Spencer LaVere Smith

prof at UCSB. neuroscience, optics, imaging, computation http://slslab.org/ also http://labrigger.com/ signal: laveresmith.10

Public Key

npub1yl3knzsdqtxuzj8hn08zcrdpzrc2c50zu6tlnl08hnmnqugalt4smdq722

NIP-05 Address

[email protected]

Profile Code

nprofile1qqsz0cmf3gxs9nwpfrmehn3vpks3pu9v283wd9lelhnmeaeswywl46cpz3mhxue69uhhyetvv9ujuerpd46hxtnfduqs6amnwvaz7tmwdaejumr0dsp33fh2

Publishing to

relay.damus.io

nos.lol

Author Public Key

npub1yl3knzsdqtxuzj8hn08zcrdpzrc2c50zu6tlnl08hnmnqugalt4smdq722

Show more details

Last Notes

2026-05-07 01:37:55 UTC

- reply

Agreed. A program is not “high-risk-high-reward”, unless the failure rate is high. If it has been de-risked, it’s not really the frontier any more. So scientists are comfortable with uncertainty. And this might be one of the most misunderstood aspects of science as far as the general public is concerned.
Also, it is possible to consider separate items: (a)rigor and (b) weight of evidence. They can be correlated or not. It is possible to be careful and rigorous while also providing a only a modest quantity of evidence, just as it is possible to have massive Ns, small P values, and multiple orthogonal lines of data, all of which is of dicy quality. The crucible is whether people can build upon what is reported, or if it is a shoddy foundation that does not support further investigation.

2026-01-22 23:02:07 UTC

#uspol
There is a ton of energy and mobilization out there. Good discipline and communication too.
If some political leader took all of that and focused it on specific policy goals, stayed on message, and stayed on camera, something might actually happen.

2026-01-22 19:55:29 UTC

- reply

I'm still happy with our self-hosted mattermost.
No AI stuff, so far as what I've seen. I think you can connect to AI services, but it doesn't seem like it's forced into it.
I've used Zulip, and it's okay. At least when I used it, it wasn't quite slick enough. It was weird enough (e.g., clunky image posting) that people didn't use it much.

2026-01-13 05:17:47 UTC

Plane spotters.
https://media.neuromatch.social/media_attachments/files/115/886/093/056/966/894/original/d60d7aafb349716a.jpeg

2026-01-07 08:31:38 UTC

- reply

Cool! I wonder what the next stop is. Midway? Majuro?

2026-01-01 00:42:11 UTC

- reply

I appreciate the engineers and engineering that make these launches work. It still captures my imagination and inspires awe, even as routine as it has become. And I appreciate all of your photos of them.

2025-12-29 21:41:11 UTC

"I don’t even focus on people. I am at war with certain ideologies and ideas, and I want them expunged. I want to turn them into phrenology. That’s what I want. But I don’t want the people out."
- Ta-Nehisi Coates
https://media.neuromatch.social/media_attachments/files/115/805/020/883/856/439/original/5889aaea2656cf4b.jpg

2025-12-28 01:04:48 UTC

- reply

Wow! Interesting. Are both in the same neighborhood? Have there been many (non-cosmetic) modifications or updates over the years?

2025-11-15 21:38:59 UTC

- reply

wish granted

2025-11-12 09:30:16 UTC

- reply

Here’s a US study that has it at $2.56 for every $1. https://news.harvard.edu/gazette/story/2025/03/nih-funding-delivers-exponential-economic-returns/ These analyses are hard to compare, though. Maybe it suffices to say that they both have positive returns.
I like the artist income project, and basic income programs in general seem to have a good track record. At least as far as I’ve read.

2025-10-30 21:55:20 UTC

- reply

Amen. Microscope software is like this too. There is so much room for improvement— refining the user experience making an “easy” mode.

2025-10-27 06:51:33 UTC

- reply

They deorbit, like most modern satellites https://dailygalaxy.com/2025/10/starlink-satellites-falling-more-than-ever/ so hopefully the debris risk is low.

2025-10-14 07:22:05 UTC

- reply

I hadn’t thought about it until just now, but it seems feasible to unify a lot of his films into a single universe, with Ice Cube playing the same person in the films, but under different names. Boyz in the Hood, Friday, and Barbershop are the early years, and then he gets into law enforcement: Ride Along and then Jumpstreet. And then progresses to covert ops in XXX. Plausible.

