i am once again bewildered by how people are “vibe coding”. obviously if you don’t understand the code, you will not be able to maintain it???
this is somewhat independent of LLM assistance. code that the LLM generates and you read and understand is not necessarily any more unmaintainable than “someone else’s code” or “code you copy-pasted from something you looked up online”.
https://blog.jacobtrefethen.com/ai-optimism/ Jacob Trefethen’s bull case on AI for medicine; not a lot that’s new to me there (except for BDNF being unstable and bad at crossing the blood brain barrier; TIL)
https://www.rationalistjudaism.com/p/thats-bats Orthodox rabbi debunks truly bizarre theological hypotheses. no, when the Talmud refers to an egg-laying bat that nurses its young, they were not referring to the platypus (which they could not have known about), or a nursing bird (which does not exist). they simply lived over a thousand years ago and thus knew less about zoology than we do today.
there is a consistent strain of Jewish theology that views the Sages as divinely inspired in matters of law, ethics, and spirituality but not necessarily in natural sciences; from this POV it’s perfectly fine to admit they were factually incorrect about bats.
https://www.neurology.org/doi/abs/10.1212/wnl.0000000000008163 disorders of consciousness show anatomical connectivity disruption on the “ascending arousal network” that links the brainstem, hypothalamus, thalamus, and basal forebrain, in particular brainstem-to-hypothalamus and brainstem-to-thalamus connectivity.
the thalamus has fast tonic spiking during awake states and switches to sleep spindles and then slow waves during slow-wave sleep, “entraining” the rest of the brain into sleep EEG patterns. you can recreate this effect optogenetically.
central thalamic DBS at 50 Hz reverses the effects of general anaesthesia in macaque monkeys, while 10 Hz or 200 Hz produces absence seizures
lousy results in humans for thalamic stimulation to rouse ppl from disorders of consciousness
https://jnnp.bmj.com/content/74/11/1571.short patients who recover from a persistent vegetative state with L-Dopa have parkinson-like symptoms & “diffuse axonal injury involving the substantia nigra or ventral tegmental area.” ⇒ if it looks like parkinsonism and it involves damage to the same brain regions, try treating it like you’d treat Parkinson’s
https://austinvernon.substack.com/cp/168201770 Austin Vernon thinks AI could reshape logistics; first, by automating trucks and cars; then, by allowing more efficient freight transport via a shift to lots of smaller vehicles rather than a few large trucks. cars have better miles per gallon and can go everywhere. last mile costs are 70-80% of shipping costs; automation helps here.
I was slightly too young to be dating during the early 2000s, and I was leading a very sheltered life, so I didn’t quite grasp that 2010′s feminism was a reaction to a pretty raunchy & sexist ambient culture.
links 8/11/25: https://roamresearch.com/#/app/srcpublic/page/08-11-2025
https://www.nationalgeographic.com/travel/article/how-to-spend-a-day-accra-ghana-arty-capital
https://blog.val.town/vibe-code
i am once again bewildered by how people are “vibe coding”. obviously if you don’t understand the code, you will not be able to maintain it???
this is somewhat independent of LLM assistance. code that the LLM generates and you read and understand is not necessarily any more unmaintainable than “someone else’s code” or “code you copy-pasted from something you looked up online”.
https://blog.jacobtrefethen.com/ai-optimism/ Jacob Trefethen’s bull case on AI for medicine; not a lot that’s new to me there (except for BDNF being unstable and bad at crossing the blood brain barrier; TIL)
https://www.cbc.ca/radio/asithappens/ice-mummy-tattooos-1.7601132 Siberian ice mummy’s tattoos revealed; they’re cool!
https://www.rationalistjudaism.com/p/thats-bats Orthodox rabbi debunks truly bizarre theological hypotheses. no, when the Talmud refers to an egg-laying bat that nurses its young, they were not referring to the platypus (which they could not have known about), or a nursing bird (which does not exist). they simply lived over a thousand years ago and thus knew less about zoology than we do today.
there is a consistent strain of Jewish theology that views the Sages as divinely inspired in matters of law, ethics, and spirituality but not necessarily in natural sciences; from this POV it’s perfectly fine to admit they were factually incorrect about bats.
