ACT-R the theory is embodied in ACT-R the software | Hacker News Reader

ACT-R the theory is embodied in ACT-R the software

act-r.psy.cmu.edu

31 pointsTomte10 comments

Explainer of the ACT-R theory. There’s a lot more at the link than what I’ve quoted below https://www.scotthyoung.com/blog/2022/02/15/act-r/

> ACT-R argues that we have two different memory systems: declarative and procedural.

> The declarative system includes all your memories of events, facts, ideas and experiences. Everything you consciously experience is part of this system. It contains both your direct sensory experience and your knowledge of abstract concepts.

> The procedural system consists of everything you can do. It includes both motor skills, like tying a shoelace or typing on a keyboard, and mental skills, like adding up numbers or writing an email.

> ACT-R explains complex skills as an ongoing interaction between these two systems.

slightly off the topic. Scott Young is a great youtuber and podcaster. I don't understand why he is not as popular as 3b1b.

If it makes you feel better I've never heard of either. Maybe he is as popular!

Related. Others?

ACT-R: A cognitive architecture - https://news.ycombinator.com/item?id=10750795 - Dec 2015 (4 comments)

ACT-R (Cognitive Archtecture) - https://news.ycombinator.com/item?id=7864148 - June 2014 (3 comments)

Act-R is written in Common Lisp. The repo download is here[0].

[0]: http://act-r.psy.cmu.edu/actr7.x/actr7.x.zip

ACT-R is a laudable effort in formalizing a model of human cognition. It is one of the very few concrete models of human cognition there is.

Unfortunately it's wrong flat out wrong (as opposed to not even wrong). We don't encode the world in symbols, we don't store such symbols in memory slots and we don't reason about such symbols using rigorous symbol manipulation.

I had a similar impression when I took a course on the matter ages ago. I don't have a better cognitive model, but the methods we were shoehorned into with ACT-R were so different from the actual cognitive processes I undertake in completing the same tasks that it all felt so pointless. Short of results derived from massive hyperparameter tuning to overfit existing data, has ACT-R shown any real-world explanatory power?

Yes, very much so. During the last 20 years, a major project has been to predict sequential patterns of brain activation during various tasks based on the activity of ACT-R modules (which have been associated with different brain regions) in models of the task.

All models are wrong, but some are useful. For several decades, ACT-R has been, and continues to be, one of the most useful and influential theories of cognitive architecture devised.

When you say we don't store symbols (in memory slots), would you also disagree that we store symbols at all?

I've never been a fan of symbolic AI. It always felt like the fuzziness of neural nets and pattern matching were a better fit for how we seem to process information, at one level. But it seems like we also perform symbolic logic, typically on demand, possibly as a layer on top of or overlapping the NN foundation.

Reply on news.ycombinator.com