Yohan J John
@dryohanjohn
🧠 computational neuroscience | neurophenomenology 2.0? 🤖 Science Writer, Kempner Institute for the study of Natural and Artificial Intelligence at Harvard University (views here are my own)
"As AI-augmented research proliferates, the scope of scientific inquiry is contracting." www.ft.com/content/0c63... [paywalled]
As it turns out, it will also remove hyphens indiscriminately. What rough beast, its hour come round at last, slouches...
"One organism might peer out from another organism, but not live there." I'm in awe of VanDerMeer's ability to conjure the uncanny.
I used to generate spiral patterns using center-surround networks just for fun... Never really thought of a use for them.
A very different approach to "chill music to work to". :P #NowPlaying
"Oscar Wilde once defined fox hunting as the unspeakable in pursuit of the uneatable. Were he alive today he might describe the quest for artificial general intelligence as the unfathomable in pursuit of the indefinable."
Good god that Taylor Swift album is atrocious. Perhaps it will put the final nail in the coffin of nihilistic millennial poptimism.
"Signs are produced and circulated at a growing speed but the human terminal of the system (the embodied mind) is put under growing pressure, and finally it cracks." This was written in 2009, so he had no idea that generative AI was just around the corner.
Nice! Just published a closely related essay. yohanjohn.substack.com/p/to-model-b...
Haha. Just last night I added this footnote on "all models are wrong" to my latest substack essay. yohanjohn.substack.com/p/neuroscien...
Today I learned that the manifold hypothesis might be traceable all the way back to Helmholtz. From 'Operationalizing Kant Manifolds, Models, and Mathematics in Helmholtz's Theories of Perception' doi.org/10.7551/mitp... #psychology #neuroscience #philosophy
I'm starting a series of essays that view neuroscience through the lens of Baudrillard's concept of hyperreality. Here's the first part: yohanjohn.substack.com/p/neuroscien... #neuroscience
Picked up this book after hearing about it in a video essay. Maybe this year will be the first time I attend a Ren Faire.
" Both are attempts at post hoc stabilisation of a wholly synthetic cognitive artefact." 💯
Yup. Here's a footnote in Wiktionary. en.wiktionary.org/wiki/granular
Recently came across this book (from 1962!), quite by fluke. It's where the term 'brown sahib' came from, apparently. I've only read a few pages, but it is savage! And cuts a bit close to home! archive.org/details/dli....
Looking into Sterling & Eyer (1988), the paper that coined the term 'allostasis'. They're not mincing words! #neuroscience
Just finished 'Martyr!' by Kaveh Akbar. A wonderfully life-affirming book about death. (Or possibly vice versa.)
This is a cool paper, both in terms of results and in terms of methodology. The distinction between 'fractured entangled' and 'unified factored' reps is handy. 'Questioning Representational Optimism in Deep Learning: The Fractured Entangled Representation Hypothesis' arxiv.org/abs/2505.11581 #ML
"... those who stayed shifted their attentions away from original research and toward hermeneutics: interpreting the scientific work of metahumans." Ted Chiang wrote this in the year 2000 in a sci fi essay in Nature (!?). Reminds me of research on interpretability in ML. 😛
I've always liked the Haber spiral diagrams since they convey something about global organization. doi.org/10.1523/JNEU...
Pet peeve: data/simulation plots with multiple lines in different colors, but the same line style and width. Not everyone can make out different colors! Also, perceptual redundancy is generally a good thing. Is there a review of the psychophysics of data visualization?
This is a subtle way of looking at instruction fine-tuning. "...models perform better when they are trained to distinguish their own language from that of an interlocutor."