Ever one to look to fields anew for inspiration, a friend of mine, the irrepressible Wael Elrifai, has suggested that I put my organisational sills, as demonstrated by the highly successful Data Migration Matters series, to good use and promote an AI event.
Wael is an interesting guy. With more degrees than the sun and a truly cosmopolitan outlook and lifestyle he wowed last years DMM5 conference with his presentation on data migration and MDM. Now if you can make that interesting then you can do anything. Owner of the niche consultancy Peak he and I have taken to chatting about the IT landscape as we see it from our respective perches. This usually happens in a bar in Islington where they do the most amazing cocktails. And as usual, after a couple of long island ice teas, followed by a couple of whisky sours, the topic turned to his knowledge domain - artificial intelligence.
AI is an interesting topic in its own right. And after telling me about his experiments with a model airplane - trying to get it to learn how to fly on its own, we turned our thinking to the uses of AI in a more business centric environment. (As an aside, I think learning to fly an airplane demonstrates the limits of AI, it takes and awful lot of crashes to learn how not to crash, which is hardly a good model for the human learning processes).
An obvious possible target for AI was big data - with all those possible relationships and insights lying around, should it not be the mathematicians with their self learning algorithms, not the software engineers with their more ham fisted approach who will pay dividends? Who is more likely to find the unforeseen?
I was by now, in a subject area way outside my comfort zone but heck, the evening was making me mellow, so I suggested that we needed to expand the conversation, maybe with more people who knew what they were talking about, like him, plus a few like me who could bring naive intellect to bear.
An idea was born.
This idea was for me to organise a small event one afternoon/evening where Wael and at least one other guest speaker would present on AI and we could all then discuss and sharing insights. The format will be a round table rather than a formal lecture. Aside from Wael we think the other speaker will be Dr. Giorgio Orsi a post doc from Oxford who has been working around AI and the semantic web. And I would announce it at DMM6.
Wael and myself will be at DMM6 to kick this off. We are looking for only a small group at first and to take it from there. So if you are interested either drop me a line or button hole one of us at DMM6.
Oh and my idea for a role for AI in data migration? Well I was thinking about profiling. Seems to be uppermost in my mind at the moment. I may have to blog about it in its own right soon. As we all know, there is no excuse for not profiling data these days. The problem is that if you point a modern profiling tool at a large database with multiple tables and allow it to do its stuff, you end with a mountain of results. How do we sift the wheat from the chaff? Seems to me that an AI approach might just be useful. Unfortunately I do not know enough about AI to know if this is right or not. But I will after this round table exercise!