In this video, Graeme Cox explains his first principle of planning an AI Strategy - Value.

VIDEO TRANSCRIPT

The question is really, how can I make use of AI to improve my organization? What are the what are the deeper integrations that I can look at beyond you know, marketing having, having a tool to help it improve its English language output. My software developer is having a co pilot that allows them to write boilerplate and base code faster than they could before. Where, how can I look at the the core date for my organization and use it for effect Whether that is adding new features and capabilities to my customer facing tooling, where if you have a SaaS platform, or customer applications that you that you deliver, or if that is in internal efficiency and productivity in improving access to internal data or use of that data to to provide insights that weren't otherwise available?

I like to run a a workshop based approach, working closely with the CIO CTO and the technology leadership to understand, the business and its needs to, to come up with a stack ranked list of opportunities and, and initiatives that could be undertaken based on the four pillars that I'm showing on screen at the moment.

And just run through them very briefly.

They are always start with value.

Many AI Skunkworks projects that I see, when when I go into organizations are have been initiated by the by people in the technology team, and they have done some great work. You know, I I'm a technologist and a software developer myself, and, you know, I I I love this stuff.

But far too often that is has been initiated and driven by what seemed cool and what looked possible.

Rather than what actually will make a difference to the business.

So the first, step in the process is actually coming back to looking for some way of measuring ROI on defined initiatives.

So a workshop process starts with an ideation session, blue sky ideation, the the the just the what could be done.

Outside technology altogether where possible, looking at where areas of the business could potentially be removed from a an information provision, prediction, optimization, or other use case.

And and starting to from that to to look in the workshop at what the potential value is and then take that away to actually try and draw out what the long term ROI of that might be.

ROI doesn't have to be measured in pound notes. Obviously, it's great if it can be. But there are a number of areas around efficiency improvements that are quite hard to drive back into bottom line improvement.

There are steps along a road towards an AI transformation around developing AI maturity, which can show measurable value in terms of improving the capability of your the organization, but again, that's parts of a stepping stone to a longer term picture. So ROI doesn't have to be measured in in mon monetarily.

It's great if it can be.