Hot new paper today from UCSF Neurosurgery with additional info about the function of Wernicke’s area.
Might be worth reading by @Fionnindy , who unfortunately seems to disallow replies to his posts. What a pity.
But I do maintain that biological functioning is likely pretty reducible as science progresses; a definition of consciousness that demands a permanent inability to “read” another’s mind seems unlikely to last the test of time, given BCI and/or techniques like OPM-MEG.
xitter.gq/fionnindy/status/21235…
Each word you say is planned by a small patch of brain above your left ear, about a quarter of a second before you say it. It sits in a temporal lobe region textbooks describe as only for listening. Our paper today in @ScienceMagazine: science.org/doi/10.1126/scie…
@UCSF @NeurosurgUCSF
1
143
Ohhh. I had only read Nagel's WIRED interview but not the underlying bat paper. Great reminder from Prof de Bruijn below to read the paper!
I've only skimmed a few pages and I think I see why this is all going so strangely for the tech crowd.
Nagel doesn't define "what is consciousness." I'd wager that is the point of the paper, on how to assess the claims and experiences of other actors besides oneself, but ultimately this whole thing is pretty curious because... without a definition, tech folks can't prove anything from first principles.
Consciousness for a folk definition might be something like awake, aware, and experiencing. That comes from a human tradition of decades-to-centuries where the only entities that we believe are awake, aware, and experiencing are also alive (meaning: biological). [I'm unclear if religious scholars claim entities like angels or deities are conscious?]
We have no historical conception of non-biological life, but we certainly do see early signs of "awake, aware, and experiencing" characteristics of LLMs, virtual fruit-fly connectome simulators, etc.
So the next step doesn't seem to just stipulate that consciousness can only be found in biological life; it seems reasonable to ask what is consciousness, at what point does the simulation of consciousness become functionally equivalent with biological consciousness, and does non-biological consciousness have different characteristics and moral values ascribed to it compared to biological consciousness (coming from us, and coming from them).
Where this goes I believe is the need for philosophy papers to develop formalization just like mathematics (which is ironic, since mathematics gets its formalization from philosophy, I think). Every term needs to be defined -- in this case, Nagel doesn't define consciousness in the bat paper, as far as I can tell -- and therefore we can't really figure out if octopus are conscious or not (given their huge neurons and pain avoidance, etc etc), nor whether are datacenter-powered Figure robots with superhuman intelligence and pain avoidance, etc etc.
I really do need to read the underlying paper more closely though; I only skimmed it.
Replying to @JakubTomek13
i don’t know if bacteria are conscious—- nagel’s claim is that consciousness is a subjective fact, ie a fact from a pov, and not assessible in the way one does third-personal facts. There’s something it’s like *to be* certain things, a bat eg, and that is consciousness. Read the bat paper it’s really very good.
Here's what Gemini said about what I wrote: "The passage works well as brainstorming and captures a real tension in AI ethics: How do we build moral frameworks for entities that act conscious without knowing if they actually are? However, as a logical argument, it is unsound. It demands a formal, third-person mathematical definition for a phenomenon (subjective experience) whose defining feature, according to the very paper being cited, is that it resists third-person reduction. Furthermore, it treats a stipulative definition as a magic key that can unlock an empirical/metaphysical mystery, while smugging functionalism into its premises under the guise of neutral inquiry."
1
163
(My earlier thoughts on this, below)
xitter.gq/christopherbaer/status…
I had read the WIRED article this morning too, and as a “philosophizing non-philosopher” (to quote Edmund Pellegrino) and without prior exposure to Nagel, the argument Nagel made didn’t seem very convincing to me. It basically stipulated itself out of biological exceptionalism?
I agree a TI-83 calculator is not conscious, both because it doesn’t write out that it is conscious, AND because we can monitor its computations and at no point does it do anything related to topics of consciousness.
When we put an exact representation of a fruit fly central nervous system into a connectome simulator (Google/HHMI), and then it can fly around in a virtual box with fruit-fly-like behavior — and when we can monitor certain neurons for behavior like pain, or hunger — I become uncertain whether simulated consciousness is possible or not, and furthermore whether that is meaningful, or not.
That doesn’t mean simulated consciousness is necessarily the same thing as traditional consciousness. But neither does it preclude that perhaps some, or even all, properties of consciousness may be shared in practice between simulated and “original, biological” consciousnesses; and this may be the case potentially only sometimes, and potentially not always, based variably on programming, hardware, or situations.