2025-10-07 20:44:26 UTC

I can kind of understand the value of grad student TAs. Every year, this material gets more and more reflexive to me, and I get further and further from where the students are at. Me: "One slide is enough for the Helmholtz decomposition. They probably already know it from high school." #iteachphysics

2025-10-03 07:23:53 UTC

- reply

Yes. Well said. Accountability matters. https://labrigger.com/blog/2025/09/22/accountability-is-a-way-to-determine-truth/

2025-09-28 04:02:49 UTC

You can probably guess what the color code is.
https://media.neuromatch.social/media_attachments/files/115/279/931/109/339/711/original/797211a71fb82732.png

2025-09-21 23:49:59 UTC

I was trying to recall some movie from the 1980s that was some sort of dystopian sci fi with heavily modded junkyard vehicles in a desert landscape that wasn't Mad Max. I used 2025 technology to search for it, and it turns out there are a metric ton of movies that fit that description.
https://en.wikipedia.org/wiki/Equalizer_2000
https://en.wikipedia.org/wiki/Battletruck
https://en.wikipedia.org/wiki/Cherry_2000
https://en.wikipedia.org/wiki/Spacehunter:_Adventures_in_the_Forbidden_Zone
https://en.wikipedia.org/wiki/World_Gone_Wild
(btw, none of these are what I was looking for... still searching)

2025-09-07 06:01:03 UTC

- reply

At least one is. Thanks for sharing this.

2025-08-22 19:08:47 UTC

- reply

I agree that the first two, kana, aren’t hard to learn.
But that third one, kanji, takes some work. Thousands of characters to learn and 2-6 pronunciations for each (or more, like 生). (By contrast, Mandarin characters often have a single pronunciation.) Still, you pick it up as you go. It’s fun to learn.

2025-08-03 00:53:09 UTC

I use Overleaf (LaTeX) for papers and grants but it wastes my time with odd behavior, so I’ve been trying Typst: https://labrigger.com/blog/2025/08/02/switching-to-typst/
If you like Overleaf, you might like Typst even more. Here at 10 quick start tips for Overleaf users trying Typst https://labrigger.com/blog/2025/08/02/10-quick-start-items-for-overleaf-to-typst/

2025-07-31 01:24:18 UTC

- reply

Maybe it’s worse. At least the formula is deterministic. The star thing isn’t, so it can be a fig leaf for all kinds all of the stuff you mention later in the thread like nepotism, bias, …
I don’t have a perfect solution either, but I’ll stick my neck out. When I review others, I think “Well, this institution hired a person in subsubfield X. Do other people in that subsubfield find their output, in whole, valuable?” So input from those experts, who have read (and hopefully cited) the work is useful. But I think at max 2 bits of dynamic range per reviewer. Finer grading than that is not reliable. Letters can be nice, but even just a three point scale rating (basically average or over a standard deviation above or below) can be helpful.

2025-07-17 03:50:24 UTC

- reply

Whole cell patch clamp in vivo is still the highest fidelity approach (and my first love). But it doesn’t scale. 1 at a time for pros, 2 at a time for a few people on the planet, and up to 4 at a time is basically just Jouhanneau and Poulet.
Electrode arrays are great for high speed and they catch single spikes, but sample mostly cells very close to the shank. That said, the shank can be long and it is as easy to sample 3 mm deep as it is to sample 0.5 mm deep.
Calcium imaging densely samples many more neurons than electrode arrays. Orders of magnitude more. But the cells have to be superficial (roughly < 1 mm for 2p, < 1.6 mm for 3p). Or you’ve got to remove some tissue. Imaging is slow, but multi spike events are larger so still detected and in ephy spikes are often binned into broader time windows, so the difference is often not that large. You can see spindles in calcium imaging.
Each of these three techniques have unique strengths relative to the others. I like them all.

2025-07-05 21:03:39 UTC

- reply

Agreed. That would be nice.

2025-07-05 11:14:03 UTC

- reply

re: item 4.
https://www.3doptix.com/academy-post/optical-design-software/
https://github.com/UMassIonTrappers/PyOpticL?tab=readme-ov-file
https://www.quadoa.com/index

2025-07-05 11:10:19 UTC

- reply

1 is brilliant. Do it.
2 is not my cup of tea. Little too close to home.
3 is cool. Fun idea.
4 would be fantastic. There are several things close to this for optics.
5 seems like a fun project for a student.

