Sramana Mitra: My thesis on agentic AI is exactly what we discussed earlier: solving specific problems with deep domain knowledge. The production systems that are getting real traction and generating significant revenue are those solving major enterprise problems.
A great case study is Autonomize. The health insurance authorization process still relies on fax machines. Autonomize digitized the process, added agents and automation, kept humans in the loop, and delivered massive efficiency gains. These are real production systems generating hundreds of thousands of dollars.
>>>Sramana Mitra: Yanni, you said you are investing in superintelligence. Tell me more about what you’ve invested in and your thesis on superintelligence.
>>>Sramana Mitra: Yeah. So, Yev, I want to make a point, and I want your thoughts on it.
We hear a lot from people saying, “This is just a wrapper on top of ChatGPT.” But there is a lot of domain knowledge in specific verticals that you can solve very well with a wrapper on top of ChatGPT, because the domain knowledge is where the real work is happening.
>>>Sramana Mitra: Very true. This is very true. I want to underscore that you should not raise money at the highest valuation you can, because you’re pricing yourself out of the exit market and out of further fundraising. If you don’t meet milestones that are that aggressive and far ahead of the curve, you will be stuck.
>>>
Yanev Suissa, Managing Partner and Founder at SineWave Ventures, discusses changes in Venture Capital and his firm’s investment thesis.
>>>Sramana Mitra: Earlier you said you wanted to discuss how venture capital is changing. We have a little time, so I want to explore that.
There’s a point of view circulating that even though venture capital is focused on vertical AI right now, AGI could make vertical AI irrelevant. And if AGI arrives during the timeframe when current gen-AI companies mature and prepare for exits, those exit opportunities might disappear. These are existential threats for venture capital. How do you think about that?
>>>Sramana Mitra: Another important point: the probability of building a true unicorn—not an artificial one inflated with liquidation preferences—is extremely low. Many overfunded companies become “zombiecorns” or fail due to overfunding.
>>>Sramana Mitra: Great. That clarifies things. Now, let’s talk about AI. What’s happening? What are you seeing? What are you investing in? What do you like to invest in?
>>>