Yesterday I was invited to represent ThoughtRiver and raise questions with an evidence panel looking at the impact of AI on existing and future business models.
The evidence panel was chaired by Stephen Metcalfe and was well-stocked with industry veterans including Mike Lynch (Invoke Capital), Husayn Kassai (Onfido) and David Kelnar (MMC Ventures) amongst others. There was a full audience including Richard Tromans from Artificial Lawyer and Richard Susskind.
Some of the most interesting observations were about data. In AI, data is the fuel that makes the engine work:
a) data explainability - the inherent difficulty with some AI algorithms (e.g. deep learning) of explaining why a decision has been made, with impacts on discrimination, insurance and other areas of policy (Mike Lynch)
b) data asymmetry - the danger that startups of the future will be shut out from innovation because incumbents will have hoarded data assets (Husayn Kassai)
c) data rarity - for AI innovators wishing to avoid direct competition with Google, Microsoft etc, the strategy of exploiting intelligence opportunities in more rarified domains where the datasets are more difficult to access (David Kelnar)
Overall the sense is that outside of genuine experts (most of the people in the room), people are still some way from understanding what AI is and what it can do. As if we had slipped back 180 years and were discussing this new thing, 'commercially available electricity'.
This is where products like our own Contract Intelligence platform - and others mentioned on the panel - will educate the market in the next 5 years. By packaging up AI as easy to use automation tools; making the complex simple for the end user. A 21st century lightbulb.
Tim Pullan