consciousness and the thalamus:
https://www.nature.com/articles/s41467-022-32840-w the thalamus controls deep slow oscillations during non-REM sleep
https://www.nature.com/articles/s41467-022-29852-x you can induce deep sleep in mice by knocking down a particular GABA-A receptor in certain neurons in the reticular nucleus of the thalamus
https://www.jneurosci.org/content/jneuro/37/47/11441.full.pdf the ventral posteromedial nucleus of the thalamus has sleep spindles during focal seizures; meanwhile the rest of the brain has slow sleep-like EEG waves
https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1440-1819.1956.tb01086.x low-frequency low-voltage stimulation to the dog thalamus puts the dog to sleep
https://journals.physiology.org/doi/abs/10.1152/ajplegacy.1951.168.1.260 low-frequency stimulation to the cat thalamus also puts the cat to sleep
https://www.cell.com/current-biology/fulltext/S0960-9822(23)00067-2 certain neurons in the thalamus are active during sound-induced waking from sleep and artificial stimulation of those neurons has the same effect
there’s a brainstem-to-thalamus-to-cortex anatomical pathway here involved in controlling wake/sleep
https://onlinelibrary.wiley.com/doi/abs/10.1111/epi.13045 DBS to the thalamus wakes up epileptics
you can also induce sleep with slow pulsed stimulation to other parts of the brain;
https://www.sciencedirect.com/science/article/abs/pii/001448866290081X basal forebrain
https://www.sciencedirect.com/science/article/abs/pii/0014488663900098 cerebellum, hippocampus, amygdala, visual cortex, and motor cortex
https://link.springer.com/article/10.1007/s00540-014-1889-4 propofol anaesthesia reduces blood flow in multiple brain regions but especially the thalamus
https://academic.oup.com/bja/article-abstract/106/4/548/231329 anaesthesia reduces blood flow to the thalamus and precuneus, which returns during awakening
https://academic.oup.com/bja/article-abstract/115/suppl_1/i27/234261?redirectedFrom=PDF EEGs during anaesthesia are slow-wave, like deep sleep
https://academic.oup.com/bja/article/66/4/490/281824 various sites of reduced brain metabolism during anaesthesia
https://www.nejm.org/doi/full/10.1056/NEJM199405263302101 Karen Ann Quinlan, the famous legal case of a “persistent vegetative state”, had damage primarily to the thalamus, not the brainstem or cortex
https://www.cell.com/trends/cognitive-sciences/abstract/S1364-6613(05)00299-8 the thalamus is impaired in the vegetative state, but some parts of cortex are still active. “primary” sensory cortex regions still respond to sensory stimuli, while “secondary” (higher-order pattern-recognition) regions are offline.
https://academic.oup.com/brain/article-abstract/123/7/1327/380151?redirectedFrom=PDF the thalamus is almost always (80%) abnormal in the vegetative state. subcortical white matter also usually damaged.
https://orbi.uliege.be/bitstream/2268/162980/1/Giacino_et_al_DoC_State_of_the_Science_Nat_Rev_Neurol_2013_epub.pdf overview article of disorders of consciousness
https://pmc.ncbi.nlm.nih.gov/articles/PMC1738081/pdf/v073p00355.pdf more on the vegetative state
https://www.neurology.org/doi/abs/10.1212/wnl.0000000000008163 disorders of consciousness show anatomical connectivity disruption on the “ascending arousal network” that links the brainstem, hypothalamus, thalamus, and basal forebrain, in particular brainstem-to-hypothalamus and brainstem-to-thalamus connectivity.
https://www.cell.com/neuron/pdf/S0896-6273(24)00280-0.pdf
the thalamus has fast tonic spiking during awake states and switches to sleep spindles and then slow waves during slow-wave sleep, “entraining” the rest of the brain into sleep EEG patterns. you can recreate this effect optogenetically.
central thalamic DBS at 50 Hz reverses the effects of general anaesthesia in macaque monkeys, while 10 Hz or 200 Hz produces absence seizures
lousy results in humans for thalamic stimulation to rouse ppl from disorders of consciousness
https://www.brainstimjrnl.com/article/S1935-861X(21)00009-7/pdf ultrasound? not really
https://thejns.org/view/journals/j-neurosurg/aop/article-10.3171-2025.3.JNS241092/article-10.3171-2025.3.JNS241092.xml deep brain stimulation? also no
https://www.epilepsybehavior.com/article/S1525-5050(13)00480-0/pdf the vegetative state is awake but unconscious; usually involves damage to thalamus but not brainstem
https://en.wikipedia.org/wiki/Vegetative_state
https://www.nature.com/articles/s41562-025-02121-9 general anaesthesia makes individuals’ and species’ fMRI results look more similar, in particular in the default mode networks
https://jnnp.bmj.com/content/74/11/1571.short patients who recover from a persistent vegetative state with L-Dopa have parkinson-like symptoms & “diffuse axonal injury involving the substantia nigra or ventral tegmental area.” ⇒ if it looks like parkinsonism and it involves damage to the same brain regions, try treating it like you’d treat Parkinson’s
https://austinvernon.substack.com/cp/168201770 Austin Vernon thinks AI could reshape logistics; first, by automating trucks and cars; then, by allowing more efficient freight transport via a shift to lots of smaller vehicles rather than a few large trucks. cars have better miles per gallon and can go everywhere. last mile costs are 70-80% of shipping costs; automation helps here.
https://andrewrettek.substack.com/p/personal-trainers Andrew Rettek on personal trainers
https://www.historicaltechtree.com/ history
https://rakhim.exotext.com/ai-is-impressive-because-we-ve-failed-at-semantic-web-and-personal-computing yet another anti-AI take that doesn’t actually mean anything bad about AI. a lot of what it does is, indeed, what people wanted the semantic web to do. but as i understand it, that would have been a massive coordination problem, and sidestepping coordination problems with tech is what is known as winning.
https://www.redhat.com/en/topics/api/what-is-a-rest-api
https://aws.amazon.com/what-is/restful-api/ this is AWS’s explanation of a REST api, which is much worse.
for a company where “writing” is allegedly a core value, why is Amazon documentation always the most impenetrable?
https://www.humaninvariant.com/blog/working thoughts on high-paid, low-effort jobs
https://thespinoff.co.nz/internet/29-08-2023/i-coined-dick-is-abundant-and-low-value-it-still-haunts-me-10-years-on
I was slightly too young to be dating during the early 2000s, and I was leading a very sheltered life, so I didn’t quite grasp that 2010′s feminism was a reaction to a pretty raunchy & sexist ambient culture.