2025-06-25 01:02:54 UTC

- reply

To put numbers to it, we can calculate the golf courses per capita. People use units of people per golf course and lower is better.
Iowa is the #3 US state on this metric. With a golf course for every 7000 people. https://golfblogger.com/golf_courses_per_capita/
Note that this is better than even Scotland, which has 9000 people per course.
https://aussiegolfer.com.au/countries-with-most-golf-courses-per-capita/
Just a silly bit of trivia.

2025-06-24 17:06:15 UTC

Here’s a map of golf courses in the US. Note that Iowa, Minnesota, and nearby areas have a bunch of courses distributed across rural cropland. In other states, courses are concentrated in or near large population centers. The Midwestern US is home to bunch of affordable public golf courses. Farm kids can golf.
https://media.neuromatch.social/media_attachments/files/114/739/414/868/565/886/original/934d21f40c5b63e0.jpeg

2025-06-21 13:47:00 UTC

- reply

During my PhD my advisor Tom Otis told me about them https://www.sainsburywellcome.org/web/people/tom-otis Maybe not uniformly wholesome, but I recall some positive messages. Haven’t listened to them in many years.

2025-06-21 12:21:50 UTC

- reply

Regurgitator

2025-06-19 23:22:59 UTC

- reply

Not to mention the nearby Dunluce Links and Slieve League!

2025-06-19 15:00:43 UTC

The vibe at 7:00am at a city park is the same in every civilization on Earth.
Same mix of people, in the same ratios, sharing the same polite smiles, greetings, and chit-chat. People taking their dogs on morning walks, people getting some exercise in before they start their day, and others just enjoying the vibe. Instantly recognizable. The dev team just did texture swaps for all of these expansion packs because it was already perfect.

2025-06-17 04:38:34 UTC

This is my new favorite use of AI.
Che-Hang Yu just put together a tour de force in non-inertial angle amplification. It seems like magic.
Then he used AI to make a fake podcast discussing the manuscript. It's wonderfully hilarious! The performance is quite good, and it nails a bunch of podcast cliches. Here I listen to it and add a bit of commentary. Check it out!
https://www.youtube.com/watch?v=qO1nenu6LiI&feature=youtu.be
(and here's the link to the preprint: https://www.biorxiv.org/content/10.1101/2025.06.13.659647v1 )
https://media.neuromatch.social/media_attachments/files/114/696/853/025/081/493/original/9814052f1abdd46c.png

2025-06-16 23:49:40 UTC

Federal workers are honorable and valuable members of society. They shouldn't have to work overtime to fix this mess, but this quote is another example of the integrity that can be found within their ranks.
https://media.neuromatch.social/media_attachments/files/114/695/710/758/842/196/original/7e0835c2360a6814.png

2025-06-15 05:47:58 UTC

A discussion on pre-K and early childhood education made me think of first person survivor bravado, which is saying: "Eh, the system is fine. I made it through."
Our parents and grandparents made the system better when they had the chance. We might not have fared so well with the system as it was a generation or more before us.
So when people talk about making improvements, I listen. It's not about making kids soft, or participation trophies, or lowering the bar.
It's about being more effective at pulling people through to the light at the other end of the tunnel, despite challenges they face.
It's about not letting the next Einstein or other impactful minds fall by the wayside if we can do better.
It's about helping a larger fraction of kids develop into more functional, healthier, and happier people.
Yes, I'm fine too. And I'm also optimistic that we can do even better.

2025-06-14 22:24:25 UTC

US universities are great bang-for-the-taxpayer dollar in terms of improving health, tech, and the economy.
US government grants to private companies (e.g., defense contractors) are often about 50% or more in indirects. By contrast, universities typically charge < 25% in indirects. Plus it has been declining even in just the past 15 years.
PDF source: https://www.congress.gov/crs_external_products/R/PDF/R48540/R48540.1.pdf
https://media.neuromatch.social/media_attachments/files/114/684/040/050/732/195/original/86ae29a837521cab.png

2025-06-09 20:37:08 UTC

- reply

Works fine for me. I have spent $0. Dozens of users. Bandwidth demands are low.

2025-06-07 13:40:27 UTC

- reply

They’re not motorized and manually moving them can cause extra motion— eg an injection needle can move laterally 0.2 mm while twisting the z dial. Sometimes that’s no big deal and manual is okay.
The display is handy. Zero at a landmark and then move to the target coordinates. Saves reading markers on the stereotax. Ergonomics like this sometimes matter (eg doing many surgeries in a day) sometimes they are less important.

