Subscribe
Share
Share
Embed
The official CEO.com podcast featuring unfiltered conversations with the leaders shaping our world.
Clint Betts: George, thank you so much for coming on the show. What an honor to have you. You are the CEO of Prospero AI. Tell us what it is and how you became the CEO.
George Kailas: I’ve been at Prospero from the very beginning—employee number one, essentially. But the story starts before that.
I came from a value investing background. I got my first job at a value-investing fund when I was 17. I taught myself accounting just to qualify for that role. After that, I spent years working on the buy side, building my own algorithms, and eventually running an AI-as-a-service business focused primarily on financial services.
Over those years, I realized that retail investors were operating at a significant data disadvantage. Even smaller hedge funds were often at a disadvantage compared to large institutions.
That realization became the foundation for Prospero.
Initially, our goal was actually pretty modest. We wanted to help people stay out of the way of institutions and reduce costly investing mistakes.
But the signals we developed ended up performing incredibly well. Today, we have an app that allows people to use those signals directly and a newsletter that teaches investors how to interpret them. The newsletter picks have outperformed the market by an average of 63% over the last four years.
What started as an effort to simplify investing ultimately turned into something much more powerful.
Along the way, we discovered that our models could do things like predict news events before they happened.
Clint Betts: Which is remarkable.
What you’re really doing is democratizing investing for everyday people who don’t have access to quantitative analysts and massive institutional research teams.
Did you always expect to build the community you’ve built, or did that come later?
George Kailas: Honestly, both.
Building a community was always part of the vision.
One of the reasons we kept the platform largely free for so long was specifically to build trust and grow that community.
I started an AI company 15 years ago, and I’ve always tried to stay ahead of trends. One thing I became convinced of early is that trust would become one of the most valuable assets in an AI-driven world.
As AI becomes more prevalent, trust becomes everything.
Part of our long-term vision is creating research that retail investors genuinely trust. If enough people trust it, the information becomes more valuable, which creates a virtuous cycle.
We also have broader ambitions.
One example is public-interest data collection. Imagine being able to ask users simple questions and aggregate information in real time to identify emerging trends—whether that’s economic activity, public health issues, or other important signals.
Traditionally, institutions have had access to that kind of information before the public does. We think there’s an opportunity to make some of those insights more accessible.
Clint Betts: You mentioned predicting news events. What exactly do you mean by that?
George Kailas: The first time it really struck me was with Alibaba.
This was around 2022 or 2023.
We noticed two of our most important signals spiking far above normal levels. Based on that, we added Alibaba to our portfolio.
Shortly afterward, Ant Financial announced a new funding round and the stock jumped.
Later, the signals cooled off, so we exited.
Then the exact same pattern appeared again. Our users even noticed it and started asking whether something similar was happening.
Not long afterward, Alibaba announced a major corporate reorganization and the stock surged again.
That was the first moment where I thought:
“Maybe we’ve built something bigger than simply helping people avoid mistakes.”
Even one of my advisors, who’s a well-known derivatives expert and usually very skeptical, became convinced after seeing that.
And we’ve seen many examples like that since.
Clint Betts: How many people are using the platform today?
George Kailas: We currently have around 15,000 monthly active users on the app and about 10,000 weekly active readers of the newsletter.
Clint Betts: A lot of people are fascinated by prediction markets like Polymarket. Do you see Prospero expanding into that world?
George Kailas: People ask me that all the time.
Personally, I’m skeptical of prediction markets as investment vehicles.
The odds are generally much worse than traditional investing.
If you buy stocks, historically the market trends upward over time. The probabilities are generally working in your favor.
With prediction markets, there are often numerous possible outcomes, which means the odds are frequently structured to benefit the market maker rather than the participant.
There’s also the issue of information quality.
At Prospero, one of our most important metrics is net option sentiment. We intentionally isolate institutional behavior because institutions tend to have better information and stronger incentives.
In prediction markets, you often can’t distinguish between informed participants and uninformed participants.
That makes extracting useful signals much more difficult.
Clint Betts: How worried are you that AI is driving the stock market to irrational levels?
George Kailas: I don’t think AI is artificially propping up the market.
I think there’s enormous real value being created.
Now, do I think there are bubbles forming? Absolutely.
But bubbles are tricky.
People often assume that because something is overvalued, it’s about to collapse.
History shows that’s not necessarily true.
A bubble can continue expanding for years.
If you become overly focused on timing the collapse, you can miss substantial gains.
The more important question is whether you have a plan.
How will you recognize when a bubble is actually bursting?
What are your exit criteria?
What are your risk controls?
Those questions matter far more than trying to predict the exact timing.
Clint Betts: You’ve talked about overbuilding compute. What do you mean by that?
George Kailas: There are a few aspects to it.
The first is the social side.
Many communities are pushing back against data centers because they consume enormous amounts of power and water while creating relatively few long-term jobs.
