Unlocking Hidden Patterns: A New Approach to Enterprise AI
Guest: Mark Anderson, Co-Founder and CEO of Pattern Computer
Host: Seth Earley, CEO at Earley Information Science
Published on: August 15, 2025
In this episode of Earley AI Podcasts, host Seth Earley welcomes Mark Anderson, co-founder and CEO of Pattern Computer, for a fascinating exploration of what lies beyond the current AI mainstream. With a career grounded in technology, strategy, and scientific innovation—including receiving the Alexandra J. Nobel Award for his contributions to computing and medicine—Mark brings nearly a decade of experience in developing proprietary pattern recognition technologies that move far beyond traditional machine learning models.
Together, Seth and Mark dive deep into the journey of Pattern Computer, unveiling its revolutionary Pattern Discovery Engine—a platform with the unique ability to make discoveries in data that have eluded conventional approaches. Mark explains how his passion for science and the shortcomings of the classic scientific method sparked the creation of new mathematical and architectural foundations in AI, leading to major breakthroughs not only in medicine but also across enterprise applications.
Key Takeaways:
-
Origins of Pattern Computer: The story behind the formation of Pattern Computer and its foundational mission to turn pattern discovery from an art into a science.
-
A New Approach to AI: How the Pattern Discovery Engine goes beyond finding incremental improvements, enabling true discovery by flipping the traditional scientific method.
-
Breakthroughs in Medicine: The real-world impact of Pattern Computer’s approach, including the discovery of gene patterns and the development of new drugs for triple negative breast cancer.
-
Pattern Discovery vs. Large Language Models: The critical differences between pattern discovery engines and LLMs, and how these technologies can work together to combine human-friendly communication with genuine scientific discovery.
-
Explainable AI and Ethics: Why true explainability, interpretability, and ethical data are at the heart of next-generation AI—and how Pattern Computer is leading the way with interpretable outputs and transparency.
-
Enterprise & Science Applications: Use cases in aerospace, mining, healthcare, and more, where Pattern Computer’s approach has led to major discoveries in seconds—successes that eluded brute-force methods for years.
-
Advice for Organizations: How businesses and innovators can access and test the Pattern Discovery Engine for their own complex data challenges.
Insightful Quotes:
“Instead of having a hypothesis and then you run, you want to go again, it’s the opposite. You’re not allowed to have any hypothesis. You can’t bring your bias to the game. And instead of that, you have good data. You run the data and you generate the hypothesis. That’s the right way to solve problems.” - Mark Anderson
"We don't use LLMs directly, it's not part of our technology, it's not part of things that we sell to our customers or clients... Our ancestral tree is science." - Mark Anderson
Links
LinkedIn: https://www.linkedin.com/in/markandersonpredicts/
Website: https://www.patterncomputer.com
Ways to Tune In:
Earley AI Podcast: https://www.earley.com/earley-ai-podcast-home
Apple Podcast: https://podcasts.apple.com/podcast/id1586654770
Spotify: https://open.spotify.com/show/5nkcZvVYjHHj6wtBABqLbE?si=73cd5d5fc89f4781
iHeart Radio: https://www.iheart.com/podcast/269-earley-ai-podcast-87108370/
Stitcher: https://www.stitcher.com/show/earley-ai-podcast
Amazon Music: https://music.amazon.com/podcasts/18524b67-09cf-433f-82db-07b6213ad3ba/earley-ai-podcast
Buzzsprout: https://earleyai.buzzsprout.com/
Podcast Transcript: Pattern Discovery Beyond LLMs - A Scientific Approach to AI
Transcript introduction
This transcript captures a conversation between Seth Earley and Mark Anderson about Pattern Computer's revolutionary Pattern Discovery Engine, exploring how it differs from large language models, its breakthroughs in cancer drug discovery, real-world applications across industries, and the importance of explainable AI for genuine scientific discovery.
Transcript
Seth Earley: Okay. Welcome to the Earley AI Podcast. My name is Seth Earley. I'm your host. Today we're going to be talking about a different approach to AI. And while most of the world is focused on large language models, there are other groundbreaking technologies changing the way we understand and use data. And our guest today is at the forefront of this movement. Joining me today is Mark Anderson. He's co-founder and CEO of Pattern Computer. Mark has led the company for nearly a decade, building the pattern discovery engine into a platform that delivers insights traditional models cannot reach. He's known for his work as a technologist, a strategist, and author. And he spent his career helping organizations see the future of technology and put it to work. He's recipient of the Alexandra J. Nobel Award for scientific discoveries for fundamental contributions in computing and medicine. Mark, welcome to the show.