2025-06-06 01:12:07 UTC

I got these catalogs when I was a kid and they ignited my imagination in a way that Amazon, Temu, Alibaba, etc. cannot. There was organic and authentic enthusiasm for the possibilities and wonder in each item description.
If you're into this kind of stuff, it's worth checking out. Maybe we're a dying breed and can't keep this company afloat, or maybe there is some market for this stuff that social media can help the company reach.
https://arstechnica.com/gadgets/2025/06/american-science-surplus-is-fighting-for-its-life-heres-why-you-should-care/

2025-05-18 06:17:45 UTC

- reply

You wrote "Ultimately, we need to start building new multi-country organizations..." -- doesn't HFSP merit a mention? https://en.wikipedia.org/wiki/Human_Frontier_Science_Program Imagine if we expanded this program.

2025-05-16 03:19:19 UTC

- reply

Your animations are important tools, and I expect they will have some longevity. Thank you for sharing them so openly.
I do wonder about the longevity of videos, especially short forms. Old lecture videos from Gilbert Strang might bear preservation, updating as formats evolve. Text, especially products that end up in libraries, have excellent odds at preservation. I am less certain that 20 year old 30 second tictoc videos have the same preservation potential.

2025-05-16 03:11:44 UTC

- reply

Agreed. And in the thick of it, I’ve got to get a pencil and paper and work it out myself. Making some stupid mistakes, some sign errors, finding them and fixing them— after that I might get it. Neither text nor video can take the place of that.

2025-05-12 14:58:59 UTC

- reply

Fun article, and fun proposal. I've read Gertner's book, and other items on Bell Labs. And I've spoken with people who worked there, and I work with ex-Bell Labs people now. The guy who hired me here came from Bell Labs.
All of these accounts of Bell Labs are right in a lot of ways, and I like the points emphasized in this article about "managing" creative researchers and scientists.
I would add that there is a lot that can never be replicated-- it was a smaller community world-wide, and there was a broad frontier that could be navigated with relatively small interdisciplinary teams. There is still a lot to explore, and a lot of the points mentioned in that web post are still valid, but you could follow them perfectly and still not have anything close to Bell Labs. And that's okay. We should recognize the time that we're in, and do the best for progress where we're at now.
Also, the tensions between Bell Labs and the government wasn't that great of an arrangement. Who gets to decide what kind of research gets funded? I bet Google would claim that it is already spending a lot on "fundamental" research. How about just sending that 2% to the NSF?
I think the model we have now is pretty good. NSF/NIH/DOE/DOD/etc. funding allocated to Programs, and (after peer review) going to university labs for execution.
Needs more support, and how we handle taxes and funding of universities is another question.

2025-05-09 13:24:40 UTC

- reply

https://media.neuromatch.social/media_attachments/files/114/478/093/333/487/630/original/420b48ba0f67773d.jpeg

2025-05-09 00:30:08 UTC

Just received my seemingly weekly reminder from UC Santa Barbara to:
https://media.neuromatch.social/media_attachments/files/114/475/047/561/585/974/original/27e36a5236da7bf0.jpeg

2025-05-06 22:49:02 UTC

I’m not leaving.

2025-05-04 18:40:52 UTC

https://media.neuromatch.social/media_attachments/files/114/451/021/779/936/688/original/e3631b6beeb7161c.png

2025-05-02 19:37:22 UTC

- reply

I understand, and I've heard good people say that before.
If each member of their editorial board sent at least one manuscript there every other year, that alone could help raise its profile. https://nbdt.scholasticahq.com/editorial-board
I hope that over time (established in 2018) it rises on people's lists of places to publish their work. It is indexed, and it seems highly sustainable.
I kind of wish the name was a bit less awkward (but it still better than PNAS).
Konrad discusses the journal in this 2020 video (they went with "NBDT" to be broadly inclusive). https://www.crowdcast.io/c/nbdt-neurons-behavior_1

2025-05-02 14:08:26 UTC

- reply

I like NBDT and I hope people publish there more. Part of the discussion at its inception was just how dirt cheap and simple scientific publishing can be, and NBDT is the proof that neuro can publish that way as well as com sci and other fields can.

