I wanted to spend this conversation with Lukas talking about something that’s becoming impossible for investors to ignore now… AI. Specifically, how AI is changing the research process. We’re building tools internally ourselves at GeoInvesting, so this isn’t theoretical for me anymore. I use AI every day now, especially for speeding up research, understanding business models faster, identifying patterns, and helping organize information. But one of the biggest points I tried to make throughout the discussion is that AI still doesn’t replace real investing skill. At least right now, the edge still comes from understanding businesses yourself, developing judgment through experience, and doing the manual work that teaches you how markets actually behave. AI can absolutely improve a strong process, but it can’t magically create one for you.
One thing I really agreed with Lukas on was the importance of still starting with the numbers. He explained that when he looks at a company, he still opens the filings first and decides pretty quickly whether it’s even worth deeper research. I think that’s important because a lot of newer investors are starting to rely too heavily on AI before they’ve actually learned how to analyze companies on their own. Where I think AI becomes incredibly useful is on the qualitative side: understanding industries, learning new concepts quickly, exploring business models, and helping you think through ideas from multiple angles. I also talked about how I’m personally using AI almost like an investing partner now: challenging assumptions, testing frameworks, spotting inconsistencies in management commentary, and even helping identify potential catalysts or opportunities I may have otherwise overlooked.
We also spent a lot of time talking about whether AI will eventually destroy the investing edge, especially in microcaps and nanocaps. Both Lukas and I agreed that while AI is becoming an incredibly powerful tool, investing still remains deeply personal and judgment-driven. Lukas made the point that conviction itself is something investors still have to develop on their own through experience, mistakes, and independent thinking, not something an AI model can simply hand you. I agreed with that, but also talked about how AI can strengthen conviction when used the right way by helping investors explore industries faster, challenge assumptions, and think through situations from multiple angles. Ultimately, we both came back to the same core idea: the investors who combine strong manual research skills with AI tools intelligently will probably have the biggest advantage going forward.
Skull Sessions is a collaboration with Geoinvesting.com, a full-stack microcap research platform and MS Microcaps LLC , home of the Microcap Quality Index (MSMqi).
Key Topics Discussed
How AI is changing investment research
Why AI should enhance, not replace, manual analysis
The importance of reviewing financial statements yourself
AI as a tool for qualitative research and business understanding
Whether AI will destroy the investing edge in microcaps
Building AI workflows around an existing investment process
Conviction building in investing
Using AI to challenge assumptions and management claims
Stocks Discussed
RCM Technologies (RCMT): Engineering and staffing company with exposure to industrial and data center-related demand.
ABVC BioPharma (ABVC): Small biotech name used as an example for analyzing sharp post-selloff situations with AI.
Bandwidth Inc. (BAND): CPaaS/software communications company discussed in the context of SaaS durability under AI pressure.
Cipher Mining (CIFR): Bitcoin mining company linked to broader AI/data center infrastructure themes.
