Awesome, but it's such a shame that US VCs are too busy putting $250B into data centers so we can have AI slop, when just 1/10000th of that, a rounding error, would have allowed a trial like this in the US.
And the funding system assures we can never do this within a university
And the funding system assures we can never do this within a university
In three patients with refractory epilepsy, hippocampal chemogenetic therapy plus low-dose clozapine was associated with a mean seizure-frequency reduction of 92.6%. Adverse events were grade 1 or 2. Full study results: nej.md/4Agtyhz
2
9
36
4,866
Chemogenetics has obvious advantages in achieving on-demand neuromodulation, but VCs here don't like local gene therapy as it requires people knowing what they are doing, not just pushing pills (it's not quickly "scaleable").
1
0
2
216
But DBS and stents require even more expertise, and those are profitable businesses. So there's a business opportunity for someone with a long timeline.
Unfortunately high interest rates and fast-flip approach toward "investing" in the last 2 years has made those people rare.
Unfortunately high interest rates and fast-flip approach toward "investing" in the last 2 years has made those people rare.
0
0
2
197
Prof. Michael Lin retweeted
In three patients with refractory epilepsy, hippocampal chemogenetic therapy plus low-dose clozapine was associated with a mean seizure-frequency reduction of 92.6%. Adverse events were grade 1 or 2. Full study results: nej.md/4Agtyhz
3
47
149
42,481
Prof. Michael Lin retweeted
This is such a bad result.
California voters appear to have approved Proposition 38, which was authored by billionaire Gary K. Michelson and will fund his immunology research institute to the tune of $4.2 billion. But the institute had only 3 employees as of September. @mzinshteyn calmatters.org/politics/elect…
14
5
145
11,914
Prof. Michael Lin retweeted
Scientists pay tribute to Marta Shahbazi Alonso and Tamara González Costa who died on their way to a developmental-biology meeting
go.nature.com/4iTyNO1
go.nature.com/4iTyNO1
44
1,098
3,357
172,288
Cool work from former postdoc Jihye Seong's lab (at Seoul National University College of Medicine)
mechanoLight: A genetically encoded fluorescent biosensor for visualizing 5-HT2AR mechanotransduction in live cells | PNAS pnas.org/doi/10.1073/pn…
0
0
7
1,100
Prof. Michael Lin retweeted
I am proud of the clear, logically consistent writing we pore over by hand in the lab. AI still can't quite automate that, at least not yet. The problem is… I don't think logical consistency is valued by humans as much anymore anyway.
The problem of AI for academic writing, and why you don't notice it in your own AI writing:
AI loves to use vague language. It will produce something that’s consistent with what you want it to say. The problem is that it will also be consistent with other interpretations that you didn’t intend. This confuses readers: which interpretation was meant?
But you won’t notice the problem when you proofread your AI’s output. That's because you’re checking whether the output is consistent with what you meant (which it will be), without thinking about how it could be misunderstood (which we're usually unskilled at, due to the curse of knowledge)
1
5
44
6,727
Prof. Michael Lin retweeted
Public preview of BioEmu2! @MSFTResearch AI for Science’s new foundation model for biomolecular structure, dynamics and function. Open weights, MIT license.
github.com/microsoft/bioe…
Movie: BioEmu2-simulated binding of intrinsically disordered peptide PUMA to MCL-1 receptor. 🧵⬇️
github.com/microsoft/bioe…
Movie: BioEmu2-simulated binding of intrinsically disordered peptide PUMA to MCL-1 receptor. 🧵⬇️
24
246
912
73,510
The study suggests young researchers make more disruptive discoveries, and older ones fewer but more lasting ones.
Perhaps experience, time, and awareness of long-term bottlenecks might be advantages for older researchers to make the more difficult and lasting breakthroughs.
Perhaps experience, time, and awareness of long-term bottlenecks might be advantages for older researchers to make the more difficult and lasting breakthroughs.
Experienced scientists produce ‘disruptive’ work less often than early-career researchers, but when they do, their innovations are more likely to have a lasting effect
go.nature.com/4sb85DD
1
1
14
1,708
Another possibility: postdocs and young PIs develop and train on the latest methods (RNAi, CRISPR, now AI), and these provide finding that can overturn earlier conclusions, so we get large bursts of disruptive work by younger researchers, but these aren't grand paradigm shifts
1
0
2
392
BTW the definition of disruptive research here is a paper that is cited out of proportion to the papers it cites, i.e. has a high cited/citing ratio. Just working on technology would do that. When you reference a new tech, there’s usually no need to cite what happened before
0
0
0
171
Prof. Michael Lin retweeted
I’m so sick of reviewers using ChatGPT to review papers. It’s atrocious having to respond to the same garbage from all 3 reviewers where each comment requests 5 new analyses. Just don’t agree to review papers if it’s too much work 🤷♂️
Even if you used AI to review my paper- at least ensure the asks are genuine and then choose the most pertinent/relevant comments to include in your review? Is that too much work?
I’m sure journal editors recognize that this is a real problem but can’t do much to rectify this issue.
Even if you used AI to review my paper- at least ensure the asks are genuine and then choose the most pertinent/relevant comments to include in your review? Is that too much work?
I’m sure journal editors recognize that this is a real problem but can’t do much to rectify this issue.
22
28
226
30,576
Prof. Michael Lin retweeted
Interesting. If I understood this correctly, TIME once again turns out to be the critical measurement axis that unlocks this mystery.
1/ We discovered something totally unexpected about the neural structure of consciousness that overturns decades of thinking on its biological basis and brings us significantly closer to the recipe for generating conscious awareness.
New preprint: biorxiv.org/content/10.648… (led by @HesseJanis, with @FrankLanfranchi, @_conor_mcgrory, @Yuelin_Shi, and @EngelTatiana)
2
4
40
7,811
Prof. Michael Lin retweeted
Want to get clean calcium signals from tons of deep-brain neurons in freely moving animals?
Check our paper science.org/doi/10.1126/sc… !
We show how correcting GRIN lens aberrations transforms MINI2P performance, revealing sparse HD coding in sensory thalamus!
@ScienceAdvances
Check our paper science.org/doi/10.1126/sc… !
We show how correcting GRIN lens aberrations transforms MINI2P performance, revealing sparse HD coding in sensory thalamus!
@ScienceAdvances
1
18
66
5,849
Prof. Michael Lin retweeted
Autism cannot be reduced to a simple excitation/inhibition imbalance. Inhibition is not one dial: PV, SST and VIP interneurons regulate different compartments, timescales and network states. A circuit can appear globally “balanced” yet remain computationally disorganized.
3
16
60
1,891
Prof. Michael Lin retweeted
In 1952, Alan Hodgkin and Andrew Huxley developed a mathematical model describing how neurons generate electrical impulses.
Using the giant axon of a squid, they studied how sodium (Na⁺) and potassium (K⁺) currents influence the electrical activity of nerve cells.
Using the giant axon of a squid, they studied how sodium (Na⁺) and potassium (K⁺) currents influence the electrical activity of nerve cells.
1
3
14
993