2025-04-28 21:42:49 UTC

Glass is dead.
Humankind is overdue for moving past glass for state-of-the-art optics. We’re never going to be completely rid of glass, but we have been coasting on 17th century technology for 21st century applications.
https://labrigger.com/blog/2025/04/27/glass-is-dead/

2025-04-17 05:18:43 UTC

UCSB was one of the first 4 nodes on the internet. https://arstechnica.com/gadgets/2025/04/a-history-of-the-internet-part-1-an-arpa-dream-takes-form/
https://media.neuromatch.social/media_attachments/files/114/351/609/880/020/531/original/dc3ab6b3720e09e3.jpeg

2025-04-17 03:15:29 UTC

- reply

And a 5-10 minute walk to the beach. Be well, Neighbor!

2025-04-16 02:59:43 UTC

Social media bots can trace their roots to a little program written in the 1960s.
“What I had not realized is that extremely short exposures to a relatively simple computer program could induce powerful, delusional thinking in quite normal people.”
https://labrigger.com/blog/2025/04/15/eliza-exe/

2025-04-13 00:20:12 UTC

- reply

Yeah! There’s a special scheme that Nashville session players use. I wonder what other systems might be out there.

2025-04-12 19:24:27 UTC

- reply

Nashville itself has developed its own system. It’s interesting.

2025-04-12 17:55:10 UTC

- reply

I found this comment: In most countries where I've had the chance to play, chord sheet notation follows the UK/US norms but orally people use the longer local form. For example, in France a songbook will show Bb7, but people will say "si bémol 7". from: https://www.jazzguitar.be/forum/theory/77636-musical-terms-different-languages.html

2025-04-12 17:51:27 UTC

- reply

Exactly. I’m not a sophisticated musician, but I can read music, and I’ve played several instruments, I played in rock bands, I was an ASCAP-paid … not much, but something. So I was surprised to learn that the US note names were a US-only thing. I wonder how people learn guitar chords in other countries.

2025-04-12 17:44:26 UTC

- reply

Exactly. It’s a US thing. As far as I can tell, all other countries use fixed solfege. I had no idea that A-B-C… for note names was a US exclusive thing.

2025-04-12 16:08:46 UTC

Ikuko (my amazing wife) is an excellent pianist. Her fingers wizzing all over the keys playing intricate classical pieces. She is teaching our young daughters piano now, and something just happened that blew my mind.
She learned piano in Japan. I learned piano (the tiny bit that I know) in the US. It turns out that we learn "do-re-mi-fa-sol-la-ti-do" completely differently. https://en.wikipedia.org/wiki/Solfège
She asked me, "is there an 'I' (letter i)".
Me: "What?"
"Like 'do' is C, 're' is D... is there an I?"
"Um. 'do' can be whatever the root is.
"No, it can't be. It's always C."
"Um, right here on wikipedia, it says that you shift it."
"Wikipedia's wrong. It is always C."
"Okay, maybe this is a difference in how we learn music between Japan and the US."
"No, what I'm explaining is German."
"Okay, but anyways, I still don't understand."
"If do is C, and re is D, then..."
"Then it's C-D-E-F-G-A-B. It starts over."
I went over to the piano and said the letters that correspond to each key. She immediately understood and went on with the lesson for our daughter.
It blew my mind that she never connected notes of the scale to letters. I guess they just use the "do-re-mi" scale, always referenced to C, and that's how they learn music. Totally logical in retrospect. But it just blew my mind that piano-simpleton me got to teach my piano-brilliant wife the names of the notes.

2025-04-10 06:25:26 UTC

maybe obvious, but I welcome the reminders. especially when we’re in the thick of a grant proposal review meeting.
https://media.neuromatch.social/media_attachments/files/114/312/236/350/203/347/original/2253564fe82c8342.jpeg

2025-04-07 06:30:26 UTC

I gave some money to https://www.janm.org just now.
https://media.neuromatch.social/media_attachments/files/114/295/269/363/565/066/original/b4c09d3e374e29a6.jpeg

2025-04-02 18:07:29 UTC

About 15,000,000 kg of rocks from outer space enter the Earth's atmosphere each year.
That mass is about the same as a quarter of a million people.
Which is about the population of Buffalo, NY.
You're welcome.

