AI Summit 2018: robots are a people’s problem
Robot development can only be done with humans at the lead
Artificial intelligence (AI) can help automate and solve many problems in the industry that require a lot manual work, but there is a human factor that is paramount to managing these systems and keeping them in track, according to the members of the panel “What will be the impact of AI on the Finance Sector?” at AI Summit 2018, who all seem to agree.
At AI Business’ event, held on 13 June at London’s Excel exhibition centre, the opinions were clear. AI will certainly improve how efficient our businesses run, but it is just a tool, and it needs to be monitored and managed by humans. After all, people set the rules and the criteria by which it functions, as the panellists said.
More importantly, there is the whole area of ethical applications of AI, which is not fully solved yet, and it brings many more repercussions than it may seem at first glance, as pointed out by Helen Crooks, chief data officer, Lloyd’s of London – the insurance provider, not the bank.
Among these, we find predictable outcomes like a shortage of jobs, but also the related taxation problem that comes with a smaller number of incomes to tax, or even less politically correct problems.
For reference, just a couple of years ago, Microsoft’s Twitter bot experiment ‘Tay’ was spewing racist comments in under 24 hours of interacting with users.
Although this may be an extreme case, it is possible that AI may lead to unfair and eschewed conclusions if let loose, and that would have a big impact in areas like insurance.
Richard Potter, CTO Services, Microsoft UK, pointed out that alongside ethics, he believes that there are matters of fairness, accountability and transparency in telemetry that need to be addressed throughout the development of AI.
Potter added that designing the business’ purpose and strategy is even more important than the application of AI. And all this needs human interaction.
The corollary of this is that robots will not wipe us out. At least not of us, and in most cases humans will still need to be present to monitor and keep developing this tech.
The panel also included Annika Schröder, AI programme lead, group innovation, UBS; and Sohail Raja, UK chief digital officer, Societe Generale.
Read more about the role of data and hardware in our first summary of the event here.