Traditional industrial projects often bring lasting employment benefits. Data centers generally don’t.
The second issue is technological.
Right now, the industry is operating under the assumption that AI requires ever-larger models and ever-larger amounts of compute.
But we’re already seeing a shift toward smaller, more specialized models that can solve specific problems much more efficiently.
I think there’s a real possibility that compute demand doesn’t continue growing at the pace many people currently expect.
Then there’s quantum computing.
Google recently stated that it expects to be ready for quantum-safe encryption by 2029.
If quantum computing progresses the way many people expect, it could fundamentally alter the economics of today’s compute infrastructure.
So I think there’s a possibility we’re building infrastructure based on assumptions that may not fully hold up over time.
Clint Betts: Then why are companies investing hundreds of billions of dollars into data centers?
George Kailas: Honestly, because that’s how capitalism works.
If someone is willing to write the checks, someone else will build the infrastructure.
You see these relationships where OpenAI commits to massive compute purchases, companies like CoreWeave commit to infrastructure, Nvidia sells the hardware, and everyone benefits from rising valuations.
As long as investors continue rewarding those relationships, the cycle continues.
Whether all of that compute ultimately gets utilized is a different question.
But if companies don’t build it, someone else will.
Clint Betts: What happens when companies like OpenAI or Anthropic eventually go public?
George Kailas: They’re very different situations.
OpenAI concerns me more.
Recently, they’ve started discussing advertising and making media-related acquisitions. To me, that suggests they may believe product differentiation alone won’t be enough, so they’re investing more heavily in distribution and messaging.
Anthropic is different.
I’ve never seen demand for a private company like I’ve seen with Anthropic.
People are desperate to invest.
They’ve grown incredibly quickly despite having far less capital than some competitors.
And when you look at the industries they could potentially transform—legal services, consulting, finance, and many others—the revenue projections start looking surprisingly realistic.
I’m extremely impressed with what they’re building.
Clint Betts: How do you maintain a moat when AI capabilities are advancing so quickly?
George Kailas: Our moat comes from structure.
For example, one of our signals measures profitability. Rather than simply throwing raw data into a model, we aggregate multiple factors—EBITDA, EPS, forward projections, simulations—and convert them into a structured score.
That structured approach makes it easier for both humans and machines to interpret.
We’ve spent years refining those systems, testing them in live markets, gathering feedback, and continuously improving them.
That process creates advantages that are difficult to replicate with a prompt alone.
Clint Betts: What does a typical day look like for you?
George Kailas: I spend a lot of time looking at markets, analyzing signals, talking to investors, and live-streaming.
But beyond that, I spend a lot of time thinking.
I’m a big believer that creativity requires space.
I studied the philosophy of creativity when I was building my first company, and one of the biggest lessons I learned is that forcing great ideas rarely works.
So I go on walks.
I consume information.
I give myself time to think.
I also make myself available to the team whenever they need help. If someone is blocked, I’ll take a call at 10 p.m. if necessary.
I think a CEO’s most important jobs are supporting people, generating differentiated ideas, and creating an environment where great work can happen.
Clint Betts: How large is the team today?
George Kailas: We only have one other full-time employee.
But over 50 people have contributed to the company through equity arrangements, advisory roles, and part-time involvement.
We’ve had to be extremely scrappy.
A lot of investors wanted us to raise prices much earlier, but that would have conflicted with our long-term vision of building a large, trusted user community.
Clint Betts: How have you funded the company?
George Kailas: We’ve raised a little over $1.5 million.
That includes roughly $538,000 through crowdfunding and the rest from angel investors.
We’ve been fortunate to have one particularly supportive angel investor who has played a significant role in helping us grow.
Clint Betts: Where do you see Prospero five years from now?
George Kailas: Ultimately, we want to become a major wealth management platform.
The signals are just the beginning.
Many wealth managers spend enormous amounts of money acquiring customers.
We’ve taken the opposite approach.
We’ve focused on building trust first.
Our users have seen the signals work. They understand the methodology. They’ve already developed confidence in the platform.
That creates a very different foundation for future products.
When we eventually launch wealth-management offerings, users won’t be making a leap of faith. They’ll already have years of experience with the underlying system.
That’s a powerful advantage.
Clint Betts: Finally, we end every interview the same way.
At CEO.com, we believe the chances one gives are just as important as the chances one takes.
When you hear that, who gave you a chance that helped get you where you are today?
George Kailas: The first person who comes to mind is Bruce Greenwald at Columbia Business School.
He runs the Value Investing Program there.
When I was 16, he gave me an opportunity and helped open doors that completely changed my life.
He introduced me to people who eventually gave me my first hedge fund opportunities. He advocated for me when most people would have overlooked a teenager with very little experience.
There’s no question I wouldn’t be where I am today without that support.
Clint Betts: George, thanks so much for coming on. Seriously, what you're building is remarkable.
George Kailas: Thank you. It’s been great, Clint.