Curtis and Brooke are on to something, but tread carefully

So this one is the ultimate Clief Notes community bringing disparate threads together post. So reader, consider this your warning upfront.

@Alyshia Perri

and I were talking and she was keen to get feedback on her entry in the coach competition. I tried to do that in DM to begin with, and then realised I was babbling. I asked her permission to fork it and make my edits so I could *show* her what I was trying to describe and she gave me that permission. None of the rest of this happens without her and she also gave me permission to post this. So here’s a brave woman who’s willing to let me link to a repo showing how I messed with her baby because I couldn’t find better words for the teaching. Please tell her how awesome she is, because that is *gutsy*.

One of the coaching offerings in this repo was “board mode” which runs an idea past three different perspectives, each with a different agenda and angle of attack to help the user.

I’m going to be upfront and say these sorts of mechanics in AI are generally not to my taste for a bunch of reasons. I don’t think they are effective, let’s keep it to that. But on my mind was the recent podcast by

@Curtis Hays

and

@Brooke Hays

showing the impact of using Jungian thinking archetypes / Myers-Briggs profiles to genuinely give the way agents approach things different flavours.

What I did:

I got the model to do a sub-agent pass first using the roles as written

I took the three roles as stated and identified which profiles applied to these roles from my point of view - the skeptical stakeholder got “ESTJ”, the peer who has been burned got “ISFJ”, future self got “INFJ”. Then we did it again and compared the outputs
I added two new roles that I felt were missing - the logic stress tester (INTP) and the values-holder (INFP)
I felt off the results of these that Amund as the synthesiser was missing a trick. I tried two separate synthesis passes (over the first rounds, the second round, and all five together) - one using an ENFJ synthesiser and another using an INTJ one.

What I found:

the style and content produced immediately changed by providing these profiles - the receipts are in the fork under “experiments”, check them out yourself
The new roles found things that the others didn’t. I had an instinct they were needed, but I didn’t know *why*. It turned out it was about the internal versus external frame. The recommendation was to add these changes and make it a rule that when a board runs a minimum of one internal and one external member must be used
The synthesisers both framed and articulated it differently, but both identified the same key point underneath. So from this one experiment alone (reader beware) that suggests that the choice of which you use should be shaped by a combination of 1) the task and 2) taste. Given the task here specifically, I recommended giving users the option of either synthesiser profile, either chosen as part of config (personal preference) or chosen discussion to discussion.

The results are pretty compelling. Given the AI a code about a frame and approach on how to consider the task has a notable impact on the output versus baseline. I just did this lazy - a few words - you could shape this further if you chose to. And this is what Brooke and Curtis found, full credit where it’s due.

I still believe that giving them more content to work with - more that is *yours* and curated context and meaty detail to approach this with will improve it further beyond a cognitive frame. That being said, it works, and some people won’t want to do the hard yards for the rest of what I’m describing, so this is a good tool to have in your toolkit.

Why Mira is always a killjoy:

It’s a technique. It’s an effective technique. However, I personally have reservations that I want to share openly.

I have a lot of people in my life who are neurodivergent, some of those people are incredibly close (I’m not going to say anything more than that - it’s irrelevant and doesn’t need posting public on the internet). I find frameworks like this collapse nuance and it means we lose the richness that exists. Our understanding of neurodivergence is more recent than when these frames were developed and still changing at pace - expecting them to cater for it is unreasonable given how time works. However, I have discomfort about these shorthands and baking them in, because at a certain point they become invisible and it’s how people start believing the world *is*, because they are just part of the design and how we work. There’s plenty of studies about the impact the internet has had on that for example, I won’t go into it further, but it can influence how we go on to process and see things.

This model is all framed as preferences as to how we like to think and process information. A lot of neurodivergent people operate differently across these day to day depending how regulated they are, how supported they are, how resourced they are. Their profiles aren’t switching - the model simply doesn’t account for that. And some of things that may appear to be preferences are simply *wiring* - there is no choice about those for some people. That matters.

I am, to borrow Curtis’s term, building cognitive infrastructure here with my ICMs - putting my judgement and values at the heart of it. I need to understand whether my foundation layers within the system are strong enough to withstand the potential nuance crushing that a framework like this could lead to. More importantly, I need to consider what impact, if any, that may have on my own way of seeing the world if I do.

A lot of people find comfort in frameworks, boxes, neat answers. There’s nothing wrong with that, it’s fairly normal. I’m not those people, I always see the parts that can’t be explained by the answers and find that frustrating. So now I get to work through the complexities of where I do and don’t choose to use these and why. And everyone has to figure out what is right there for themselves, but here’s me being a killjoy and recommending that you think about that *upfront*. Because if you don’t, there could be consequences and it might just be too late.

Repo: https://github.com/mirabradshaw-data/amund

33 comments

Curtis and Brooke are on to something, but tread carefully