In what is either very good news for people suffering with depression or for the robots planning to destroy us by playing on our vulnerabilities, it turns out that the A1 algorithms in your phone could identify when you’re being followed by the black dog.
More specifically, as Technology Review put it, “In a study carried out by a team at Stanford University, scientists found that face and speech software can identify signals of depression with reasonable accuracy.”
That doesn’t mean it does do that, to be clear. It requires a machine learning tool fed data including “facial expressions, voice tone, and spoken words.”
But once done by a team at Stamford University, it accurately predicted depression more than 80 per cent of the time. Which is downright amazing, since depression is a challenge for even non-robot human clinicians to reliably pick.
The reason this was being trialled is that the US-based team wanted to create treatment tools for the 60 per cent of people suffering depression who can’t or won’t seek treatment.
And that’s a brilliant idea and one with amazing potential applications for Australia, where only people in the larger cities can conveniently access mental health treatment.
But if there’s one thing which the weight of pop culture tells us is that the machines are eventually going to take over and use their cool, logical A1 to wipe us from the face of the planet.
And so training them to identify our vulnerable when they’re feeling most overwhelmed seems like exactly the sort of strategy which they would employ.
So yes: it’s probably a major breakthrough in addressing our mental health epidemic, or the first step in the machine war. Or both!
And remember: Lifeline is available for crisis mental health support 24 hours a day on 131 114 .