Published on January 30, 2026
Watch the video of this panel on Gamma Prime’s YouTube channel:
Panel talk with Michael Terpin (Founder and CEO at Transform Group), Christian Sauer (CEO at Soonami), Catrina Wang (GP at Portal Ventures), Amit Mehra (Partner at Borderless Capital) and Anna Shakola (Head of Business Development at Cointelegraph Accelerator)
Anna Shakola (Head of Business Development at Cointelegraph Accelerator):
Michael, going back to your vision for 2030—five or six years from now—what is overhyped right now and what is underhyped? What should we pay attention to, and what should we slow down on?
Christian Sauer (CEO at Soonami):
I think overhyped is a little bit…
Anna Shakola:
AI and memecoins?
Christian Sauer:
AI and making money, and underhyped is, to me, a little bit the purpose. There was this nice book, Start With Why. I think for a lot of people, the why is at the moment just, “I want to make money.”
And I think that doesn’t really help. So I would love to see more purpose in a lot of these projects, that we really understand why they are there.
Where does society benefit? And that’s at least how I entered the market five years ago, when we all wanted to build a decentralized layer for humanity. And now we’re ending up pumping meme coins.
So I think that would be my most underrated part of the industry.
Michael Terpin (Founder and CEO at Transform Group):
I’d say overrated would be layer ones and layer twos that are specific to AI. I don’t think you need a new layer one. I don’t think you need a new layer two.
Its only purpose is to basically do what the other hundred layer twos are, but use take-care coins instead of polygons. I think underhyped is probably that 90% of the agents are going to die, but the 10% that survive are going to be giant and used by millions of people profitably in the next five years. Because again, do you want to, when five years from now you’re doing hundreds of transactions a day, whether it’s trading like you would AI Quant, or I’ve got another portfolio company that’s Jacques Voorhees, which is Eric Voorhees’ dad.
Eric and I are the seed investors in it, called IceCap AI. And it’s a shopping bot that basically goes into a store and answers all the questions like a salesperson as an agent, and it increases the lift of sales, because a physical store has 40% sales. If you walk in the front door, you walk out with something 40% of the time.
An online store has 4%. And because it’s a horrible experience if you have questions, how do you get answers? You look it up in a database file, a help file.
Maybe you’ve got a really bad bot, but if you actually have a human-sounding, thinking agent—not a chatbot—that basically helps you through the sale and is trained like a salesperson: “Oh, you’re looking at that. I have another color of that dress. You want to see how it would look on you?
We have a sale. Would you like to know about the sale going on right now?” All it has to do is raise the sales rate by 2%, and that’s a 50% increase from 4% to 6% conversion.
And so that’s another use case that I’m bullish on, which is sort of AI sales agents.
Catrina Wang (GP at Portal Ventures):
What’s overhyped, I think, in the context of crypto, is autonomous agentic trading. I’ve seen many, many trading decks doing that. But realistically, as humans, we still want a level of control.
Imagine you have an AI agent that’s doing all your crypto trading for you. I just have no say. Who’s going to do that?
We want a co-pilot. We don’t necessarily really want autonomy for AI agents. Even on a technical standpoint, AI is largely non-deterministic, whereas smart contracts are.
So that mismatch can cause a lot of issues as well. What’s underhyped? Actually, I think there’s real opportunity when it comes to going back to the three pillars.
There’s data, there’s algorithm. Algorithm is actually what BitTensor is doing, and it does make sense.
But when it comes to data contribution:
Before, why would I want to contribute my data for additional training? OpenAI is doing that. But now, with token incentives and on-chain verifiability and ownership tracking, I can actually get a share if they are able to develop the LLM based on my data contribution and somehow profit off it. Then there is very distinct proof that I own a share of this.
When it comes to data contribution, I haven’t really seen a lot of credible startups doing this, but it would be quite interesting.
Amit Mehra (Partner at Borderless Capital):
I agree with most. I think generally, within the firm, we are not very excited about the infrastructure side. I think one of the things we are seeing is a lot of startups, for example, will go and basically use the VC money raised to buy a number of GPUs and data centers.
And actually, it’s very capital-intensive, but it’s not scalable. I think we are a little bit more bullish, and what’s underhyped, I think, is the challenges that centralized big LLM companies are facing.
So there are things in AI that just can’t be done. It’s just not scalable to be done in a centralized way, like refining models or, for example, data attribution or governance. So companies that are hitting that, I think, are kind of underhyped, and they will really grow big.
Christian Sauer:
One last thought. On AI agentic trading, imagine if everyone used it—then all these AI agents would just be fighting each other. That’s an interesting dynamic. And more broadly, look at AI in marketing. It’s now incredibly easy to generate marketing content and LinkedIn posts with AI. But in the end, there are only so many eyeballs.
If everyone has access to the same AI-powered marketing tools, the impact gets diluted. This will happen across many markets: AI enables something powerful, but once everyone uses it, the advantage disappears.
