Fergus Meiklejohn
@airuyi.bsky.social
Studied philosophy, especially like the thinking of Kant, Wittgenstein and Heidegger. Working on figuring out how to build something that can think for itself
Reposted by Fergus Meiklejohn
Delighted to share our discoveries about one of the brain's neurotransmitter systems:
www.biorxiv.org/content/10.6...
Together with colleagues at the @alleninstitute.org, we have learned a lot about a tiny cluster of neurons in the brainstem locus coeruleus (LC) that releases norepinephrine (NE). 1
www.biorxiv.org/content/10.6...
Together with colleagues at the @alleninstitute.org, we have learned a lot about a tiny cluster of neurons in the brainstem locus coeruleus (LC) that releases norepinephrine (NE). 1
www.biorxiv.org
April 13, 2026 at 4:51 AM
Delighted to share our discoveries about one of the brain's neurotransmitter systems:
www.biorxiv.org/content/10.6...
Together with colleagues at the @alleninstitute.org, we have learned a lot about a tiny cluster of neurons in the brainstem locus coeruleus (LC) that releases norepinephrine (NE). 1
www.biorxiv.org/content/10.6...
Together with colleagues at the @alleninstitute.org, we have learned a lot about a tiny cluster of neurons in the brainstem locus coeruleus (LC) that releases norepinephrine (NE). 1
Reposted by Fergus Meiklejohn
New preprint on our favorite theme: same connectome + different neurons = different dynamics.
@shtakasu.bsky.social looked at detail balance (E+I inputs sum to 0) in RNNs w/different activation functions, finding that it stabilized some networks and destabilized others.
arxiv.org/html/2608.30...
@shtakasu.bsky.social looked at detail balance (E+I inputs sum to 0) in RNNs w/different activation functions, finding that it stabilized some networks and destabilized others.
arxiv.org/html/2608.30...
September 8, 2026 at 1:44 PM
New preprint on our favorite theme: same connectome + different neurons = different dynamics.
@shtakasu.bsky.social looked at detail balance (E+I inputs sum to 0) in RNNs w/different activation functions, finding that it stabilized some networks and destabilized others.
arxiv.org/html/2608.30...
@shtakasu.bsky.social looked at detail balance (E+I inputs sum to 0) in RNNs w/different activation functions, finding that it stabilized some networks and destabilized others.
arxiv.org/html/2608.30...
Reposted by Fergus Meiklejohn
You can't blame Ed Sheeran for caving in to American stadium owners I suppose. I mean, take these guys in the 60s, they refused to play to racially-segregated audiences in stadiums in the southern states of the USA and who remembers them now?
The Beatles banned segregated audiences, contract shows
The Beatles refused to play to segregated audiences in the US, a contract for one of the group's concerts shows.
www.bbc.co.uk
September 15, 2026 at 12:04 PM
You can't blame Ed Sheeran for caving in to American stadium owners I suppose. I mean, take these guys in the 60s, they refused to play to racially-segregated audiences in stadiums in the southern states of the USA and who remembers them now?
Reposted by Fergus Meiklejohn
Jay used our clonal mollies to ask does increasing environmental variation increases phenotypic variation? Turns out.... it doesn't!
Mollies reared with pred cues swim more slowly & hide more but the magnitude of individuality remained the same! It seems individuals are unique no matter what...
Mollies reared with pred cues swim more slowly & hide more but the magnitude of individuality remained the same! It seems individuals are unique no matter what...
🐟🚨 New paper! 🚨🐟
Where do behavioral differences come from? Early life environment clearly matters...
...or not?
Early environment affected behavior, but NOT how different individuals became from each other. Those differences were seeded before birth through factors OTHER than genetic variation 🤯
Where do behavioral differences come from? Early life environment clearly matters...
...or not?
