GenAI Pioneer Reveals How to Spot Winning AI Investments

Generative AI is revolutionizing industries and capturing the attention of investors worldwide. With its ability to create text, images, video, and audio, this technology is becoming a cornerstone in media, entertainment, and gaming sectors. Recognizing its transformative potential, investors are increasingly focusing on generative AI startups.

We had the privilege of sitting down with Kevin Baxpiler, the co-founder and managing partner of ReImagine Ventures, to gain insights into the investor perspective on generative AI. With a background in early-stage investments at the intersection of AI and consumer entertainment technologies, Kevin has been at the forefront of recognizing and nurturing the potential of generative AI in the startup ecosystem. In this interview, Kevin shares his expertise on what investors look for in generative AI startups, the unique challenges these startups face, and the future outlook for this rapidly evolving technology.

Can you share with us some of the key characteristics you look for in generative AI startups when considering investment opportunities? How do you assess the potential for success in this rapidly evolving field?

So a few things are not that different to any startup really. We look at the team, what type of industry knowledge and market knowledge does the team possess and what kind of technical expertise do they bring to the table.

The difference in this case in generative AI is that often the technical expertise is much deeper and you can’t just have some really cool background from an army unit, you need to have really deep understanding about transformer models and machine learning. So, basically a deeper and more extensive knowledge of AI.

And then something that’s definitely different for GenAI is that we’re really interested in how you plan to train your model. So if you’re a GenAi startup that wants to build a vertical model or any kind of AI model, the key question is really where is the data from? Is the data potentially difficult to attain, obtain? Could it be a barrier to entry? How clean is it? Is it legal? So those are some of the questions that you can expect.

At the end of the day, another big question we spend a lot of time on these days, which is really difficult, is how fast will any of this technology be commoditized? We also look at what’s the platform risk? And by platform risk, I mean, is open AI or Google or any of those guys adding this as a feature tomorrow to their platform.

So as an investor, how do you evaluate the market readiness and scalability of GenAI solutions?

This is not an easy question to answer right now because we’re really at the infancy of this technology.

There are still many uncertainties, starting with one of the biggest ones around regulation, which right now is starting to shape up, but there’s still a lot of uncertainty. You have other uncertainties, which could be around GP or chip shortages, data centers, energy usage, et cetera. So because we’re really just the beginning of this journey, it’s not easy to answer.

Having said that, if you look at how many big tech companies are leaning in and investing heavily in this space, you can expect that the market and the market readiness and the scalability will come quickly.

SIGN UP TO OUR NEWSLETTER

The freshest data straight to your inbox

What advice would you offer to founders of generative AI startups, particularly those navigating the challenges of fundraising and building their ventures? Are there any common pitfalls you’ve observed that founders should be mindful of?

Yeah, make sure that you have a team that has technical expertise, and at the same time, commercial knowledge about the industry and the market you’re trying to enter and disrupt.

Also, make sure you have a good answer about where your data is from. What are the data sources and how did you obtain the data? Is this data available to everyone or did you have a unique angle to have obtained this data?

Talent is also really important, as I mentioned it earlier. It takes a long time to develop talent in machine learning and in generative AI. So there isn’t that much right now out there. So it’s a question we’ll ask and it’s a question you need to answer, focusing on where and how you can recruit people to your company, to your vision. You’ll compete with some of the biggest companies in the world on this, so that’s not easy.

Tell us a little bit of some success stories of companies that you’ve either funded or met in the past. What makes them unique?

We can go through a bunch of examples. We invested in Hour One in 2019. They’re developing a stable diffusion model which can create video from text. So every pixel is generated synthetically. That was fairly unique back then. We see a lot more companies entering the space these days. It’s an exciting company that recently created the digital clone of Reid Hoffman. I remember back then how the team was very strong and unique, they had the technical knowledge and the industry knowledge needed to succeed in this space.

Another example that I like is an investment in a company called Cleo, which took the approach to verticalize Generative AI. So they’re creating a model of their own, not an LLM, a smaller engine that will be able to create games from prompts. So within two or three minutes, for two to three dollars, you can create entire games that people can play. So the idea is to become the largest game store. The idea is to utilize this technology to reduce the cost and the hurdle rates for anybody to create fun experiences. And that’s really, really exciting. It’s really just the beginning, but they’re making really great progress.

Maybe lastly, the last investment made is in a company called Blue Games, which is not building its own GenAi model, but they’re productizing generative AI and they’re creating gamified content. For example, Pictionary, where they are manipulating prompts in a certain way that then creates images which you have to correctly figure out and guess what the phrase or the word is and you can play that together with your friends. And so there’s a lot less tech there, but there’s a clever way to productize the technology.

Looking ahead, what do you believe are the most promising applications or areas of growth for generative AI technology? How can founders best position themselves to capitalize on these opportunities and drive meaningful impact in the years to come

Yeah, this is the $1 billion, or actually the $100 billion, question right now. To be fair, I think nobody really knows because there’s still so many uncertainties, some of which I mentioned earlier with regards to regulation. And we don’t really also fully understand how these black boxes work, right? I’m referring to the LLM. It’s developing at a breakneck fast pace.There’s new research papers out every week. And so it’s really, really difficult to know.

However, there are a few, I’d say, guideposts that I would like to share. So the difference in this technological evolution or revolution, you can call it, is that big tech is really leaning in. If you listen to the latest earnings calls from Microsoft, Nvidia, and Salesforce, and all those guys, Google and Meta, they’re all investing billions of dollars annually into generative AI. So as a founder, you want to avoid competing with them. You want to avoid that platform risk, where you create a product that is not be offered by OpenAI today, but it’s likely they will offer it down the road.

So, I we recommend to really focus on verticals, and where you have unique insights and also access to data to train your model. You can build some probably smaller models that are a lot less costly and, and then, become the leading model and the leading company in that specific vertical. So that’s, that’s one tip.

The other are where I think a lot of people aren’t looking at right now is, but it’s still interesting is on the consumer side. So obviously on one hand you have gaming, but you also have other consumer companies that can simply make use of GenAI to differentiate and improve upon services and products that are out there right now.

What’s interesting is that these startups can raise a lot less capital. So there’s a capital light approach to this and you should be out of the gate and get to market much, much quicker. And it should become a lot cheaper for everybody to enter spaces like these. And if it’s a consumer space where you don’t compete with Microsoft, but with a CPG company, they’re much less likely to be able to compete with you so you have a lot less platform risk.