Pocket Therapy

What if you had a shrink in your pocket? 

A tiny little doctor who could think, with infinite room on their docket? 

Right there to save you from the brink of despair, an on-demand confidant.

He will hear your problem and with you, right through, he’ll walk it.

AI, just like every other technological buzzword I’ve heard since I had ears, is being touted as the solution to the world’s many problems. People don’t have access to legal representation? AI. Need to improve your policy team’s efficiency? AI. Having trouble modeling Earth’s climate? AI. Certainly promising ideas, but no more than that. These are predictions that read like science fiction, not projections. In order to estimate the actual impact that AI will have on the world, we need to find where it’s taking hold today and extrapolate from there. 

So, I’ve been on the lookout for actual applications of generative AI, especially Large Language Models (LLMs) like ChatGPT, Claude, and Gemini. There’s a lot of talk, but where are people walking the walk?

My main professional focus recently has been health and safety. I turned my search there. Mental wellbeing – A.K.A. Emotional Ergonomics – is a primary concern of those in workplace wellness. For good reason. Insurance claims for psychological services like counseling and therapy have been rising over the years. Organizations sensitive to insurance premiums are seeking scalable solutions to address the root causes of these conditions.

There are many, many researchers and journalists documenting rising cases of depression, anxiety, and other conditions in Western populations. Trauma, a term historically used to refer to an individual’s mental state, has graduated into mainstream use in discussions about epidemiology and public health. 

Leaders in both the private and public sectors are clamoring to find fixes. And AI is the current solution in problem-hunting mode. Powerful organizations everywhere are trying to force AI into cahoots with their most intractable issues. Inevitably, there are a lot of bad matches. AI will probably not solve climate change in any direct capacity. This disconnect will continue, because big, hairy problems require multi-pronged solutions, take a lot of time, and many attempts to solve.

The probability that there is a concrete use case for AI in the arena of one of humanity’s grand challenges is unlikely. A more reliable place to look is bottom-up, consumer applications of new technologies. How are individuals using commercial-off-the-shelf LLMs in the health and wellness space today, to improve their own health? What use case is strong enough to motivate people to self-select into being guinea pigs? This is where we can find a market beachhead that provides enough grounding to act as the foundation for futurist thinking.

If Google is the 21st-century confessionary, Large Language Models are poised to be our new shrinks. 

As a specific example, Pocket Therapy is a real solution in action everywhere. Apparently, as with search engines, the absence of a human allows us to shed our inhibitions. All of a sudden, not only are we able to share our most vulnerable thoughts - but now we get a second opinion. Instead of anonymously posting your relationship troubles on a Reddit forum and soliciting advice, you can now consult a digital oracle born out of the primordial informational ooze that is the internet’s textual database. Wisdom of the crowd, on tap.

Noam Chomsky famously said that language is not a communication tool. Yes, we use it for sharing information between people. But only a fraction of the time. The primary function of language is for organizing our self. Language is, 99% of the time, an internal tool. It helps us negotiate our role in our community, our expectations of ourselves, and rationalize away our guilt.

Talk therapists guide people through the process of personal integration. Clients share anecdotes, scenarios, and voice their struggles. People, today, are doing the same thing with LLMs. I’m incredibly curious if this is going to work. If Chomsky is right, I think there’s a good chance that LLMs hold tremendous potential as a tool for personal development.

Of course, the orientation of that development will be downstream of the “wisdom” held by the oracle, the formulation of the questions posed to it, and the constitution of the individual who is now the patient of a mysterious psychoanalyst. 

This edge is unfolding quickly, so we’ll see. Personally I’m counting down the days until the American Psychological Association releases a formal statement about the use of LLMs as an intervention. The cat is certainly out of the bag. Talk therapy is officially over-the-counter. 

If you take this perspective, it’s no wonder that everyone is trying to protocolize therapy via AI. I mean, seriously – what a market. Eight billion people trying to make sense of their place in the world?