Early environment affected behavior, but NOT how different individuals became from each other. Those differences were seeded before birth through factors OTHER than genetic variation 🤯
Born This Way: Individuality Is Seeded Before Birth and Robust to Ecological Stress
We reared and recorded behaviours of clonal Amazon mollies in standardized environments, with and without predation stress, from birth through week four of life, to understand how the developmental e...
onlinelibrary.wiley.com
September 15, 2026 at 6:59 PM
Jay used our clonal mollies to ask does increasing environmental variation increases phenotypic variation? Turns out.... it doesn't!
Mollies reared with pred cues swim more slowly & hide more but the magnitude of individuality remained the same! It seems individuals are unique no matter what...
Mollies reared with pred cues swim more slowly & hide more but the magnitude of individuality remained the same! It seems individuals are unique no matter what...
The problem is body consequence. How can one neuron in billions, with no possible knowledge of anything except it's local environment, contribute positively to the successful navigation of the entity in the world. It's a cell. It only wants chemicals.
September 15, 2026 at 7:49 AM
The problem is body consequence. How can one neuron in billions, with no possible knowledge of anything except it's local environment, contribute positively to the successful navigation of the entity in the world. It's a cell. It only wants chemicals.
Reposted by Fergus Meiklejohn
In thisa piece for @thenewworldmag.bsky.social I express the wish for a better level of public discourse around AI that doesn't leave us hanging on every word of the alleged tech gurus.
www.thenewworld.co.uk/philip-ball-...
www.thenewworld.co.uk/philip-ball-...
Apocalyptic AI fantasies are a stupid, annoying distraction from its real problems
The people who work in AI are not gurus and most of their predictions have been wrong. Working in the industry makes them less reliable commentators, not more
www.thenewworld.co.uk
September 15, 2026 at 7:35 AM
In thisa piece for @thenewworldmag.bsky.social I express the wish for a better level of public discourse around AI that doesn't leave us hanging on every word of the alleged tech gurus.
www.thenewworld.co.uk/philip-ball-...
www.thenewworld.co.uk/philip-ball-...
Reposted by Fergus Meiklejohn
Cory @doctorow.pluralistic.net has been writing about inflammatory AI predictions for years. Message below is that systems are being deployed & given the ability to do real things on the network, without adequate testing or guardrails. The systems have zero intentions. medium.com/@doctorow/ll...
LLMs are real, AI is fake
No, it didn’t “go rogue.”
medium.com
September 14, 2026 at 9:19 PM
Cory @doctorow.pluralistic.net has been writing about inflammatory AI predictions for years. Message below is that systems are being deployed & given the ability to do real things on the network, without adequate testing or guardrails. The systems have zero intentions. medium.com/@doctorow/ll...
Reposted by Fergus Meiklejohn
Yes, we already have artificial autonomous entities that are not aligned with human values and that we've allowed to get out of control - they're called corporations
I think the gist of warnings like this one is correct: we urgently need binding international regulations before use of AI does serious harm.
I'm also concerned that the debate is being distorted by the language & framings that those in the industry have imbibed.
www.theguardian.com/technology/2...
I'm also concerned that the debate is being distorted by the language & framings that those in the industry have imbibed.
www.theguardian.com/technology/2...
I worked at Google DeepMind. You should listen to the warnings about AI | Alex Turner
We must stop companies from allowing AI to self-improve into an uncontrollable level of intelligence
www.theguardian.com
September 14, 2026 at 2:28 PM
Yes, we already have artificial autonomous entities that are not aligned with human values and that we've allowed to get out of control - they're called corporations
Reposted by Fergus Meiklejohn
Time to repost this:
"On the existential risks of artificial intelligence"
romainbrette.fr/on-the-exist...
"On the existential risks of artificial intelligence"
romainbrette.fr/on-the-exist...
On the existential risks of artificial intelligence
The impressive progresses in machine learning have revived the fear that humans might eventually be wiped out or enslaved by artificial superintelligences. This is hardly a new fear. For example, t…
romainbrette.fr
September 14, 2026 at 3:21 PM
Time to repost this:
"On the existential risks of artificial intelligence"
romainbrette.fr/on-the-exist...
"On the existential risks of artificial intelligence"
romainbrette.fr/on-the-exist...
