#Connectionism
Also I don't know if you're aware how far back this history of what people mean when they talk about contemporary AI goes...
The metatheoretical calculus of connectionism — or: how we reason over artificial neural networks
YouTube video by Olivia Guest
www.youtube.com
November 14, 2026 at 12:16 PM
And that's only scraping the surface.

Basically, I define the field of NeuroAI as work that adopts ANNs as the core theoretical framework for understanding neural computation, i.e. studies who's intellectual ancestry can be traced back to connectionism and the theories of the PDP group.
November 16, 2026 at 8:28 PM
I did a full academic event yesterday without a single chat or mention in 8 hours of even the phrase "AI" let alone anything related to it in any sense (except my affiliation and when I said "connectionism" in the Q and A), probably for the first time I can remember since like 2020. It was healing.
November 11, 2026 at 6:56 AM
4️⃣ Guest, O. & Martin, A. E. (2026). A Metatheory of Classical and Modern Connectionism. Psychological Review. olivia.science/doc/guest_ma...
November 4, 2026 at 7:16 PM
Congrats! I'll try to get to it soon — it's always nice to see more people working on artificial neural nets' abilities (or lack thereof; like my & @andreaeyleen.eurosky.social's series of papers). I think connectionism (at least in cognitive, neuro, psych scis) has had a long often uncontested run.
November 4, 2026 at 7:48 AM
🧵 Here are some key takeaways on the multiple returns of connectionism, and their role for theory and history of #AI. What phenomena would a future resituated and resituating media theory of artificial intelligence need to address?
🔥 Your #AI is on Fire! Here's a piece to cool us down a bit in the midst of a super-controversy: “The Multiple Returns of Connectionism: Theory and History of Artificial Intelligence“ just came out in the Media of Cooperation working paper series. Happy to discuss this! 🔥 👉️ doi.org/10.25819/ubs...
November 1, 2026 at 4:42 PM
🧵 A second passage addresses the enduring neo-cybernetic impulse that leads to the emergence and recurrence of cognitivist and behaviorist connectionism (as with cybernetics itself, this primarily affects the human sciences).
November 1, 2026 at 4:44 PM
🔥 Your #AI is on Fire! Here's a piece to cool us down a bit in the midst of a super-controversy: “The Multiple Returns of Connectionism: Theory and History of Artificial Intelligence“ just came out in the Media of Cooperation working paper series. Happy to discuss this! 🔥 👉️ doi.org/10.25819/ubs...
October 30, 2026 at 7:01 PM
source? did she say "can't be implemented in" or "can't be learned by"?

symbolicists have acknowledged the possibility of "implementational connectionism" since 1988
October 29, 2026 at 2:45 PM
the objection that comes to my mind is different. "implementational connectionism" is a commitment about the end state, not the process used to get there. to whatever degree the counter position is correct, it's a criticism of the training process, not the capabilities of architecture.
October 29, 2026 at 9:03 PM
i think that if you are committed to soft implementational connectionism, you could reasonably say, "you see, it requires a very strong (i.e., almost temperature 0) prior in order to do this." i think that's wrong but i don't think anything anyone has done has strictly ruled it out.
October 29, 2026 at 3:06 PM
A while ago I read this interesting paper from the field of Education, called "Connectivism" (not connectionism). The central object here is a graph. It's quite interesting and easy to read: https://eclass.uoa.gr/modules/document/file.php/PPP130/DIMENTE09/elearnspace.%20Connectivism_%20...pdf
eclass.uoa.gr
eclass.uoa.gr
October 30, 2026 at 1:41 PM
A metatheory of classical and modern connectionism
repository.ubn.ru.nl
October 26, 2026 at 7:52 AM
Incidentally, the PDF link to "A Metatheory of Classical and Modern Connectionism" on olivia.science fails (tried two browsers).
October 26, 2026 at 7:37 AM
Sure, and also it depends which type of connectionism...
Another, related aspect, that I wanted to highlight in this work is the difference in stances — some much more damaging than others — that appears to have occurred around 2010. These are summarised in Table 1, here: doi.org/10.1037/rev0...; e.g. goals seem to shift from theory to prediction.

3/n
October 24, 2026 at 4:20 AM
Very kind! Maybe see this too, if useful...
The metatheoretical calculus of connectionism — or: how we reason over artificial neural networks
YouTube video by Olivia Guest
www.youtube.com
October 24, 2026 at 7:46 AM
So I did 45 minutes of me talking (this: youtu.be/4DHgVkIWhig) and then 45 minutes us all talking and chatting. What I like about that format is nothing then feels rushed.
The metatheoretical calculus of connectionism — or: how we reason over artificial neural networks
YouTube video by Olivia Guest
youtu.be
October 24, 2026 at 10:12 AM
Showing how anti-LLM I am by writing an art project in Prolog.

Connectionism shall not triumph this time! 🧐
October 23, 2026 at 3:27 AM
Casebeer - Natural Ethical Facts Evolution, Connectionism, And Moral Cognition (Mit 2003).pdf
Text PDF - 876.4 KB
Descarga en Archive.org
Pagina del item
October 22, 2026 at 5:32 PM
If you're interested in historical trends within connectionism, some take it much further back than Lovelace too:
if it's not useful nor interesting, please feel free to ignore — but it can also be helpful to understand our place in a complex web spun by so-called AI (sadly) as it has roots within our fields and therefore helps to understand our reactions to it as a (meta)theoretical force
The metatheoretical calculus of connectionism — or: how we reason over artificial neural networks
YouTube video by Olivia Guest
youtu.be
October 21, 2026 at 8:40 AM
The debate over how mathematical equations map to human thinking mirrors a fundamental argument in neuroscience about how our brain's nervous system actually operates: Symbolic Processing vs. Emergent Connectionism.

from general read trump half brin not working
October 20, 2026 at 1:49 AM
Casebeer - Natural Ethical Facts Evolution, Connectionism, And Moral Cognition (Mit 2003).pdf
Text PDF - 876.4 KB
Descarga en Archive.org
Pagina del item
October 17, 2026 at 5:17 PM
Guest, O. & Martin, A. E. (2026). A Metatheory of Classical and Modern Connectionism. Psychological Review. doi.org/10.1037/rev0...

PDF: repository.ubn.ru.nl/bitstream/ha...
October 16, 2026 at 7:20 AM
The general feeling in the field right now is that structuralism and connectionism will likely totally win out in mechanistic interpretation of models. This was a genuine surprise until yesterday or so. If the results hold up, people are expecting a complete connectomics of models.
October 16, 2026 at 7:40 PM