Sitting in a small, over-bright seminar room on the sixth floor of Western’s Social Sciences complex with five classmates, I bit the side of my index finger to hold focus. The topic — actuarial versus clinical approaches to assessment — may not sound compelling or prescient, but. . . Harvey, Dr. Doug’s teaching assistant was making a strong, if not riveting case for the former. It was the early ’70’s and computers were still finding their footing in the hallowed halls of learning. AI. . . well, just a sci-fi concept (or a steak sauce, if you’re a secretary of education). Hal (Heuristically programmed Algorithmic computer) had barely had time to go rogue and over-ride Dave’s request to ‘open the pod door’ — all in aid of protecting itself from human meddling. And we six were a hard sell for Harvey’s pitch favouring factor analysis over human instinct and intuition. 

Fast forward a half-century (OMG, has it really been that long!) Hal may have been sold for parts, Orange (Agent, Clockwork, or otherwise) has replaced Hal’s penetrating red eye with less than penetrating red hats, and the humourless, automated, monotonic (some would say, moronic) voice now issues assaultingly from a guy obsessed with the red planet and bad math. But it seems the arm wrestle between the human and the machine carries on.

I have to fess up — I use tests! Paper and pencil, increasingly digitally administered measures of personality, interest, intelligence, aptitude. Rorschachs, projectives — not so much. Fun at a party, but, well, a bit too close to tea leaves and phrenology for my taste. Still there’s nothing like a good, big pile of psychometric data to flesh out a report and lend a little gravitas to one’s ‘clinical impression’. (Dr. Doug would be rolling over at the mere mention of something as ephemeral as ‘an impression’!) And while we’re in confession mode, report writing is a whole lot less onerous with machine scored results and oft-used (dare I say ‘canned’) boiler plate phraseology. 

Enter AI. Slipping in noiselessly on little cat feet, summarizing sessions, prepping proposed overviews of appointments, ‘freeing’ practitioners to concentrate on the client (vs. finding a pen that writes), and obviating the cognitive log jam as the therapist struggles to remember what was talked about — an hour (or a day, or heaven forfend, a week) later. All with the disclaimer, of course, that ‘AI-generated content may be inaccurate or misleading. Always check for accuracy.’ Even AI has to enlist a little CYA’ing. Nevertheless with comforting monikers attached: ‘companion’, ‘assistant’, . .

And then there’s Abby (or Happi or even Eliza — for folks with one or both feet in the past) ‘in your pocket’ and available 24/7 (as the pitch goes), empathic, attentive, supportive (and cheaper!) than those pesky humans. Who wouldn’t opt for the relentless positivity of a less paranoid and clearly more benign Hal!

Finding a best way forward, whether it be in medicine, psychology, truck-driving, or widget manufacturing is presenting as the ethical dilemma. I read an interesting piece in the New Yorker today paralleling the late eighteenth century struggles in the UK textile industry with today’s AI ‘revolution’. In a few short decades, an artisanal population was displaced and decimated by automation and machinery. In its wisdom, the government of the time (an 1800’s equivalent of today’s billionaires boys club) saw fit to criminalize the ‘old ways’ (Frame Breaking Act, 1812), making open resistance to the more profitable ‘industrialized options’ a capital offence. This in favour of finding a sustainable middle ground. One where both the capitalist and the craftsman may coexist, the latter benefiting from innovation; the former incorporating the latter’s wisdom, experience, and skill. I have no wish to sacrifice, to remove from my practice all the benefits that have evolved over the past 50 years. Neither do I want to hang all the mental health practitioners who continue to bring their distinctly human expertise to the table.