2025-03-31 23:20:21 UTC

Winston visited campus today. #ucsantabarbara
https://media.neuromatch.social/media_attachments/files/114/259/603/215/021/863/original/d4a54b5407895464.jpeg

2025-03-21 23:32:54 UTC

https://media.neuromatch.social/media_attachments/files/114/203/030/367/377/574/original/4e8599e34549f37f.jpeg

2025-03-19 17:57:29 UTC

Carl Olsen (Carnegie Mellon U) passed away recently. Matt Smith and S. P. Arun wrote a nice article. https://www.sciencedirect.com/science/article/pii/S0896627325001680
Each time I met with him, he was always sharp and open-minded, and fun to chat with. A great career with a ton of positive impact.
https://media.neuromatch.social/media_attachments/files/114/190/378/574/916/792/original/23e366f210071f8c.png

2025-03-19 17:54:03 UTC

I didn’t know his name, so I looked him up. Alec Toth was barely 19 years old when this letter was written. And he had worked for the company for less than 1 year.
Following this moment of compassion, Alec Toth knocked our one of the most massive careers in his industry: https://en.wikipedia.org/wiki/Alex_Toth
https://media.neuromatch.social/media_attachments/files/114/190/358/170/085/319/original/99a501278b15c1c8.png

2025-03-18 14:41:34 UTC

Here’s a nice web interface for exploring mesoscale data on brain-wide neuronal wiring.
https://web.ece.ucsb.edu/~riichirohira/TopBM-2.06/
Riichiro Hira put it together a while back, but it is not yet well known. It’s fun to explore, and handy to quickly check something.
https://youtu.be/WUuWg7FPWvY?feature=shared

2025-03-09 17:03:03 UTC

- reply

Cool. I might add a line between the first and second one with the root noted in the radical, like is done on line 3. Just to make it quicker to recognize the symmetry.

2025-03-03 17:31:07 UTC

- reply

Here's a join link. If it doesn't work, then let me know and I'll add people manually. I just need an email address.
https://2p.ece.ucsb.edu/signup_user_complete/?id=uqr31cgtf3d7bcrzm6949rcz4r&md=link&sbr=sa
To answer your questions:
- No limit on number of people. It is super-easy to host thousands. Companies just charge this way to make money. The real bandwidth issues and storage issues are negligible.
- No conditions. It's not a big public group, so it doesn't tend to attract trolls or spam. Whoever wants to join, can.
- Go ahead and spread the word. I am grateful.
- It runs off a server on the UCSB campus. All users have access to the data. If anyone wants a dump of it (i.e., a snapshot), I'm happy to assist.
- The data has a mirror. It is also on campus, so if there was a major disaster, we'd lose both copies. But for routine drive failure we're covered.
- No limits on numbers of messages or attachments. Just whatever the software limits are. I'd appreciate it if people didn't try to game that and upload PB data sets. I'm guessing no one will. Movies, images, etc., are fine of course. I'll say something if it becomes an issue at all.
- It's free. I promise to never charge for it. I also promise to never sell any of the data. If there are any concerns about that, I'm happy to put that in writing, if someone wants to draft something up. It's just a bit of community organization I'm happy to do because I'm part of the community and I like to have good resources like this.

2025-03-03 16:32:35 UTC

- reply

I host a slack-like group… it is focused on neuro, 2p, reporters, etc. I’m happy to expand outward from there for more neuro topics. Practically unlimited storage, https://2p.ece.ucsb.edu mattermost— there are ios and android apps, desktop, browser..

2025-02-05 19:22:59 UTC

- reply

added

2025-02-03 17:28:27 UTC

- reply

brb gonna make some nanoscale disordered structures out of superconducting material using 2-photon 3d printing

2025-01-27 00:38:08 UTC

- reply

I tried Zulip for a while. It handles threads in a nicer way than Slack, but it was different enough from Slack (and looked a bit clunky) that engagement was low. Getting people to use Mattermost has been easier so far.

2025-01-25 20:28:05 UTC

- reply

Ah, thanks for the heads up!
I still pay for Slack for my internal lab communications. There's an academic discount but it's still a lot to spend for what it is. Especially given the competition. I may phase it out in favor of Mattermost. It's good to hear that your data migration went smoothly. Cheers!

2025-01-25 20:15:39 UTC

- reply

What changes with Slack's data deletion policy on Jan 28th?
I've been running a self-hosted Mattermost instance for a while now and it seems fine. Not as slick as Slack in some ways (e.g., pasting in images and videos), but it's fine.
Most importantly: no per-person billing. It's effectively free-- just using hardware we already had, and the bandwidth is not a big deal.