Reposted by Fergus Meiklejohn
Infantile amnesia is not simply a story about forgetting the past. It is an opportunity to understand how memory itself is built, write @franklandlab.bsky.social, @sjo09.bsky.social, and Nicholas Turk-Browne.
#neuroskyence
www.thetransmitter.org/memory/infan...
#neuroskyence
www.thetransmitter.org/memory/infan...
Infant memories: Lost but not gone?
Infantile amnesia is not simply a story about forgetting the past. It is an opportunity to understand how memory itself is built.
www.thetransmitter.org
September 14, 2026 at 4:12 PM
Infantile amnesia is not simply a story about forgetting the past. It is an opportunity to understand how memory itself is built, write @franklandlab.bsky.social, @sjo09.bsky.social, and Nicholas Turk-Browne.
#neuroskyence
www.thetransmitter.org/memory/infan...
#neuroskyence
www.thetransmitter.org/memory/infan...
Reposted by Fergus Meiklejohn
Approximating the Lorenz attractor with a chaotic leaky water wheel.
The orange dot is the centre of mass of the water. It traces (a projection of) the familiar attractor as the wheel spins seemingly randomly left and right.
The orange dot is the centre of mass of the water. It traces (a projection of) the familiar attractor as the wheel spins seemingly randomly left and right.
September 13, 2026 at 5:38 PM
Approximating the Lorenz attractor with a chaotic leaky water wheel.
The orange dot is the centre of mass of the water. It traces (a projection of) the familiar attractor as the wheel spins seemingly randomly left and right.
The orange dot is the centre of mass of the water. It traces (a projection of) the familiar attractor as the wheel spins seemingly randomly left and right.
Computation is everywhere in our thinking. Here is Michael Levin asserting that living things are computational entities storing and processing information. This is only one possible frame, one pair of glasses. We can take them off and use another one sometimes..
www.drmichaellevin.org/research/
www.drmichaellevin.org/research/
September 14, 2026 at 6:03 AM
Computation is everywhere in our thinking. Here is Michael Levin asserting that living things are computational entities storing and processing information. This is only one possible frame, one pair of glasses. We can take them off and use another one sometimes..
www.drmichaellevin.org/research/
www.drmichaellevin.org/research/
There is an argument for building a brain rather that the brain.
Any brain must be embodied (in some sort of world) because Being-In-The-World is the root of motivation, of all possible brain thought.
Any brain must be embodied (in some sort of world) because Being-In-The-World is the root of motivation, of all possible brain thought.
Talking about blue brain, digital twins and the likes: The danger of building a brain from scratch: trialsanderrors.substack.com/p/the-danger...
The danger of building a brain from scratch
Having a model of the brain that works in many cases seems like a good idea. Is it really ?
trialsanderrors.substack.com
September 14, 2026 at 1:47 AM
There is an argument for building a brain rather that the brain.
Any brain must be embodied (in some sort of world) because Being-In-The-World is the root of motivation, of all possible brain thought.
Any brain must be embodied (in some sort of world) because Being-In-The-World is the root of motivation, of all possible brain thought.
Memory
We don't retrieve data from somewhere in the brain; the actual neurons that fire have both been physically remade with new proteins and also different neuron groups fire for the "same" memory.
What if memory is just a very similar conversation in the brain. Our brains do "the same" thing..?
We don't retrieve data from somewhere in the brain; the actual neurons that fire have both been physically remade with new proteins and also different neuron groups fire for the "same" memory.
What if memory is just a very similar conversation in the brain. Our brains do "the same" thing..?
September 13, 2026 at 8:11 AM
Memory
We don't retrieve data from somewhere in the brain; the actual neurons that fire have both been physically remade with new proteins and also different neuron groups fire for the "same" memory.
What if memory is just a very similar conversation in the brain. Our brains do "the same" thing..?
We don't retrieve data from somewhere in the brain; the actual neurons that fire have both been physically remade with new proteins and also different neuron groups fire for the "same" memory.
What if memory is just a very similar conversation in the brain. Our brains do "the same" thing..?
The challenge with not using an engineering or reductionist approach to building is that it then becomes very difficult to figure out why the thing doesn't work.
