AI natives and AI wannabes
I had a conversation with Amanda Silver from Microsoft recently, which turned into a profile I published earlier this week. Silver, who has been with the company for nearly 25 years, gets that startups have a distinct edge when it comes to embracing AI, no matter how well-intentioned an established company may be. A freshly minted startup unencumbered by legacy technology decisions can move more freely and make choices without the limits a large organization navigating a maze of existing systems has to face.
That got me thinking about the huge advantage a company starting today would have over one that started even as recently as a few years ago. Some of the benefits are those that every startup has over a more mature company, the ability to build from scratch using the latest technology — but today, right now, there are just so many ways to build faster with AI coding tools.
That’s certainly a point that Silver understands, even though she has spent almost her entire career working for one of the largest tech companies in the world. “Microsoft internally reflects the landscape that a lot of our customers are dealing with. We have 30-year-old code bases, and we have brand new code bases, and I think that a lot of startups have a luxury in that they aren't saddled with legacy,” Silver told FastForward.
The AI-native efficiency advantage
There is this new Silicon Valley fever dream, fueled by AI tooling, that a startup can build a high-impact company with just a few employees. Sam Altman famously said one day we’ll see the one-person company reach $1 billion in revenue. I’ll believe that one when I see it, but there is truth to the idea that you can build more with less, at least to a point.
A great example of this is Anysphere, the startup behind the AI coding tool Cursor. The company hit $100 million ARR with just 20 employees, according to the NYT. It’s not a one-person outfit of course, but it represents a level of efficiency that would’ve been difficult to achieve previously. While Anysphere is now valued at over $29 billion and has over 300 employees, that early trajectory shows what AI-native companies can do.
Ray Wang, founder and principal analyst at Constellation Research, appeared recently on CNBC's Squawk Box, referring to this phenomenon as “old tech versus new tech” — and he believes the newer companies have a huge advantage over incumbents. “AI-native firms run super efficiently with staggering revenue/profit per employee, can grow 10x–100x versus traditional 20–30%, and are built around high percentages of digital labor, enabling faster automation and outcomes,” Wang told Squawk Box.
MIT professor Bryan Reimer, co-author of the book How to Make AI Useful (along with Swedish futurologist Magnus Lindkvist), says the advantage is real. “As soon as you establish a framework and you begin to invest deeply in that framework, you become encumbered by decisions of the past,” Reimer told FastForward. “So for new entrants who are really bright and really adept at using these tools, man, will they be able to move fast in ways that are harder for other organizations to come close to,” he said.
It won’t take you all the way
That is not to say that there aren’t limits to this ability, or that humans no longer matter. I spoke to a founder recently, who says AI helped him build an initial version of the program. He was clear there were limits, and once he reached those, the company brought in a professional UX designer to bring them the rest of the way. But using AI coding tools saved him a lot of time and money getting started.
There’s little doubt that AI-coding tools can help a company, especially one starting from scratch, begin building and producing a working program much faster than was possible previously. AI can also generally help companies run more leanly than in the past, but we are probably many years away from Altman’s one-person startup vision if that will ever be possible.
For now, as Reimer told me, LLMs are just another tool in the quiver. They enhance humans, they don’t replace them.
~Ron
Featured photo by Fiona Dodd on Unsplash.
