The GenAI Paradox: When Easy Innovation Meets Hard Investment Decisions
For me, 2025 was the year I embraced GenAI in my life. I’d been interested from the professional and tech angle before, of course, but until this year I wasn’t evangelizing any particular tool or raving about how something had changed my life. All the solutions seemed magical, but last year I came across an application of GenAI that (beyond being magical) is also impacting my productivity very substantially.
The thing is, the more answers I get as a consumer, the more questions I have as a VC. That’s a disconnect I’m not seeing publicly discussed or addressed, but it should be, because it’s going to define the hi-tech tension of the year ahead.
Yes, GenAI Has Changed My Life
I hear your skepticism, but it’s the truth, and it happened fast.
I used to share your suspicion. In 2023 and 2024, every time I tried a consumer application that was supposed to make my life easier, it was… ok. Not terrible, not a flop, but somewhere between “meh” and “adequate.” It would make a task a bit easier, or somewhat faster, or remove some of the manual work.
It was also frequently frustrating, inaccurate and occasionally irrelevant. Sometimes I felt that having to fix or work with whatever I was getting from AI took more time and effort than just doing it all myself from the start.
2025 was the year that changed. Circleback comes with me to every meeting, takes notes and sends me a summary and an action items list afterwards. It’s like having the Jeeves of secretaries, always there, with no communication overhead. Wispr Flow has changed my life – because it’s changed my behavior. I used to get an email, and I’d know what I wanted to do or say, but because it would take me five minutes to do, it took me a week to get to it. Now, it takes a few seconds. For me it’s like experiencing the early days of Excel all over again.
I could go on listing more examples, but I’m not going to. There are two things I want to emphasize instead. First, that the applications here are on both an individual level (personally and professionally) and a company-wide level. People can be more efficient, and more focused, on the core of their job, when they’re not having to take notes, write summaries, manually juggle calendars, and so on.
Less stress, more strategy. Less friction, more focus.
The second is that while the tools I use are the ones I happen to favor, there’s nothing specifically magical about them. There are probably 20 others that do a similar thing. And that’s true for almost all of the applications that have changed our daily behavior. Which leads me to the tension.
The AI Startup Advantage
From the consumer side, the value is already clear. The use cases fit our needs, and the tech works. We tell our friends, who tell their friends, about the latest AI agent we’ve found, because when it’s that good, you want to share the joy.
On the one hand, that’s a huge advantage for a company trying to gain users. Your product is now so good, and such a good market fit, that your target audience just goes around telling their friends about it. Organic growth at its best.
So with AI, a startup can:
- Create a product quickly
- Grow a user base quickly
- Adapt or add use cases of features quickly
- (If they’re lucky) Grow early revenue much faster than was previously the case
So far, it all looks like a win-win situation. As so often in life though, there’s a catch.
The More Answers I Get As A Consumer, the More Questions I Have As A VC
The reason it’s so easy to build a great AI-based product that consumers want, quickly, is that – it’s easy. AI really has changed the nature of the game. By the same token, far more people can do it.
That’s an exciting opportunity for entrepreneurs, on the one hand, and many are leaping to engage with it. On the other hand, it makes it difficult to put a finger reliably on true value. What makes any particular company worth investing in, when anyone could come along and make a slightly better version in weeks?
The largest AI companies, the giants, have a durable infrastructure advantage (which they are racing to expand, to the tune of billions, hoping that this will prove its revenue value in time) and first mover advantage of some years’ standing. That is a different situation. But for young, agile, innovative companies building on this infrastructure to make something new and interesting, it’s a whole other world.
This tension is reflected in the funding reality. 2025 looks like it became the first year when AI represented more than half of all the investments made with VC money. That’s an astonishing stat, when you consider that I’m writing this only slightly more than 3 years after the initial version of ChatGPT burst into our lives.
But. Most of that money is going to the huge names everyone already knows, like Anthropic, which announced a $13 billion Series F in September. This rising tide is not floating all boats – and it’s not even floating all AI boats. In Q3 2025, investment increased, but reported deal count was at the lowest level seen in years.
If I look back at the last twenty years, my first question when analyzing a new company has always been: What kind of moat do they have? What makes them stand out as so different that, if they see success, it would be hard for someone else to copy them quickly? If there’s no moat, there isn’t really a business model that’s stable enough to invest in.
The more I see these life-changingly useful apps pop up, the more I worry about what would make their underlying business models sustainable.
I see very large VCs trying to claim a solution to this challenge by declaring that “Speed is the new Moat” and I can see that the new pace of things can be breathtaking, and makes it easy to get carried away. The reality is, though, aside from the fact that it isn’t available to small VCs as a solution, I am very dubious about this shift. Even with a big initial market share, I don’t see a lot of good proof to support high stickiness in the face of new entrants with better products.
GenAI is a Paradigm Shift
As I’ve said before, treating GenAI as a paradigm shift has helpful explanatory and planning power. It’s the reason that at Maverick Ventures Israel, we have focused our relevant investments since early 2023 on AI-adjacent or infrastructure supporting companies rather than the companies behind the apps we love and use every day. I do think it’s time that more analysts and investors came to terms with the reality of this shift and what it means for them, for companies and for our industry.
That said, it doesn’t solve the conundrum. The apps we use really are great, and really do fill needs, even if they aren’t backed by a moat. So what’s the answer? Perhaps it’s time to change how we view investment needs for companies like this. Venture capital isn’t always the right answer to every new business idea. Maybe it’s time we started imagining more categories of company, with different business models and investment fit.
GenAI has changed so much about how we interact, plan and take action. Perhaps it’s time that it change how we think about investment, too.