2024-12-29 08:09:09 UTC

- reply

On the one hand, I want to remark that 3p imaging is typically right up against the damage envelope for biological tissue.
On the other hand, I want to encourage you to try it. Go for it! Can you get 5p or 10p imaging to work at all? Even with a non biological sample like simple fluorescent plastic. What are you thinking of using as an excitation source?

2024-12-22 19:12:32 UTC

I'm excited that Pacific Optica is releasing their new scan optics. https://pacificoptica.com/ventana/ For those of you who know two-photon imaging well, check out the specs on the link. These optics are unprecedented. They're easy to integrate into a rig (using Thorlabs SM2 and SM3 threads) and they provide a way bigger diffraction-limited imaging field-of-view. No more vignetting due to the scan engine.
https://media.neuromatch.social/media_attachments/files/113/698/052/768/899/564/original/c2b64614703e949f.png

2024-12-21 21:18:37 UTC

- reply

You’re correct. Most designers still strive for broad achromacity in the visible. They’re good objectives, but like you said, there is room for improvement for multiphoton. I tried the Leica and Olympus out a while back. They were okay.
I prefer the Cousa objective and the Diesel2p system’s objective (both have correction collars), but I’m biased of course. https://pacificoptica.com We do optimize for 2p and 3p performance. It makes a big difference.

2024-12-10 05:30:51 UTC

- reply

Elite ed may be overvalued, but comparing it to the 2007 housing bubble seems odd. Less than 1% of undergrads are at Ivy League institutions. 0.3%, according to recent numbers on wikipedia.
“Mass delusion about underlying value” — there might be something to that. But evaluating value for an undergrad degree is nontrivial.
I think one story about higher ed that needs to be told is how “eliteness” is no longer monopolized by the Ivy League. Perceived prestige is stuck in the usual places, but actual value and performance is spread around a bit more than it used to be.

2024-12-10 03:57:56 UTC

What would you change about academia life?
There’s a discussion on bsky about saying “no” to service requests. But really, why is it even such an issue? A healthy institution or community should be able to spread the load. This is a strange job.

2024-11-27 05:33:02 UTC

I sympathize with the moderation issue. Those of us who were online with pre-WWW internet and the BBS scene know that it didn't take algorithm-driven social media to make people act uncompassionately online.
Online or off, we've always been sharing the earth with a spectrum from opportunistic jerks to truly evil people. Massive numbers of them. We will have to fight for good our whole lives.

2024-11-27 00:13:21 UTC

Postdoc at the Allen Inst. Applications due Dec 16. https://alleninstitute.org/careers/internships-and-postbac/shanahan-foundation-fellowship/ Good compensation, $104k + benefits

2024-11-25 18:45:23 UTC

- reply

If I'm understanding you both correctly, then I think I agree with you both.

2024-11-23 04:29:35 UTC

- reply

...yes... keep going...

2024-11-22 17:49:12 UTC

Please spread the word. Applications to UCSB's Dynamical Neuroscience PhD program are due Dec 1.
https://www.dyns.ucsb.edu
It's a unique program:
- Focused on dynamical neuro: behavior, circuits, etc.
- Includes grad level math training for analysis and modeling: diff eq, linear analysis, etc. This is a big emphasis of the program.
- Integrated with AI/neuro labs and initiatives on campus for ML/AI/neuro.
- Interdepartmental: bio, physics, engineering, psych, stats, etc.
- Small classes, great community.
- Lots of cool labs-- e.g., navigation and vision and dendrites and computation, mice, flies, humans, ...
- It's a nice place to spend your PhD years.

2024-11-18 04:11:17 UTC

Something recently reminded me of the Tham Luang cave rescue in 2018. I looked it up and it was even more technical and complex than I remembered. The people who went to such lengths-- costing some their lives
https://en.wikipedia.org/wiki/Tham_Luang_cave_rescue
https://media.neuromatch.social/media_attachments/files/113/501/978/702/312/515/original/45e5a69a43764420.png

2024-11-17 19:44:42 UTC

New post: If you use visual stimulation while patching in vivo in mice, then a shroud can be useful. Here's a 3D printed one. Explanation, photos, CAD files.
http://labrigger.com/blog/2024/11/17/a-fantastic-shroud-for-visual-stimulation-during-imaging/
https://media.neuromatch.social/media_attachments/files/113/500/002/666/763/249/original/1c6ab48755169a57.png