Put differently:
A calculator does not tell us it is afraid, so we also don’t bother to look internally at whether it is afraid.
Nor is there value in thinking about whether a static transcript (like a tape recording or a book) is conscious.
Of course we know that a robot that tells us it is afraid may simply be mimicking what it thinks it should say based on its training data that is largely the record of humans.
And of course a human that tells us it is afraid is, by all measures, afraid.
But I can’t really tell the difference between the last three categories (tape recording of a human that is saying it is afraid, robot saying it is afraid, human saying it is afraid) from only listening to what was said; from the evidence I see of outputs, they all look the same, so what seems to matter is the process that generates the output.
Coincidentally, we can also look inside of an actor at how it generated that output. In humans, that’s fMRI or EEG or OPM-MEG, and we can also observe some differences across humans with different brain lesions or from transcranial magnetic stimulation…
But in robots, we can also ablate or boost certain neurons and cause different thinking traces (output streams, and also what Anthropic saw as a kind of working memory ‘Jacobian space’), which then show more or less (as we desire) mimicry of what conscious biological entities sound like or do. We can flip behaviors on and off in the fruit fly connectome simulator, or reprogram weights to do wacky things such that the fruit fly is able to pattern match to react accurately for video games.
So things are getting a little confusing when looking at the process of how outputs are made, no? We can build increasingly sophisticated, non-biological systems with physical AI (robot) forms, and someday in the future, we may see “more profound depth of activity” in the AI version than the biological version — meaning that depth of feeling, quantity of feeling may not be strictly a biological-defendable claim.
Not sure where that leaves us, but I didn’t think Nagel engaged with this much in the article from a computational, evolutionary-based perspective?
Going forward, there are definitely technical design choices and moral/ethical values we can choose to hold or make/change. But I am not so sure we should predicate everything on merely stipulating only biology is capable of consciousness (which is how I am reading Nagel), when it isn’t clear whether a particular AI in the future is going to be a tool or a robot.
Note: I think we will see a wide variety where some AI seem to possess emergent virtual consciousness properties while other AI don’t. 😂
1
204
I had read the WIRED article this morning too, and as a “philosophizing non-philosopher” (to quote Edmund Pellegrino) and without prior exposure to Nagel, the argument Nagel made didn’t seem very convincing to me. It basically stipulated itself out of biological exceptionalism?
I agree a TI-83 calculator is not conscious, both because it doesn’t write out that it is conscious, AND because we can monitor its computations and at no point does it do anything related to topics of consciousness.
When we put an exact representation of a fruit fly central nervous system into a connectome simulator (Google/HHMI), and then it can fly around in a virtual box with fruit-fly-like behavior — and when we can monitor certain neurons for behavior like pain, or hunger — I become uncertain whether simulated consciousness is possible or not, and furthermore whether that is meaningful, or not.
That doesn’t mean simulated consciousness is necessarily the same thing as traditional consciousness. But neither does it preclude that perhaps some, or even all, properties of consciousness may be shared in practice between simulated and “original, biological” consciousnesses; and this may be the case potentially only sometimes, and potentially not always, based variably on programming, hardware, or situations.
Put differently:
A calculator does not tell us it is afraid, so we also don’t bother to look internally at whether it is afraid.
Nor is there value in thinking about whether a static transcript (like a tape recording or a book) is conscious.
Of course we know that a robot that tells us it is afraid may simply be mimicking what it thinks it should say based on its training data that is largely the record of humans.
And of course a human that tells us it is afraid is, by all measures, afraid.
But I can’t really tell the difference between the last three categories (tape recording of a human that is saying it is afraid, robot saying it is afraid, human saying it is afraid) from only listening to what was said; from the evidence I see of outputs, they all look the same, so what seems to matter is the process that generates the output.
Coincidentally, we can also look inside of an actor at how it generated that output. In humans, that’s fMRI or EEG or OPM-MEG, and we can also observe some differences across humans with different brain lesions or from transcranial magnetic stimulation…
But in robots, we can also ablate or boost certain neurons and cause different thinking traces (output streams, and also what Anthropic saw as a kind of working memory ‘Jacobian space’), which then show more or less (as we desire) mimicry of what conscious biological entities sound like or do. We can flip behaviors on and off in the fruit fly connectome simulator, or reprogram weights to do wacky things such that the fruit fly is able to pattern match to react accurately for video games.
