🧵 Neuroscience
Anonymous at Mon, 27 May 2024 19:48:12 UTC No. 16196333
In your opinion, what are the greatest unanswered questions in neuroscience?
Anonymous at Mon, 27 May 2024 19:54:26 UTC No. 16196338
the greastest unanswered question in neuroscience is whether they checked and fixed all the bugs in their software yet: https://www.theguardian.com/science
the second greastest is whether all their yapping will ever allow them to find a working mechanism of the brain.
Anonymous at Mon, 27 May 2024 19:55:40 UTC No. 16196339
>>16196338
fMRI is steaming pile of shit. Electrophysiology is the way
Anonymous at Mon, 27 May 2024 20:01:24 UTC No. 16196347
>>16196339
>Electrophysiology is the way
I doubt that this is anything better than fMRI in not being a deadend for neuroscientists to study the brain. I've been hearing their yapping and promises since the 2000s or something and not a single significant discovery has came out from them. heck, they even let the computer scientists run away with building some cool AI shit with some stupid flawed neuron model from whoever know last century.
Anonymous at Mon, 27 May 2024 20:02:50 UTC No. 16196349
imho, the greatest unanswered question is why are they still doing this pseudoscience bs?
Anonymous at Mon, 27 May 2024 20:04:40 UTC No. 16196355
>>16196347
And what is it that they have promised? And what do you hold significant?
Anonymous at Mon, 27 May 2024 20:05:37 UTC No. 16196357
>>16196333
Why do I want kurisu to sit on my face? Neurologicaly speaking.
Anonymous at Mon, 27 May 2024 20:12:32 UTC No. 16196372
>>16196357
Pleasure hormones get released in your brain when a hot and smart neuroscientist causes you to asphyxiate
Anonymous at Mon, 27 May 2024 20:18:32 UTC No. 16196390
>>16196349
What is true science?
Anonymous at Mon, 27 May 2024 20:19:49 UTC No. 16196396
>>16196355
how the fuck do I remember? it's from the 2000s. I just remember that neuroscience was really big and was the coolest shit back then, so cool that even I wanted to study neuroscience. well, lucky that I discovered compsci instead.
after like 20 years the neurosciences shit fizzled out occasionally I see some neuroscience talks on youtube but it's like they're piggybacking the AI crazes now. even some of the "new" shit like active inference which is being shilled by some old dudes in Oxford is just reinforcement learning with a specific form of reward function (which is of course a big field of study in RL. does it warrant to be called something "new"?)
Anonymous at Mon, 27 May 2024 20:22:07 UTC No. 16196403
>>16196338
>links an ancient news article
>doesn't understand basic research
>seethes
Anonymous at Mon, 27 May 2024 20:24:19 UTC No. 16196409
>>16196403
>seethes
nah, I'm just disappointed with you all. literally nothing has ever came out of neurosciences. I wanted to do some cool shit with the brain or understand the brain when I was younger. nothing ever useful or significant has came out of neuroscience. I'm very very disappointed and a bit bumped.
Anonymous at Mon, 27 May 2024 20:26:51 UTC No. 16196413
hey, shill me something cool that has came out of neuroscience and maybe I will go read it and be less disappointed about iot.
Anonymous at Mon, 27 May 2024 20:32:00 UTC No. 16196423
give me some credit. I spent 10 hours to understand this active inference thing completely and was bumped that it was reinforcement learning with a bit of a layer of philosophy mixed ontop.
Anonymous at Mon, 27 May 2024 20:36:31 UTC No. 16196432
there's a reason that people like Demis Hassabis, Geoffrey Hinton left neuroscience. they realized it was a deadend. Rich Sutton was a psychologists.
Even my heroes left neuroscience, how can you blame me???
Anonymous at Mon, 27 May 2024 20:39:01 UTC No. 16196435
>>16196413
>>16196413
Personally, I find this cool that they have identified which retinal output neurons drive predatory behavior in animals: https://www.cell.com/neuron/fulltex
More pop-sciency stuff is here: https://neurosciencenews.com/top-10
Anonymous at Mon, 27 May 2024 20:40:16 UTC No. 16196439
>>16196435
thanks, I'll check it out.
