Christian Wilms @ChrisWilms23@qoto.org

Do you care about other non-Musk social networks such as Bluesky?

Please boost so we can see what people are thinking.

@ChrisWilms23 thanks for the @! @susanleemburg Indeed, we're making the Buskazi lab version for the community (and our own @t1m used it with neuropixels and regular probes) . So far, a bit over 60 labs have received drives from us and the feedback we got was mainly positive.

@BorisBarbour

Moving to the UK, as a scientist:

Startup package, what’s that?
Entry-level faculty salary lower than your postdoc salary a decade ago, what do you mean?
Child care as expensive as elsewhere where salaries are twice the amount, that can’t possibly be, right?
Buying a house costs several times the cost of a house where you lived before, come on that can’t be that bad?
Ah yes, please pay many thousands of pounds upfront so that we can sign you up with the NHS… no we can’t finance that or let you pay it over multiple months.
We hope you can still apply to the ERC, the local funding bodies are overstretched but surely it will work out?

In other words, please subsidise us from your savings and international collaborations.

This is not a red carpet but a tar pit with snakes, flying arrows and a giant rolling stone catching up to your footsteps.

#uk #academia

We hope to start with neuropixels2.0 soon (multishanks ftw!)

So far I've seen this surprisingly low-tech DIY-ish version https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8177890/

This one that looks nice, but seems almost a bit delicate for rats (ours are enthusiastically destructive): https://www.biorxiv.org/content/10.1101/2023.08.03.551752v1

And I believe there are the Atlas Neurotech headstages (but only for neuropixels 1.0?)

Neuropixels 2.0 probes are now on sale!

Multi-shank, miniaturized, higher density than 1.0.

To order: https://neuropixels.org/post/neuropixels-2-0-multishank-10mm-available-for-sale

Our paper about the probes: https://www.science.org/doi/full/10.1126/science.abf4588

Quick factcheck on zoom changing T&Cs to allow them to use your calls to train algorithms. Result: yup.

A fatal flaw in the "be really hard on yourself as the mechanism to achieve" plan is that even when you achieve you absolutely cannot believe it, because you've really overtrained being hard on yourself. You see a systematic undervaluing that robs you of true information about your work.

So, one of the hallmarks of maladaptive high achievement I look for, as a psychologist studying productivity, is inability to really celebrate.

The program for #ESISyNC 2023 is now online 🥳

✅ 2 days with:
11 speakers | panel discussion | poster session | lunch & coffee breaks | dinner & party

📆 Sept 14 + 15, Frankfurt, Germany

This year's topic: Linking hypotheses: where neuroscience, computation, and cognition meet

➡️ Register now: https://www.esi-frankfurt.de/newevent/

#event #neuroscience #computation #cognition #brainresearch #science #conference #frankfurt

Canon is selling a camera with a 1" SPAD sensor-- an array of Single Photon Avalanche Diodes. it's basically a matrix of single photon detectors. So each pixel can detect individual photons. The gain is huge, but the dark noise is non-zero, so it's not perfect. Nicely, each pixel can operate at a high bandwidth (~ GHz) so it can enable some interesting applications (though usually there are on-chip limits like binning that prevent users from having low-level access to that bandwidth).

$25,000. It's exciting to see this sensor technology to make it into products, but I haven't been able to locate any impressive numbers or example data. For example, BSI sCMOS products can match the minimum illumination spec (0.001 lux) and exceed the SNR: https://www.photonics.com/Buyers_Guide/ProdSpec/Sensors_Detectors/HWK4123/psp8497

I hope companies keep pushing in this direction-- it opens up some very interesting parameter space for scientific measurements.

Product page: https://www.usa.canon.com/shop/p/ms-500

New posting! This one is a #HowTo guide on how NOT to start a research group in #academia - in other words, all the things that I wish I'd done differently or that I wish I'd been more aware of when I started back in 2015. I hope that helps anyone else out there taking their first steps in this direction! (And any comments/additions very welcome). 👇

https://totalinternalreflectionblog.com/2023/07/30/shutting-up-shop-a-short-guide-on-how-not-to-start-a-research-group/

We are now in the middle of "Oops"

Lord Frost seems to be remarkably ill-informed about climate. Doesn't he know that the most likely outcome is that the gulf stream will switch off, freezing Europe while rest of the world fries?
See video by Stefan Rahmstorf
https://twitter.com/i/status/1684149268855107586

@stevesilberman For that take to be true, Elon would have to have had clear goals beforehand and would have needed the necessary focus and dedication over the course of months to work towards those goals.

That seems to be vastly overestimating his abilities.

Occam's Razor says he bought a fun toy and the rest is just rich guy flailing.

Attributing malice to his incompetence is a fool's errand.

I slightly wish the media would switch its messaging from "We're all gonna die!!!" to "Here's what we need to do to stop the dying" (starting with - defund the fossil fuel companies and vote environment...)

https://www.bbc.co.uk/news/science-environment-66229065

America 👏 does 👏 not 👏 have 👏 a 👏 far-left 👏 movement.

- Our far-right are explicit, open fascists.
- Our center-right are extremist nationalists.
- Our center-left are classical conservatives.
- Our far-left are moral centrist capitalist moderates.

Any objective political compass would place politicians like Bernie Sanders, AOC, or Elizabeth Warren barely left of center.

A true far-left movement would be advocating for forced property seizure by the working class. Instead, our "far-left" is like "let's don't do a genocide on trans people today please I haven't even had brunch".

It is now painfully clear that no local government, state or federal agency, or other civic institution should be using Twitter for any purpose other than directing people to alternative platforms.

Using the platform for anything other than a last ditch backup for any kind of emergency communication is clearly a disaster waiting to happen.

New posting! This one is about the number of unpaid hours in #academia, an insight that's come from the realisation of how much time I'm still committing despite now (currently) being unemployed. To what extent is academic research sustained by the massive expenditure of unpaid labour? How much unpaid overtime do you think you're clocking? Let's lift the lid...

https://totalinternalreflectionblog.com/2023/06/29/pro-bono-a-riff-on-unpaid-labour-in-academia/

#Twitter is "no longer a tool that's accelerating science. On the contrary"

https://www.axios.com/2023/06/20/scientific-community-eyes-doors-twitter

@Myndex @lambo Then go where the algorithms are. Honestly, the whole fediverse does not have to be the same. If you want your content searchable, by all means produce for an algorithm. Leave the rest of us alone.

Delighted to announce the 2023 UCL Neuropixels Course.

#Neuropixels probes are transforming #neuroscience. This free online course trains scientists to use them and process their output.

Day 1 (17 October 2023) covers the basics. Day 2 (18 October 2023) does a deep dive into 4 hot topics: the 2.0 release, the latest Kilosort, chronic recording implants, and software to track the same neurons across days.

For more info and to register:
https://www.ucl.ac.uk/neuropixels/training/2023-neuropixels-course