So things are getting a little confusing when looking at the process of how outputs are made, no? We can build increasingly sophisticated, non-biological systems with physical AI (robot) forms, and someday in the future, we may see “more profound depth of activity” in the AI version than the biological version — meaning that depth of feeling, quantity of feeling may not be strictly a biological-defendable claim.
Not sure where that leaves us, but I didn’t think Nagel engaged with this much in the article from a computational, evolutionary-based perspective?
Going forward, there are definitely technical design choices and moral/ethical values we can choose to hold or make/change. But I am not so sure we should predicate everything on merely stipulating only biology is capable of consciousness (which is how I am reading Nagel), when it isn’t clear whether a particular AI in the future is going to be a tool or a robot.
Note: I think we will see a wide variety where some AI seem to possess emergent virtual consciousness properties while other AI don’t. 😂
Kind of in the same direction…
I think we now have to accept that the crazy AI people (I use the term affectionately) are, at least, directionally correct.
- There is at least a small chance that recursive self-improvement (RSI) in the next couple of years leads quickly to “superintelligence”, which in turn leads to extremely rapid transformative effects, including a possibility of total disaster.
- Even if, like me, you are very sceptical of the most aggressive forecasts, they are directionally correct. You can quibble with the timelines of RSI, with the precise meaning of “superintelligence”, and with the pace of transformation and diffusion, but the basic story is right.
Short of political interventions, we will build systems with capabilities superior to ours across all domains within the next couple of decades. The limitations of current systems look solvable, progress has been shockingly fast, and the resources being thrown at AI are enormous. This process will likely be greatly accelerated once AI systems outcompete us at designing better AI systems, and it will lead to a speed and scale of transformation, as well as novel catastrophic risks, that almost no mainstream elites or normies are pricing into their models of the world.
That, at any rate, is what I now believe with a high degree of confidence.
1
76
Preprint from 11/16/2026
A meta-analysis of 912 neuroimaging studies finds no single core brain region or network underlying the self, challenging the prevailing view of a shared cortical midline or insular anchor. Instead, self-related processing emerges from contextually determined, shifting coalitions of distributed brain systems, suggesting that selfhood is defined by dynamic configuration rather than a fixed neural substrate.
biorxiv.org/content/10.64898…
1
57
This is a very astute summary (from the engineer who coined “open source” I believe) of the central issues in clinical medicine and public health right now: how to ration care, how to think about health equity/justice, what is socially vs individually determined.
So I recently posted on here that one of the reasons I'm opposed to an AI pause is that I don't want to die.
An entire thundering horde of blithering idiots immediately started screaming at me that I'm a selfish bastard who wants to burn the world down.
These people seem to assume I have some kind of supervillain bunker I plan to retreat into, cackling madly and clutching my extended lifespan to my chest, rather than having to live in the same future world as everybody else.
This is a deranged response. I had to think for a while to even begin to mentally model the kind of bugfuck craziness that could produce such thinking.
Now that I think I have done so, I have a public service announcement to make:
Some people will try to convince you that life extension, or wealth in general, is a zero-sum game. Then in order for some to win, others must lose. For any individual person to achieve an extremely long lifetime, the world must be plunged into misery for others.
These people are not your friends. They are not saying these things to help you or save you. They are telling lies which, if you believe them, will control you. And control is their goal.
Did we learn nothing from the last century? The self-anointed scream that a vast catastrophe is heading towards us all, unless we surrender our wealth and liberty to them. The label on the catastrophe changes: overpopulation, pollution, global warming, rogue AIs. The prescription never changes. It's always a grift, always a lie, always a power-grab.
Don't get fooled again.
24
Just downloaded this 499 page book on robophilosophy and it has a gripping start:
"Whether we acknowledge it or not, we are, in fact, in the midst of a robot invasion. The machines are everywhere and doing everything. They may have begun by displacing workers on the factory floor, but they now actively participate in many aspects of intellectual, social, and cultural life."
!!