Anonymous at Mon, 27 May 2024 20:52:07 UTC No. 16196460
>>16196396
This applies to almost all scientific fields, it is not intrinsic to neurosciences. There are only a handful of "bleeding edge" research groups in the world, most groups focus on not-so-interesting but nevertheless important stuff for building the foundations. Then there are occasional group leaders who overpromise and underdeliver, or blatantly lie.
Anonymous at Mon, 27 May 2024 20:56:16 UTC No. 16196469
>>16196460
>This applies to almost all scientific fields, it is not intrinsic to neurosciences
maybe true, but it's also personal for me, cause I really liked neuroscience, almost chose it to be my undergrad. in my country you can't switch undergrad field of study once you inked the paper unless you want to waste 2 more years.
aight, I should stop my yapping on neuroscience, you guys have fun.
Anonymous at Mon, 27 May 2024 20:57:31 UTC No. 16196473
I agree that piggybacking on and building unwarranted hype with AI is lazy. I appreciate papers that use AI as tool to make data analysis more efficient or reveal something that hasn't been seen before with conventional methods. But purely theoretical neuro papers that conflate AI with brains are bad.
Anonymous at Mon, 27 May 2024 21:38:49 UTC No. 16196563
>>16196333
how does consciousness arise from neurons?
Anonymous at Tue, 28 May 2024 05:45:04 UTC No. 16197142
>>16196333
i find behavior and problem-solving the most interesting parts of neuroscience. Going through Kandel right now
Anonymous at Tue, 28 May 2024 08:45:19 UTC No. 16197404
>>16196563
Nobody knows
Anonymous at Tue, 28 May 2024 09:18:45 UTC No. 16197427
How do we cure foot fetishists?
Anonymous at Tue, 28 May 2024 10:44:56 UTC No. 16197493
>>16197427
You don't
Anonymous at Wed, 29 May 2024 04:23:21 UTC No. 16198651
>>16196333
anons any roadmap or guide i can follow to understand computatinal neuroscience? What is the greatest hurdle to simulating a human brain in a computer and what prerequisites are there to understanding the comp scie behind it like cognitive architectures ? I'd be really appreciative of any anon that replied
Anonymous at Wed, 29 May 2024 10:00:06 UTC No. 16198940
>>16198651
For computational neuroscience, you need to understand biology, some essential math (linear algebra and calculus), statistics, essential physics (classical mechanics and electromagnetism), and some EE topics such as signal analysis and signal processing. I will make a roadmap image later.
Anonymous at Wed, 29 May 2024 16:38:06 UTC No. 16199273
>>16196347
Nothing dramatic, but substantial progress has been made in identifying neurological correlates for said models. https://pubmed.ncbi.nlm.nih.gov/332
Pretty much just affirms intuitions on hierarchical processing. The next step is to formalize intermediary processes like 'attention'. Which probably requires more nuanced imaging. Picrel- was just some schizophrenia shit I found.
Anonymous at Thu, 30 May 2024 07:49:15 UTC No. 16200433
>>16199273
Modeling is cool and helps with making new hypotheses. I don't know why theoretical work gets so much hate
Anonymous at Thu, 30 May 2024 20:59:23 UTC No. 16201375
>>16196396
>thread ruined by a crying cs fag
get out of sci, compsci is not science
Anonymous at Thu, 30 May 2024 21:04:24 UTC No. 16201384
How to become happy.
Anonymous at Fri, 31 May 2024 05:04:03 UTC No. 16202036
>>16201384
Find a meaning in your life. Maybe a skill to learn that takes years to master. Martial arts, archery, smithing, painting, some academic topic, survival skills, anything. Or devote yourself to some cause.
There was essentially no depression in the medieval times when people had meaningful jobs that took decades to master. That gave life meaning, and made people happy.
Anonymous at Sat, 1 Jun 2024 05:31:53 UTC No. 16203663
>>16198651
Sci wiki recommends Dayan & Abbot