Mark Anderson: Thank you, Seth. Good to be here.
Seth Earley: So you were on a while ago, and it's great to have you back. I want to kind of refresh people's memories and just provide the story. Your approach, your 10th anniversary. Tell us about the journey. How has Pattern evolved from the early startup days to where you are now? How and what does the pattern discovery engine enable that other technologies cannot?
Mark Anderson: Right. So we have a pretty cool foundational story. I'll do it in a few sentences. As you know, I'm a science guy from Stanford, and so I had the scientific method tattooed in my brain when I was five years old. And then somewhere along that run, it occurred to me that the method wasn't perfect. The scientific method is you have a hypothesis, and then you test it. And what I realized was that the hardest part is coming up with the hypothesis in the first place. And if you're biased, or if you're looking in the wrong direction, you'll never find what you're looking for.
So the insight was to flip it. Instead of having a hypothesis and then you run it, it's the opposite. You're not allowed to have any hypothesis. You can't bring your bias to the game. And instead of that, you have good data. You run the data and you generate the hypothesis. That's the right way to solve problems. And so that was the founding insight. And from that, we said, okay, we need to build new mathematics, new architecture, new everything to make that work. And so we spent about three years in stealth mode building the Pattern Discovery Engine.
Seth Earley: And what does it actually do? Like, give me a concrete example of how it works differently than traditional approaches.
Mark Anderson: Sure. So we don't use LLMs directly, it's not part of our technology, it's not part of things that we sell to our customers or clients. Our ancestral tree is science. The idea originally had nothing to do with the ancestral tree of LLMs. We're looking for patterns that lead to discoveries. So let me give you a concrete example. We started with cancer drugs because that's hard. We picked the top five cancers. And the first one was triple negative breast cancer. Horrible disease. It has a very high mutation rate. It's hard to keep up with it if you're in a patient or in the lab. And it has no direct treatment today.
So in our very first outing, we used a data set out of Berkeley National Laboratory called MetaBrick, and it's a fairly large data set of women who had this disease and died. The paper is published, they're looking at one, two, three genes. And that was normal back then. It had been around for 20 years. Everybody had been in there. We found between 50 and 100 new patterns of genes that were in patterns that had not been seen before. And so, instead of going through what's called GWAS and trying to find things, we just found them. Literally, from that data set, we could see them. That was a discovery. That was not an incremental improvement. That was a discovery.
Seth Earley: So from those discoveries, what happened next?
Mark Anderson: From those discoveries, we then created drug combinations. And those drug combinations were tested in vitro, meaning in the lab, against triple negative breast cancer cells. And they worked. And they worked really well. In fact, they worked so well that we were able to get patents on them. And we now have multiple patents on drug combinations for triple negative breast cancer. So the typical success rate in drug discovery from inception to clinical trials is about 0.01%. So, one in 10,000. For us, it's 14%. So, 14,000 times better.
And I think that's gonna continue right on through. When we were doing this at full speed, which we're not doing right now because we're doing some more things, we were creating five novel, new, patentable drugs against one of the top five cancers every month. Of those five, we were getting one per month that would test to be effective. That's a crazy number against the top five cancers. It's unheard of.
Seth Earley: That's remarkable. So you've obviously proven this works in the medical field. What about other applications?
Mark Anderson: We're now launching Pattern DE online. Pattern DE is the pattern discovery engine writ large, integrated everything that we've built so far into one service, SaaS, we call it Discovery as a Service. And it's built so you can, just like OpenAI or something else, you can come to a website, bring in data, run it. And instead of doing what chat would do, you're looking for major scientific or business discoveries that were unavailable before.
Our very first influencer came out of the academic world. They had been working with Rob Knight's lab at UCSD in microbiome, and they'd been working with the people at Oak Ridge National Lab using the Summit supercomputer, and for 18 months, they couldn't crack this puzzle. Big puzzle. The puzzle was, kids around the world in farm country had asthma incidences that were high. They didn't know why. And they were wondering if it was caused by the bugs in the soil. They solved the problem using Pattern DE in 56 seconds.
