Artificial Intelligence is meant to change our lives radically in the upcoming years and people are starting to doubt its ethical consequences
The increased use of robotics and artificial intelligence for basic human interaction has the scientific community divided.
“Some people in our society like face to face. There is a generational conflict between baby boomers and millennials,” said Simon Montford, a moderator at yesterday’s AI TechWorld conference in London.
The contextual ability to understand, which is inherent in humans but has been seen to be absent in machines, was a major point of discussion.
How to teach robots to understand slang and sarcasm?
“You have to gather a specific dataset otherwise it is impossible to include,” iterated Javier Rodriguez Zaurin, head of data science at Simply Business.
Natural Language Processing (NLP) is a technology with limitations, but can also come in handy in its rudimental stage.
“Businesses naturally assume all people speak English. That is not the case in an airport. 50% of our customers are foreign,” stressed Dougie Anscombe-Stephen, digital innovation lead at London Luton Airport.
Quick and easy language exchanges can therefore be helpful when solving inquiries in a timely manner, speakers suggested.
An audience member expressed surprise that no linguistic experts are used to create and implement NLP in business models.
AI in healthcare
The ethical issues become even more relevant when AI is applied in the healthcare sector.
“People are hesitant to AI because machines can make mistakes too,” said Arjun Panesar, co-founder of UK charity Diabetes.co.uk.
However, most experts agree prospects for early disease diagnosis and predictive analysis are promising.
The effects on healthcare insurance are yet to be determined.
“The misuse of data is the biggest risk,” argued Mark Tsimelzon, chief engineering officer at Babylon Health.
“Still, insurance companies want to keep you healthy because it is in their own interest,” he added.
Issues such as patient consent concerning data, and developing appropriate regulation, will be crucial in the implementation of AI technologies going forward, speakers highlighted.