99% of AI Startups are Wasting AI | Radical AI, Joseph F. Krause

Key Concepts

• Artificial General Intelligence (AGI) for Scientific Discovery: The core mission of Radical AI, aiming to create AI capable of human-level intelligence for scientific breakthroughs.
• Material Science Discovery: The specific domain Radical AI is targeting, characterized by long development cycles and significant barriers to innovation.
• Novel Discovery vs. Optimization: The distinction between creating entirely new solutions versus incrementally improving existing ones, with Radical AI focusing on the former.
• Bias to Action: A crucial entrepreneurial principle emphasizing immediate execution and learning from results over extensive planning.
• Inverse Design: A paradigm shift in material science where solutions are designed from the problem backward, enabled by AI and autonomy.
• 51% Rule: A decision-making framework encouraging timely action when confidence reaches 51%, prioritizing learning from outcomes.
• Mission-Driven Culture: The emphasis at Radical AI on hiring individuals motivated by a profound purpose rather than just employment.

Radical AI: Revolutionizing Scientific Discovery with AGI

The Problem with Current Scientific Discovery

The transcript highlights a significant inefficiency in the current scientific discovery process, particularly in material science. This process is described as being "150 years old" and taking an "incredibly long time, typically 10 plus years" from novel discovery to a scalable material system, often involving "exorbitant amount of cost." The speaker, Joseph Krauss, co-founder and CEO of Radical AI, contrasts this with the focus of many AI companies on "low-hanging fruit" and "small problems," questioning why AI isn't being used for grand challenges like curing cancer.

Radical AI's Mission and Approach

Radical AI is dedicated to building "artificial general intelligence for scientific discovery," starting with material science. Their core objective is to "reinvent the way we do the scientific process." They aim to move beyond incremental "one, two, or five percent improvements" and tackle "novel discovery," which is identified as the most challenging and impactful area. The company believes that success in these areas will "fundamentally reshape human trajectory."

The Power of AI in Material Science

The transcript emphasizes AI's ability to overcome human limitations in scientific research. A key capability is AI's power to "index information." The example of AlphaGo is used to illustrate this: AlphaGo made a move in Go that human experts had never conceived, demonstrating AI's capacity to process vast amounts of data and discover novel strategies.

Applied to material science, AI agents can "index millions of scientific publications" and learn from existing research at an unprecedented speed, estimated to be "about 370 times the speed" of a human scientist. This allows for the simulation of "billions of materials" and testing of "thousands of them," connecting information in real-time.

This capability enables a shift towards "inverse design," where materials are designed backward from the desired solution to a problem, rather than discovering materials and then searching for applications. This is a process that human scientists struggle with or can only achieve over "very very long time frames."

Overcoming Barriers to Innovation

Radical AI's vision is to "remove materials as our biggest barrier to some of our most important industries" such as automotive, aerospace, manufacturing, defense, climate energy, and semiconductors. By accelerating material discovery, they aim to unlock innovation across these sectors.

The Formation of Radical AI

The genesis of Radical AI involved a confluence of ideas and experiences:

• Joseph Krauss's Background: Krauss's journey was shaped by a desire to pursue something he was "the best in the world at" and "truly love[d]." His military service in the US National Guard instilled the importance of putting the "mission of the team... before yourself." His graduate studies at Rice University, focused on impactful research, and his experience at the Army Research Lab, where he witnessed the disconnect between fundamental science and practical application, solidified his desire to "commercialize science."
• Encounter with Kevin Ryan: Krauss's pursuit of entrepreneurship led him to Kevin Ryan, a prolific New York entrepreneur. Ryan challenged Krauss to "prove out" his belief that material science was the future, leading to a six-month internship where Krauss learned the value of a "bias to action."
• Jorge's Insight: One of Krauss's co-founders, Jorge, a venture capitalist at Ali Corp, was researching AI at a fundamental level. He questioned why AI was being used for "low-hanging fruit" and not for grand challenges like curing cancer.
• Intersection of AI, Material Science, and Robotics: After extensive research, the team identified material science's fragmentation and slow progress as a prime area for AI implementation. Their deep dive into publications at the intersection of material science, AI, and robotics led them to their third co-founder.
• Herd Cedar's Expertise: Herd Cedar, the third co-founder, had built an "autonomous scientific lab" at Lawrence Berkeley National Lab, bringing crucial expertise in robotics and automation.

Together, these three co-founders envisioned a "new paradigm of scientific discovery" driven by AI and autonomy, transitioning from a human-driven to an AI and autonomydriven process. They felt compelled to build this company themselves, recognizing that "this is the way science was going to go."

Funding and Early Stages

Radical AI was incubated within Ali Corp, with Kevin Ryan committing to invest. Recognizing the capital-intensive nature of building a "full stack solution" in material science, they aimed for a "fairly large pre-seed round." They presented a "100-page pitch deck" to Kevin Ryan, who, impressed by their vision, became the sole investor for their pre-seed round, which was raised in "about 45 minutes."

Culture and Decision-Making at Radical AI

Radical AI fosters a "mission-driven culture." They explicitly state, "If you are looking for a job, this is not the place for you. If you are looking for a mission, then you should come work here." This ensures that every team member is deeply committed to the company's goals.

A key cultural tenet is embracing failure as part of the discovery and learning process. They operate under the "51% rule," inspired by SpaceX, which encourages making decisions when confidence reaches 51%. This is balanced by considering the "size of the decision" and the "risks if the decision is wrong." The goal is not just speed but "quick decision-making but effective decision-making." They acknowledge that even with extensive deliberation, 100% certainty is rarely achievable, and learning from outcomes is paramount.

The Long-Term Vision

Radical AI's ambition extends far beyond immediate goals. They believe their company will "way outlive myself and and the two co-founders," potentially existing for "100, 200 years old." Their ultimate aim is to create a world "not today thought possible" by removing materials as a bottleneck to innovation. Joseph Krauss reflects on Steve Jobs' quote about connecting the dots, emphasizing that while the path forward may be uncertain, past experiences, even frustrating ones, are "imperative" for future breakthroughs. The core message is to "never ever give up no matter the scenario" in the pursuit of answers.