EP 29: Alan King - Purpose First, AI Second

things like that.

What do you find is a pet peeve of things people get wrong then? So looking at that mix between pushing forward and making sure you're doing it safely and ethically, what are some of the big mistakes you see over and over in different organizations?

Well, I think it really depends on the organisations.

And if you're talking to commercial organisations, they're often quite keen to go very fast and perhaps take risks.

Their appetite for risk is higher.

If you're talking to, say, charities or government or associations, then there's always a much greater focus on ethics and governance.

And often they have trustees and their structures are such that it's not a case of one or two key decision makers and it's kind of greenlit.

It often has to go sort of via committee.

So you have to make sure that you present information in a way that will bring people with you to make sure that everyone's comfortable with the potential outcomes.

So there's an interesting point that you make about risk.

Is looking at risk as a first assessment a good way of deciding what to do with AI when you're looking at a problem?

Yeah, I think so.

One of the things that you have to understand is risk really is the centre point of probably how to go forward.

Because if you don't actually acknowledge that at the gate, you can then get yourself into a lot of problems down the line.

And there's always a knock-on effect.

So we always do... a readiness assessment effectively understand what's going on in the business already or the organization who's already using ai because there will be ai happening whether people have sanctioned it or not there will be shadow ai taking place and potentially you've got data leakage you've got information flowing to places where it shouldn't be potentially users or customers data being exposed to tools all around the world where it could be breaking the eu gdpr acts or forthcoming ai in your act so to build that profile straight away is really important and then that allows you to kind of lock in and also it helps you understand what the risk appetite is for the organization itself and as i said some will be more ready for risk.

They'll be prepared because they want to gain some kind of competitive advantage.

They won't want to break GDPR, but they might be prepared to be more progressive in how they apply what they do.

Other organisations will be very pulled back and they'll be very worried about how they're perceived.

Risk is a big problem for perception, say it's a membership organisation.

If their members perceive that AI is exposing the organisation to various issues or problems, that can be a big credibility issue and then that can have a knock-on effect for the members as well.

So I think ultimately it's about getting that sort of definition, that line in the sand, you understand where you are.

And then from there, you can start to think about how you structure things,

what that's going to look like.

I suppose that makes sense because under the surface, AI is essentially a large statistical engine running