In certain circles there seems to be a lot of preening around the concept of “AI Native.” When I start asking what that means, it’s difficult to get a definition.
For some, it’s start-up companies bringing products to the market that are wholly AI based. And like many startup, there is a lot of bragging about double/triple digit growth, enormous market caps. (Hmmm, I’ve seen this movie before….)
Some of these things focus on everything being done within the AI itself, but with a few exceptions, like art and video, these systems depend on outside data and information to do things–at least business applications. In short, their value exists only through access to data from other places.
Perhaps the most reasonable explanation is the offerings have been built natively using AI tools, rather than previously developed products with AI tools bolted on. But if AI Native is a meaningful distinction, then I might also make the same claim for developing software natively in Python. Does “Python native” mean anything important? Will it raise the valuation if I went to VCs claiming this and asking for investments.
To exist in the real world, these tools have to interface with all sorts of different systems. And this underlies the complexity and challenge of making any system work. The fact that something is AI native is absolute nonsense, because for anyone to get value from the tool, it has to be connected to non AI based systems, processes, and technologies.
And few of these tools exist in a vacuum, operating completely autonomously. There are humans involved in some way–whether it is defining the work that needs to be done, leveraging the output produced, or using it to help them think differently.
We miss the real value of AI when we make it all about AI and the technology. It is never the AI itself, but what it enables us to do, how it is integrated into all the other messy stuff that exists in the real world.
We are leveraging these tools in existing businesses, all with legacy–perhaps antiquated systems. There are processes, technologies, disparate data, entrenched workflows. There are huge numbers of disparate systems, that need to interconnect, sometimes clumsily, but all to get work done.
The fact that something is AI-Native is about as meaningless as my vision of Python Native. What we are talking about is tools and technology, that must be integrated into all these systems and human interactions to produce meaningful value.
And that’s the challenge we seem to miss in many of these discussions and the difficulty we are discovering with leveraging the power of AI in the most impactful ways.
I’ve been advising the founders of an AI native tool. It’s a fantastic tool, but the challenges they face are those that I’ve outlined. Where do they get the data, how do they integrate it in a meaningful way, how to they present it in a way that maximizes the impact, how do they inject changes back into the legacy systems. The challenge isn’t the AI native tool itself–isolated from everything else (leveraging LLMs) it works beautifully. But it’s when integrated with everything else, the challenges arise–and then every user is different.
Other AI native companies I work with see similar issues, some are managing it by substantively narrowing the scope of what they are looking at, others are putting very tough guardrails to more effectively manage how the tools are used.
And other technology companies are enhancing the capabilities of their legacy platforms, augmenting them with AI, providing entirely new capabilities and use cases.
And I could go on.
The point is, for some reason we are making it about the AI. Whether we call it AI Native, AI enhanced, AI supercharged, AI compatible. The real point, just like everything we have seen in preceding technologies, it’s how we integrate these into our work, improving everything we do.
We will know we have succeeded when we stop talking about the AI, knowing it is an inseparable part of how we live and get work done.
Afterword: Here is an outstanding AI generated discussion of this post. Enjoy!
Leave a Reply