"Robophilosophy: Philosophy of, for, and by Social Robots" @mitpress is now out and available. This book has been ten years in the making and includes ground-breaking research from the @Robophilosophy_ Conferences. Get it open access at:
direct.mit.edu/books/oa-edit…
Very neat. Strange, but neat.
in case you were curious: fable 5.2 and I confirmed that if you record yourself singing and whistling every note in your range, you can play basically any song using only your voice
no song is playing in this video. every sound is one of 64 clips of me holding a single note
1
25
CI/CD efficiency is definitely one of the things that is always worth checking on, often weekly, for vibe coders like myself. I was up to 10k minutes/month in GitHub Actions usage at one point, since LLMs tend to not optimize much here unless directly asked to do so.
In this case, Dan reduced wall-clock duration by 65% with a one line fix:
“ •apps/e2e/playwright.config.ts: run 4 Playwright workers on CI instead of 1 (local runs keep Playwright's default)…”
What a disaster. I feel like GitHub should be helping users with this type of stuff. The entire “self-hosted” GHA ecosystem shouldn’t need to exist except for big biz customers. 😵💫
only one person (@foundmalek) replied with their OSS repo, so I took a look:
a one line change reduced their CI time by 65%!
PR'd and merged 🚀
you can just do things
1
46
Now that’s a good story.
Replying to @mdonnino
Dr. Raymond Adams (considered father of neurology) & colleagues did large case series putting forth link between thiamine & WE in 1950/60s though clinical trials were less rigorous at the time. Shortly before he died, I met him (age 96) & he shared a bit of a bombshell with me 2/
27
Thanks for explaining, and stepping back for a second, this is shockingly bad — nobody should ever use bitly short URLs if they are apparently long term compromises. What was bitly thinking?!
Ok time to ruin it for everyone by telling people how this work.
Old bitly links no longer work, just go back and look for a big account post with a dead bitly link.
Bitly allow custom urls if it's not taken already.
You register it and now the post links to your stuff and you get free internet points.
1
1
122
Update
Earlier today, we identified an issue that allowed bad actors to repurpose certain old, inactive links to redirect people to content unrelated to the original posts. We have suspended the accounts involved and are disabling the affected links. Our Trust and Safety team is working to address the issue and strengthen protections against similar misuse in the future.
Readers added context they thought people might want to know
Bitly documentation states that bit.ly back-halves cannot be re-used even after account deletion, yet a similar vulnerability allowed repurposing of old links as documented in a February 2025 case Bitly claimed to have fixed.
support.bitly.com/hc/en-us/artic…
disruptionist.com/p/crypto-scamm…
ransomnews.com/burger-king-tw…
19
Harvard prof with a very good review of all the different terms and concepts related to aging… precision of definition here helps a lot IMHO.
Our new paper, “Simply Aging,” is out in @PNASNews!
What is aging, when does it begin, how do we measure it, and what would it take to extend lifespan? We tackle these questions directly and simply.
📄 Read: pnas.org/doi/10.1073/pnas.26…
💬 Discuss: simplyaging.xyz/
20
Quite a result; we really still don't know basic facts about what the brain does (in this case, mouse brain functional neuroanatomy).
I still can’t believe it. My PhD project of 8+ years at @HillmanLab, @ZuckermanBrain and @ColumbiaBME was published today in @NatureComms! We present a newly discovered body map in the secondary motor cortex (M2) of the mouse brain.
25
Cool new preprint detailing Univ of Michigan brain model they trained on 150 TB of brain scans, if I understood correctly
arxiv.org/html/2611.15094v1
Open weights model I believe!
A new brain imaging benchmark was just released on @huggingface 🩻 🧠 NeuroBench is a 2.74 TB dataset of MRI and CT studies to test how well vision-language models actually read brain scans.
This kind of benchmark is so important as medical imaging data is often private. With that public, anyone can check how models really perform.
huge kudos to @mlins_lab for this release 🚀
🤗 dataset: hf.co/datasets/mlinslab/neur…
📜 paper: hf.co/papers/2611.15094
1
1
48
Might be worth trying to do some brain surgery on the brain model:
Want to try weight-space programming? Today we're releasing dspy-wsp, a prototype weight-space optimizer that works like any other @DSPyOSS optimizer.
Using dspy-wsp, and about ~$5-10 of Astra, we improved every decision model we tested. cmpnd.ai/blog/weight-space-p…
1
24






































