm: from human-driven science to AI- and autonomy-driven discovery.

Do you have competitors doing the same thing? How do you think about them?

No one is building the full-stack solution to discover and sell materials at scale. That’s our business model: become the largest materials company in the world.

Many companies in ‘materials + AI’ focus on one side of the flywheel—the AI engine that predicts materials. They’re not building self-driving labs to test at high throughput, scale, and sell those materials. If someone ships a Nobel-worthy AI model, great—we’ll use it. It will improve our predictions, our data, and our system.

I tell the team: competitor or not, our job is to run them over. Don’t worry about what others are doing. Build the best company and the best technology in the world, and the rest takes care of itself.

How do you think about storytelling as a founder?

The founder’s job—especially the CEO’s—is to be unbelievably concrete on the vision and flexible on the details. That’s a Jeff Bezos principle. Whether you’re talking to investors, talent, or customers, they want two things: where you’re going and why you’re the team to get there. What makes you different?

The data room matters, but it’s table stakes. People need to believe the world will look the way you say—and that you’ll make it so. Good storytelling isn’t making things up; it’s explaining why what you’re doing today leads to the outcome you’re promising.

Beyond storytelling and vision, what’s your leadership philosophy?

First, culture. We’re obsessive about it. If you’re looking for a job, don’t come here. If you’re looking for a mission, do. We look for three things in everyone.

Our culture is built around first-principles analysis — asking ‘why’ about everything. We follow the 51% rule, making decisions once we’re just over halfway confident, knowing it’s faster to correct mistakes than to wait for certainty. And above all, we’re driven by relentless ambition: we’re trying to achieve what many believe is impossible, something that could change civilization itself. Failure happens daily, sometimes for years, but the accumulated learning is what ultimately builds the breakthrough.

Second, how we operate. I even banned the word strategy in an email because I’m allergic to over-planning. Execute, learn, iterate, move forward. Fail fast and often. We model this at the top. In monthly all-hands, we share where we failed, what we did right, where we’re going, why we changed, and what we won’t change because it compromises the vision. It’s okay to try, fail, and learn.

Biggest challenge going from zero to one?

Building a deeply interdisciplinary team stepwise. We have five technical buckets: software, machine learning, materials science, autonomy, and mechanical engineering. We didn’t start with someone in each, but we needed all of them to build the flywheel.

We had to choose where to start, build momentum, and add the rest without losing focus. Sometimes we lacked a key person or weren’t ready for a capability—robotics, automating instruments, or the lab’s operating system. You need a lot of people to build that. We reached 20 people in about eight months—very aggressive—but in the first three to four months, even while hiring two to three people every two weeks, gaps remained. Only once the whole team was in place could we see how to build our internal “System One.” Bridging that gap was hard.

What must happen to go from one to 100 in five years?

First, never lose the culture. Second, team—people build companies and technology. You need incredible people to build incredible technology. Third, ship meaningful results to the industry.

We’re not an academic lab, a national lab, or a nonprofit. We’re a materials company. We need to put products—novel materials—into customers’ hands so they can see the difference versus current systems. In five years, we want real materials through customer testing and optimization, so we can plan scaling and manufacturing for real applications.

How do you get the best out of yourself personally and professionally?

Discipline. I served in the military, and discipline is one of life’s keys. I run the same schedule most days. Personally, everything starts with discipline, and it bleeds into how you operate professionally. I constantly ask: Am I the best version of myself, and where can I improve? There’s always something to improve. Tackle it and move on to the next. Do that forever and you’ll always be improving. The day you stop learning and improving is the day you stop growing.

No leader in the world knows everything. The one who thinks they know everything is probably wrong. I keep a learning mentality, and discipline helps me execute across what matters, so I’m disciplined across the board, personally and professionally.

And that’s it! You can follow and connect with Joseph over on LinkedIn and Twitter, and don’t forget to check out Radical AI’s website.

BRAIN FOOD 🧠 

TWEETS OF THE WEEK 🐣 

TOOLS WE RECOMMEND 🛠️

Every week, we highlight tools we like and those we actually use inside our business and give them an honest review. Today, we are highlighting Hyperbound—the AI sales coach to help you quickly evolve and reinforce your playbooks as you scale.

See the full set of tools we use inside of Athyna & Open Source CEO here.

HOW I CAN HELP 🥳

P.S. Want to work together?

1. Hiring global talent: If you’re hiring tech, business or ops talent and want to do it 80% less, check out my startup, Athyna. 🌏

2. See my tech stack: Find our suite of tools & resources for both this newsletter and Athyna here. 🧰 

3. Reach an audience of tech leaders: Advertise with us if you want to get in front of founders, investors and leaders in tech. 👀