You Really Need an AI Strategy Now

ChatGPT, announced only a few months ago, is really starting to have an impact. And suddenly, everyone is thinking about their AI moves. Or they should be.

First, the bad news. ChatGPT and its ilk are going to make a bit of a mess of things. They are certainly good enough to use, and they will be put to use, for good and bad, in a wide variety of ways.

• The AI genie is out of the bottle. There is no turning AI off. Ask the US or China if they'd like to win or lose the "AI race." When nations care about not losing the race, the race is on.
• No one knows how to regulate this stuff yet. Until that happens, the race is pretty much unrestrained. So messiness is pretty much guaranteed.
• Companies large and small are going to need an AI Strategy. While ChatGPT is the bright shiny object, and Google and Bing are working their "large language models" overtime, you need to figure out what AI-features you'll need in 2-3 years.  (Or maybe 1 year). And then a gameplan to get there.  For many companies that are still working to leverage data better, the hill you're climbing just got a bit taller.

The truth is that building AI models that actually work require tons of training data. Who has that? Companies. That's who. Companies like you. The core thing that all AI does, is find the patterns in large data sets. And you have that data.  You may have to clean it, improve it, and make it useable, but you almost certainly have it.

That's not just good news, it's actually very good news. In your industry, you may well have a very special data set. If you go to work thinking about how to use it, you have a chance to be first, or at least early in the game.

And it's pretty important that you try. In a few years time, leaders will have figured out how to leverage AI. Maybe to sell more. Maybe to innovate faster. Maybe to reduce headcount and shift their economics.  So it's not good to be far behind.

Of course, to do that, you'll need more than just an idea. Knowing how to do this is important. Knowing what's feasible, and not feasible, that's also important. Knowing how to test it and how to deploy it. And how to update it. And then how to sustain it. You'll need a whole lot of things you probably don't have.

Sounds like you need a plan. We can help. From getting your data cleaned to charting the course to building the models, to deploying the results, we can help. Contact us below.