If behaviour doesn't demonstrate self-directed thinking is it the neuron, the synapse, the connection, their organisation, something else?
If behaviour doesn't demonstrate self-directed thinking is it the neuron, the synapse, the connection, their organisation, something else?
September 13, 2026 at 6:03 AM
The challenge with not using an engineering or reductionist approach to building is that it then becomes very difficult to figure out why the thing doesn't work.
If behaviour doesn't demonstrate self-directed thinking is it the neuron, the synapse, the connection, their organisation, something else?
If behaviour doesn't demonstrate self-directed thinking is it the neuron, the synapse, the connection, their organisation, something else?
If we say "what is the state of the brain at this instant?" (or map the connectome at that instant) we are implying that the brain is computing states on every tick: state 1 is input -> compute -> state 2 is output and input on next computational tick.
But this computational model is not proven!
But this computational model is not proven!
September 13, 2026 at 4:37 AM
If we say "what is the state of the brain at this instant?" (or map the connectome at that instant) we are implying that the brain is computing states on every tick: state 1 is input -> compute -> state 2 is output and input on next computational tick.
But this computational model is not proven!
But this computational model is not proven!
If we model the brain using computational words we are putting our thinking in a box: looking at the world through a prior frame.
If we call seeing a coffee cup a "process" of "information" leading to a "representation". We are already slanting our thinking towards "the brain is a computer"
If we call seeing a coffee cup a "process" of "information" leading to a "representation". We are already slanting our thinking towards "the brain is a computer"
September 13, 2026 at 4:17 AM
If we model the brain using computational words we are putting our thinking in a box: looking at the world through a prior frame.
If we call seeing a coffee cup a "process" of "information" leading to a "representation". We are already slanting our thinking towards "the brain is a computer"
If we call seeing a coffee cup a "process" of "information" leading to a "representation". We are already slanting our thinking towards "the brain is a computer"
The deeper question is: why are our governments permitting companies to track our movements, decisions and speech?
Do our governments work for us or not?
Do our governments work for us or not?
September 13, 2026 at 4:15 AM
The deeper question is: why are our governments permitting companies to track our movements, decisions and speech?
Do our governments work for us or not?
Do our governments work for us or not?
This is a film I made about Iranian Art more than 20 years ago. Khatami was President and there was freedom, hope and a sense of fun and adventure in the air :-)
Watch it if you'd like a glimpse of Iran in those happier times.
x.com/airuyi/statu...
Watch it if you'd like a glimpse of Iran in those happier times.
x.com/airuyi/statu...
Fergus Meiklejohn (@airuyi) on X
This is a film I made about Iranian Art more than 20 years ago. Khatami was President and there was freedom, hope and a sense of fun and adventure in the air :-)
Watch it if you'd like a glimpse of I...
x.com
September 13, 2026 at 4:13 AM
This is a film I made about Iranian Art more than 20 years ago. Khatami was President and there was freedom, hope and a sense of fun and adventure in the air :-)
Watch it if you'd like a glimpse of Iran in those happier times.
x.com/airuyi/statu...
Watch it if you'd like a glimpse of Iran in those happier times.
x.com/airuyi/statu...
Reposted by Fergus Meiklejohn
Some sense in the "rogue agents" debacle.
Anthropomorphic metaphors of machines are misleading, just like machine metaphors of the brain are misleading.
Anthropomorphic metaphors of machines are misleading, just like machine metaphors of the brain are misleading.
I wrote down my thoughts about the last several weeks of AI hell.
aiguide.substack.com/p/misleading...
aiguide.substack.com/p/misleading...
Misleading Metaphors, Real Risks
What To Fear from AI Agents and How to Reclaim Our Human Agency
aiguide.substack.com
September 11, 2026 at 7:15 AM
Some sense in the "rogue agents" debacle.
Anthropomorphic metaphors of machines are misleading, just like machine metaphors of the brain are misleading.
Anthropomorphic metaphors of machines are misleading, just like machine metaphors of the brain are misleading.