Seth Earley: 56 seconds compared to 18 months with a supercomputer?
Mark Anderson: That's right. They were able to show the correlation of microbiome soil content in rural environments worldwide to the incidence of asthma in children. And they located it. So now we're working with aerospace companies, we're working with mining companies, we're working with financial services companies. In every case, what they're looking for is patterns in their data that they haven't been able to find before.
In aerospace, for example, we're looking at patterns in satellite data to identify potential problems before they become critical failures. In mining, we're looking at patterns in geological data to identify where to drill for maximum yield. In financial services, we're looking at patterns in trading data to identify opportunities and risks. And in every case, what we're finding is that the Pattern Discovery Engine is able to find patterns that traditional methods miss.
Seth Earley: So how do organizations get access to this? Can anyone use it?
Mark Anderson: Pattern DE is now available online at patterncomputer.com. You can go there, you can sign up for an account, you can upload your data, and you can run it. We have a freemium model, so you can start for free and see if it works for your use case. And then if it does, you can scale up. We also work directly with larger enterprises on custom implementations where we integrate Pattern DE into their existing systems and workflows.
Seth Earley: Talk to me about explainability. That's obviously a hot topic in AI right now. How does Pattern Computer approach that?
Mark Anderson: Explainability and interpretability are at the heart of what we do. When the Pattern Discovery Engine finds a pattern, it doesn't just tell you "here's a result." It shows you exactly what the pattern is, what data points contribute to it, and why it matters. This is fundamentally different from a black box neural network where you can't really explain why it made a certain decision. Our outputs are transparent. You can see the pattern, you can verify it, you can test it, and you can understand it. That's essential for scientific discovery and for business applications where you need to know why something works, not just that it works.
Seth Earley: And how does this relate to or differ from what people are doing with large language models?
Mark Anderson: LLMs are amazing at generating human-like text and finding linguistic patterns. They're great for communication, for summarization, for certain types of content generation. But they're not designed for scientific discovery. They're trained on human-generated content, which means they're inheriting all of our biases, all of our assumptions, all of our limitations. They can tell you what humans have written about a topic, but they can't tell you what's true that humans haven't discovered yet.
That's what we do. We're not generating text based on patterns in language. We're finding patterns in data that lead to new discoveries. Patterns that no one has seen before. Patterns that can lead to new drugs, new materials, new solutions to problems that have been unsolved for years or decades. So we're complementary to LLMs, but we're doing something fundamentally different.
Seth Earley: What advice would you give to organizations that are trying to figure out their AI strategy right now?
Mark Anderson: I think the key is to understand what problem you're trying to solve. If you need better communication, better customer service, better content generation, LLMs are great for that. But if you need genuine discovery, if you have complex data and you're trying to find patterns that lead to breakthroughs, that's where pattern discovery comes in. Don't just adopt AI because everyone else is doing it. Understand your specific problem, and then find the right tool for that problem.
And I would say, test things. We have a freemium model specifically so people can bring their hardest problems and see if pattern discovery works for them. We're confident it will, but we want people to prove it to themselves. Bring your most difficult data problem, the one that you've been working on for months or years without success, and see if we can solve it in hours or minutes.
Seth Earley: That's a great approach. Any final thoughts as we wrap up?
Mark Anderson: I think we're at an inflection point in AI. Everyone's focused on LLMs right now, and that's fine. They're useful tools. But there's a whole other dimension of AI that's about genuine discovery. About finding things that have never been found before. And that's what Pattern Computer is about. We're not trying to make existing processes 10% better. We're trying to find patterns that lead to discoveries that change everything. That save lives with new cancer drugs, that create competitive advantages for businesses, that solve problems that have been unsolved for decades.
And the exciting thing is that this technology is now available to anyone. You don't need to be a Fortune 500 company. You don't need a team of PhDs. You can go to our website, upload your data, and start making discoveries. That's powerful. And I think it's going to change a lot of industries over the next few years.
Seth Earley: Well, Mark, thank you so much for your time today. This has been fascinating. And for our listeners, if you want to learn more about Pattern Computer and the Pattern Discovery Engine, you can visit patterncomputer.com or connect with Mark on LinkedIn. Thanks for tuning in to the Earley AI Podcast, and we'll see you next time.
Mark Anderson: Thanks, Seth. Great talking with you.