Would you try a depression simulator?
Can technology really educate or merely entertain?
Last year, researchers from the University of Michigan, Ann Arbor, Middlebury College, Harvard University, and the University of Texas at Austin introduced the world to Eeyore. Named after the gloomy donkey from Winnie-the-Pooh, Eeyore is a large language model (LLM) that realistically simulates depression. Its objective: create a tool to train therapists on how to communicate effectively with patients with depressive symptoms. The researchers claimed that Eeyore outperformed other LLMs on linguistic authenticity ("the chatbot’s wording, phrasing, and tone closely match how individuals with depression speak") and cognitive pattern authenticity ("the chatbot realistically reflects depressive thought patterns like selective abstraction and overgeneralization without exaggeration").
When I read about Eeyore, it made me wonder: can such a simulation be used to create awareness in the general public about how it feels to live with depression? That's what today's piece asks.
Support my work from India through UPI: goswami.t@okicici
or from elsewhere in the world by clicking here.
In 2013, developer Zoë Quinn released Depression Quest, an interactive text-based game that sought to simulate living with depression. The game presented players – who were asked to think like someone with depression – with different situations from everyday life. You had to choose from multiple options to respond to each situation.
Here's one scenario from the game:
You are a twenty-something person who's had a hard day at work. You are going to be home alone this evening, and you know you should use the solitude to work on a project you've been haphazardly tending to for the past few months.
But you panic at the mere thought. The prospect of facing what feels like an insurmountable task exhausts you.
"The stress is weighing down on you like a heavy, wet wool blanket." What are you going to do?
1. Order some food, grab a drink, and hunker down for a night of work (in the game, this option was crossed out, signifying that a person with depression was unlikely to choose it).
2. Reluctantly sit down at your desk and try and make yourself do something.
3. Turn on the TV, telling yourself you just need a quick half hour to unwind from work.
4. Crawl into bed. You're so stressed and overwhelmed you couldn't possibly accomplish anything anyways.
Each of these options would lead to different consequences and more choices. The point was to show how depression could makes it hard for you to do 'normal' tasks, how it forces you to adopt behaviours that aren't 'right' for you, and how it could trap you in a painful spiral.
Depression Quest didn't claim to be prescriptive or make any definitive claims about the nature of depression. It only aimed to spread awareness about depression and help fight stigma.
(The game received wide critical acclaim – but it also nearly destroyed its creator's life. Gamer bros were incensed that it introduced politics into gamer culture. This would be the start of Gamergate – a vicious misogynistic harassment campaign in the gaming industry.)
Long before Quinn, in 1972, American psychiatrist Kenneth Mark Colby simulated a person with paranoid schizophrenia in a program named PARRY (for paranoia).
A group of experienced psychiatrists analysed a combination of real patients and computers running PARRY through teleprinters. Another group of 33 psychiatrists were shown transcripts of the conversations. The two groups were then asked to identify which of the 'patients' were human and which were computer programs. The psychiatrists were able to make the correct identification only 48% of the time — a figure consistent with random guessing.
