Hi, Bryce from Columbus, OH

[Bryce_Bate]

My name is Bryce. I had a 40-year career as a software developer (Sr. Systems Engineer) at Ohio State University working on the educational side of computing. I’ve been dabbling in AI since the days of Roger Schank’s rule-based LISP systems, through topic modeling using gensim and LDA, and now LLMs. None of my former colleagues understood my passion (obsession) with this area.

Now retired for several years, I’ve been working on an AI-based application for kids to help teach critical thinking skills through interactive conversations discussing mystery stories. My background ranges from a degree in psychology to doctoral work (A.B.D.) in philosophy where I focused on epistemology.

I’ve been following Jeff Hawkins’s work for years and now the TBP. I’m fascinated by the playground it may afford us in studying the way in which the interactions of simple elements, combined with a base set of principles, grow in complexity to result in the emergence (that magic word) of properties and features that sum to something we call “intelligence” or even “understanding.”

I think there is a convergence of ideas by a (non-exhaustive) list of people, perhaps organized roughly like this, based on my recent but limited readings:

Where and How Models are Built:

These thinkers describe the physical “machinery” that allows intelligence to emerge.

• Jeff Hawkins
• Gaurav Suri
• Charan Ranganath

Thinking as “Internal Doing”:

These thinkers explain that “understanding” isn’t a passive state, but an active, physical simulation.

• Lawrence Barsalou
• Lisa Feldman Barrett
• Benjamin Bergen

Prediction and Efficiency:

These thinkers explain the “operating system” of the mind.

• Andy Clark
• Nick Chater

Culture, Language, and Tools:

These thinkers explain how our “simulations” are shaped and directed by the world around us.

• Lera Boroditsky
• Andy Clark (again)
• Benjamin Bergen (again)

I don’t know if I’ll be able to contribute to the code base. That hasn’t been my role working with Claude (where was Claude when I needed it?). But, I hope that interest in the broader picture may help spark ideas within the team. We’ll see.

This is something very special. I want to play some small role, if I can. This project has a purpose that is meaningful to me. At almost 73, finding meaning and purpose is important. Thank you all for giving me this chance to help make purposeful meaning with you.

[brainwaves]

Welcome to the forum @Bryce_Bate! We look forward to your contributions in whatever form they take.

[vclay]

Hi @Bryce_Bate ,

welcome! That sounds like a very interesting background, and your interests align really nicely with ours.

I see you’ve already made several other posts on the forum, which is also a great way to contribute to the project. This is a huge endeavor and we are happy about anyone interested in following this exciting journey in one way or another. There are many ways you can help us move this project forward besides code contributions. For example, by sharing relevant ideas you’ve come across in your career or by creating educational resources around our project (I see you’ve already started sharing some videos, as well as the related researchers you mentioned in this post ) Here is a longer list for some more ideas if that helps Ways to Contribute

Best wishes,

Viviane

[Bryce_Bate]

@vclay

welcome! That sounds like a very interesting background, and your interests align really nicely with ours.

I see you’ve already made several other posts on the forum, which is also a great way to contribute to the project. This is a huge endeavor and we are happy about anyone interested in following this exciting journey in one way or another. There are many ways you can help us move this project forward besides code contributions. For example, by sharing relevant ideas you’ve come across in your career or by creating educational resources around our project (I see you’ve already started sharing some videos, as well as the related researchers you mentioned in this post ) Here is a longer list for some more ideas if that helps Ways to Contribute

Best wishes,

Thank you @vclay for your encouraging words. I started by thinking I’d be an observer of this exciting project. Now, I’m finding myself totally caught up in an idea I’m still formulating. I’m actually playing in the code! What I’m starting to do is treat Monty as a something to be studied through experimental psychology. To be more precise, experimental “synthetic psychology.” I like that term. It’s not mine. I think it’s from Valentino Braitenberg’s work on embodiment in machines (1984). That’s what I’m looking at, what Benjamin Bergen (“Louder Than Words”, 2012) calls “embodied simulations.” But more to the point, I’m interested in the predictable breaking points in these simulations and how language can effectively “blind” the physical sensors.

I’m working on demonstrating, following classical work in experimental psychology in Cognitive Dissonance Theory (Festinger, 1957) and the Priming Effect (Meyer & Schvaneveldt, 1971) as starters, how simulations can lead to failures in Monty’s internal “mental state” resulting in behavioral failures.

Already, I’m seeing some evidence of how “priming” Monty’s Goal State Generator (GSG) can produce a kind of “inattentive blindness” (Chabris and Simon, 2010) where Monty remains blind for a predictable time (i.e., taking more steps in recognition) to what the sensors are presenting as reality.

I think that demonstrating how Monty is susceptible to the same cognitive failures as humans could be a prerequisite for building more a robust AI. That’s the practical benefit of this synthetic psychology approach. It’s like studying and applying Behavioral Economics for machines. If I can investigate some of these “breaking points,” I think it also helps validate the Thousand Brains Theory (TBT) by showing that there is alignment of the TBT to the Monty implementation, even in how cognitive failures occur.

I’ll share more data as I get into it. Your thoughts and suggestions will be much appreciated. Somehow, this project has opened up an opportunity I wasn’t expecting, something really purposeful to me. It’s pretty exciting to think I’m actually watching the priming effect and cognitive dissonance surface in a machine in a way I can literally see these things in the logs. It’s a bit mind-blowing, to be honest.

I’m hoping this synthetic psychology approach can help show that Monty isn’t just a pattern matcher—it’s fallible just like humans. I think we can leverage that to great advantage.

– Bryce

[vclay]

Hi @Bryce_Bate

that sounds super exciting! I saw your recent post about this and am looking forward to hearing more about where your experiments with Monty go It would certainly be really exciting to see the same kind of perceptual and cognitive errors humans display, demonstrated in Monty. I am happy to hear that this has turned into a meaningful task for you and if you ever run into questions around the code or how to test certain ideas, feel free to post those.

Best wishes,

Viviane