Access Ai – It’s simple, we’re going to use data to stop suicide. Just that.
Originally published in Access Ai magazine, November 2017
Pete Trainor speaking at a Silicon Beach event last year
A London-based designer who provides mental health charities with artificial intelligence (AI) technology is calling for designers and engineers to focus on developing technology that can cure social ills.
Pete Trainor stated in a tweet yesterday, “It’s simple – We’re going to use data to stop suicide. Just that”
By ‘we’, Trainor was referring to an AI company he co-founded and launched last year known as US.ai, which is partnering with suicide-prevention charities and NGOs to give them access to big data.
US is a commercial business that supplies AI to banks, but it also has a social mission to work with third sector groups, particularly those charities tackling suicide and the mental health problems, such as depression, that lead to it.
Depression is the leading cause of disability worldwide, a major contributor to the overall global burden of disease, and at its worst can lead to suicide, the WHO reported in February this year.
Although there are many known and effective treatments for depression, fewer than half of those affected globally (in some countries less than 10%) will seek help.
Trainor argues that there is an increasing rate of mental illness, being driven partly by technology.
“We spent the last 20 years creating apps to make life frictionless”, he said, “by doing that we may have created detrimental harm to peoples’ psyche, their psychology.”
Talking to SU
US have created a conversational agent, known as SU – so named because it is “a reflection of us”. SU can help charities provide treatment to mental health sufferers. The hope is that people will have conversations with the system via their phones, and it will offer advice and support.
“The insight was that a lot of the men these charities were trying to have conversations with didn’t actually want to speak to a person – they were happy to speak, they had things they needed to say, but they didn’t want to speak to a human”, Trainor said.
If the technology proves successful, then it could overcome two of the biggest barriers that face mental health patients who need help: a lack of resources among healthcare providers and a social stigma attached to their conditions.
The power of data
One of the men who spoke to SU under controlled testing conditions, called Michael, was suffering with a drinking problem.
As Michael spoke to SU, the system collected data about the keywords which appeared frequently in interactions over four weeks.
When US reviewed the data, they found that ‘marriage’ and ‘divorce’ were topics that Michael had brought up frequently with the bot, even though he had never mentioned them to his human councillor
By gaining access to the data which SU collected, Michael’s councillor recognised that marital problems could be a cause of his drinking, an insight which could help her to save Michael’s life.
“We were not gonna save a life with a piece of AI”, Pete said. “We were gonna save a life by giving the people, the humans, more information and data than they had without it.”
“It’s going to be a really amazing couple of years when these charities, social enterprises, and NGOs are suddenly given that fuel that will give them all of that data and insight. There’s no downside to that at all.”