Predictive Processing Therapy (PPT) I - The revolutionary approach to harnessing breakthrough cognitive science for mental health. It took me 60 seconds to develop based on a podcast.

Till Gebel
May 25
2 min read

Updated: May 26

NOTE: This is a fun experiment with generating a new therapeutic approach from a single postcast by Lisa Feldman-Barret.

Deconstructing sadness: Lisa Feldman Barrett's 2018 book "How Emotions are Made" impressed me from the start: the book begins by deconstructing the difficult feeling of sadness.

predictive processing — Click on the image to go to the app

It reminded me of a psychedelic experience: Once, I had experienced sadness in a new way: in a ceremony, I had been deeply sad, but only at the physiological level. All signs of sadness were present (tears, movements), but I watched this display curiously without feeling the burden of sadness.

Sadness as a construction, based on the prediction what one will feel

Sadness, as all therapists agree, is a particularly difficult emotion to bear, so much that we often silence sadness with anger. This is particularly true for men, who still are not supposed to be weak. Despite all the rumours that men now can be sensitive and show tears, it seems that most women still prefer the other kind of men, if one digs deeper. Thus, men in particular cover up sadness by turning it into anger.

So I listened to Lisa Feldman´s recent podcast on "The Diary of a CEO". She now has integrated predictive processing theory ("the Baysian brain" hypothesis). It interprets the quality (or also the qualia) of our experience as a result of the predictions our brain makes at any moment.

This contradicts the explanation of emotions as as reaction to objective reality, based on a fixed set of culturally invariant emotion-patterns.

And obviously, it is somewhat "liberating".

Emotions, predictive processing and Buddhism

Lisa Feldman-Barret´s theory is also quite close to the Buddhist concept of "emptiness of self". In this view, emotions too are fabrications, constructs, devoid of independent self-existence (although they are felt as real).

Predictive Processing explained: Artificial Intelligence turns a podcast into a "new therapy"

Obviously, in her theory there are quite a few handles for people to reduce their "suffering from emotions". In the podcast, she recalls a few examples for possible (self-) interventions and strategies.

And, not surprisingly, they show a lot of overlap with all kinds of existing therapies and with mindfulness / meditation.

Using AI to create a new therapeutical approach

So, as a fun exercise, I asked Claude 4 to create a new therapeutical framework from Predictive Processing, based only on the podcast transcript.

Here is the entire prompt, together with the transcript:

Create an interactive fascinating web application showcasing this new type of breakthrough therapy based on neuroscience . Make it a standalone HTML application. Add examples for specific interventions.

To be frank: not all ideas that Claude extracted are new. It combines elements of NLP, mindfulness, ACT etc. It´s definitely a cognitive type of interventions

But, this was just the result of 60 seconds. Have to focus more!

Have fun!