You can go slow with AI

I spent some time today learning about generative AI and integrating it into applications. There are a dizzying array of options and technologies to learn, and new innovations are being announced literally every day. It’s an amazing period in human history and I wonder if and when the innovation is ever going to slow down.

It feels like no matter what you build, it will be obsolete within a few months. Just adding AI to an application isn’t enough. Companies need to commit to a long term strategy and to building entire teams and departments dedicated to AI.

It’s really an extension to the big data movement. Machine learning and generative models are essentially a new and powerful way to leverage your data. Organizations have been gathering valuable data about their business and trying to understand it for years now, so in a way many organizations are actually well prepared for this new wave of innovation through AI.

Your data is your differentiator. Anyone can apply a model trained on public data. Only you can create a model that is trained on your particular business. In a way that gives data-driven organizations additional time to consider the right technologies and approaches, without the need to always be at the leading edge all the time. Even if you are a little behind the times on technology you will still be able to provide unique value based on your unique data.

It’s worth taking the time to understand the approaches and how you can apply them in your domain with your particular use cases. However there is no escaping the fact that you’re going to need to build up the teams and capabilities to continuously evaluate and improve as the technology gets more and more powerful.