The for-profit AI companies see their social good will as a potential for valuation and profit boosts. They will milk to the bottom to meet those ends: rate limits, price increases, and so on. It is not a question if the majority users will feel dejected but when.

And then for top-down veto power society already has a system for that with modern federal democracies albeit it also needs more advancement to more complex yet faster structures. 2/2

I think what Solarpunk has to do with is local social networks having veto power on larger social networks, esp. on matters of tech and energy use. That is what fast neuronal synchronization has taught me about bottom-up causality and transferring this insight to social analysis. 1/2

PPS: More thought preamble to comparing Pinotsis & Miller 2023 to Fakharian et al. 2025 (ephaptic coupling and motor control).

The meta is, as in spiking, there is ephaptic top-down veto power of networks on circuits and vice versa, in the same region like the hippocampus.

bsky.app/profile/neur...

PS: [Idea is that local neuronal circuits have a switch-based system for allowing external control. A hypothetical compatibilism framework.]

I’m using my Paul Feyerabend []’s to bridge ideas from Danny Dennett/Kevin Mitchell with ideas from the community of Buzsaki alumni that influenced my PhD.

Axo-axonic veto power. Just one cell-type instance. I want to go back to the idea of local circuits having self-determinism [note: my bias of (1)] the same way that a some bacteria has when sharing a common extracell matrix (2).

1 news.vt.edu/articles/202...

2 www.nature.com/articles/nat...

Are there instances for neurons where the duration of state A correlates linearly with the chances to transition to state B but not vice versa? 🧪

Trying to understand motor control via the idea of brain oscillations makes me doubt the idea of brain oscillations syntax without an explicit description of the biophysical substrate.

Thanks. Uff yea I’ve rarely done that in the case a direct repost might trigger a knee-jerk block or bring too much attention to myself for a hot take. Definitely in the “use sparingly” category because it is situational.

What is subtweeting?

Hypothesis: Neurons are often like uptight bureaucrats. It doesn’t give a crap what the networks says, if an input signal doesn’t match it will keep escalating an output error message until something changes.

Not sure. I can only see whatever those different glia do functionally is on timescales larger than a lot of spiking behavior (>10s of ms). Their membranes are not excitable with the speed that neurons, myocytes or specialized sensory cells like hair cells. Or if they did it would been seen by now.

Yup it’s a hot new trend to study glial function outside of the classic interpretation. Maybe still needs a bit more encouragement but it is definitely a good strategic place for career focus in the longterm.

Hypothesis: A theme of a kind of a “brodmann-ification” (lack of better word) of how ephaptic coupling is used in different brain regions with different geometry and cell-types over different spatial and time scales while many principles stay constant. Just with like synaptic diversity.

2/2

Rapid modulation of striatal cholinergic interneurons and dopamine release by satellite astrocytes - Nature Communications Astrocytes perform intriguing roles in neural circuits. Here, the authors show that in striatum, astrocytes adopt a satellite-like proximity to cholinergic interneurons, modulating excitability and do...

Another study related to ephapses and motor control systems but within within the striatal circuits of the basal go/go-no pathways.

“Rapid modulation of striatal cholinergic interneurons and dopamine release by satellite astrocytes”

www.nature.com/articles/s41...

There is a theme emerging…

1/2

Edit: slow-wave, connectomics/network*

Traveling waves lit has clues. Josh Jacob’s group had this blurb in ref to Anastassiou et al., Pinotsis & Miller. Note: Steinmetz group found spiral waves to be synaptic.

“Planar, spiral, and concentric traveling waves distinguish behavioral states in human memory”

www.nature.com/articles/s41...

With time context you can look at “shadows” cast by the ephaptic interaction from reference area to target area (make it pair-wise).

This is preamble to the comparison ephaptic/synaptic/gap junction control of saccades (cerebellum;spike jitter). Link below:

pmc.ncbi.nlm.nih.gov/articles/PMC... 3/2

Paper: “Cell-class-specific electric field entrainment of neural activity”

www.cell.com/neuron/fullt...

“Slow periodic activity in the longitudinal hippocampal slice can self‐propagate non‐synaptically by a mechanism consistent with ephaptic coupling”

pmc.ncbi.nlm.nih.gov/articles/PMC...

2/2

To buttress the slow wave the connectomics/network methods of ephaptic control of engram formation (SEF <-> FEF saccade paradigm from Pinotsis & Miller 2023) one can use electrophys signatures like entrained slow traveling waves (Durand lab) and cell-class-specific entrainment (Anastassiou lab). 1/2

Edit: meso-scale, needs*

PPS: The issue with LFPs/E-fields is they have the same issues as rate-based predictive models. You can have a plethora of neural tissue structure produce the same e-field that ends up entraining or shifting spike times. Forget about “epiphenomena”, how do you even validate the current sources!?

PS: I had some rants about methods before I get back to writing the thread and other items on different instances of ephaptic coupling and motor control like saccades.

I’ve had at least three FEP folks unfollow me across platforms. Sorry I just have the ambassadorial patience of Warren Mansell.

Okay I’m back to methods I like. You have to randomize to create null sets if you can’t do direct causal manipulation.

- Randomize spike times
- Randomize the neuron/tissue orientation for the sims of e-field generation
- ephaptic coupling, esp. at the meso-scale needs, needs control system methods

PSS: I regret a bit for dropping Matthew Cook’s class on computational models because I couldn’t see the link back then. It was frustrating because not know the links was making the problem sets boring. If I knew curiosity would bring me back, even in a limited context, I would’ve kept to it.

PSS: I regret a bit for dropping Matthew Cook’s class on computational models because I couldn’t see the link back then. It was frustrating because not know the links was making the problem sets boring. If I knew curiosity would bring me back, even in a limited context, I would’ve kept to it.

Ephaptic conduction molding memory engrams Memories are programmed in the brain as connected neuronal ensembles called engrams. However, the method by which the brain forms engrams during memory encoding is not understood. We have created a me...

PS: There is a sim paper for ephaptically-driven engram formation with cellular automata which makes me happy. It’s a different kind of traction, esp. bc cellular automata has ties to halting problem studies (Busy Beaver challenge) and Matthew Cook’s work.

pmc.ncbi.nlm.nih.gov/articles/PMC...

In vivo ephaptic coupling allows memory network formation Abstract. It is increasingly clear that memories are distributed across multiple brain areas. Such “engram complexes” are important features of memory form

Love the work. Replicating this using the same task (or across hippocampal regions with a spatial learning task) with non-Free Energy/Granger causality method is a community ask.

Is the dataset publicly available?

academic.oup.com/cercor/artic...

Shares in Allbirds surge after maker of wool sneakers announces pivot to AI Rebrand as NewBird AI sent shares up 582% in bizarre and rapid turnaround for firm that had fallen on hard times

They even got approval to ditch the benefit corporation entity to become a conventional for-profit. All for an AI pump-and-dump.

www.theguardian.com/business/202...

I can’t comment on this. You have to find the domain experts or the developing online community. I’m sure things will fork into the same topic in the future.