2024-11-17 18:25:40 UTC

A dichroic slider with a kinematic mount for adjusting the dichroic, and magnets to lock in place to the same position each time. Feels nice.
more on https://2p.ece.ucsb.edu/
https://media.neuromatch.social/media_attachments/files/113/499/685/311/129/150/original/f4fd2ca9b5f1e3e7.png

2024-11-17 17:21:10 UTC

Your optics need to be thick to be flat.
recent post on https://2p.ece.ucsb.edu/
if you're not already a member: https://2p.ece.ucsb.edu/signup_user_complete/?id=uqr31cgtf3d7bcrzm6949rcz4r&md=link&sbr=sa
https://media.neuromatch.social/media_attachments/files/113/499/441/334/698/240/original/1894e14dd7b1fe08.png

2024-11-15 16:41:33 UTC

Meeting: Vision: A Platform for Linking Circuits, Behavior & Perception
https://meetings.cshl.edu/courses.aspx?course=C-VISION
Lead by Kristina Nielsen (JHU) and Joseph Carroll (Wisconsin)
Tons of great people involved: Movshon, Sommer, Dunn, Gordon Smith, Priebe, Kastner, and more...
https://media.neuromatch.social/media_attachments/files/113/487/948/352/830/381/original/eb19208b872e9729.png

2024-11-10 06:08:45 UTC

topical
https://media.neuromatch.social/media_attachments/files/113/457/162/446/456/035/original/dd7919250af61bfb.jpeg

2024-11-10 03:44:14 UTC

There is less engagement on mastodon. And that’s okay. Probably a good thing.

2024-11-07 23:06:31 UTC

UCSB has a bunch of high level admin jobs all under search at the same time:
- a chancellor (for the first time in decades)
- a vice chancellor for research
- a dean for MLPS (math, life sciences, physics)
- a dean for the CCS (honors college)
- a dean and AVC for grad div
and more.
There'll be some new faces soon. Yours?

2024-11-05 05:06:23 UTC

Feeling this mood at the moment. Good luck, World.
https://media.neuromatch.social/media_attachments/files/113/428/604/275/638/517/original/545a8e3cb6f5d64b.jpeg

2024-11-04 18:19:08 UTC

Big paper from the International Brain Lab group: https://elifesciences.org/reviewed-preprints/100840 One takeaway is that using stereotaxic coordinates involves variance which "... is sizeable enough to impact probe targeting to desired brain regions and poses a limit on the resolution of probe targeting." Lots of labs do posthoc histology verification, which can help.
Functional targeting is my favorite. Here's a figure from the supplement to one of our preprints. Functional targeting is way better than using skull landmarks. https://www.biorxiv.org/content/10.1101/2023.08.27.555017v1
https://media.neuromatch.social/media_attachments/files/113/426/054/780/781/837/original/95f0cfb1e64f21b5.png

2024-11-04 04:11:04 UTC

California has an excellent voters’ guide https://voterguide.sos.ca.gov plainly worded, statements from candidates, arguments for and against measures, fact checked, budget analysis… and in a bunch of different languages. There’s room for improving how California is governed, but our voter guide is well done. Do other states do it well too?

2024-11-04 04:00:05 UTC

Dots are clinics. Darker colors are counties that are further from a clinic. There is no safe way to insert politicians and government officials into hard healthcare decisions. I trust women and doctors. Everyone should have help and care nearby. Let’s get to work and make things better.
https://media.neuromatch.social/media_attachments/files/113/422/656/163/177/289/original/9b65686a23f63164.jpeg

2024-11-04 03:52:46 UTC

US citizens: Please vote. It matters. Even if you’re in a state where the presidential vote is lopsided, your vote matters. Local matters need your vote.

2024-10-30 22:15:51 UTC

Pacific Optica has some new products coming out. Faster imaging across larger fields-of-view. I'm excited about it. The patents are in place. Just waiting on marketing material. Let me know if you're interested and want to know more. Especially if you need a new 2p imaging system soon, or just want your current one to do more. https://pacificoptica.com
https://media.neuromatch.social/media_attachments/files/113/398/547/118/509/455/original/c02561ca7314802e.jpeg
https://media.neuromatch.social/media_attachments/files/113/398/547/441/641/100/original/8fd53ddd8c47d342.jpeg

2024-10-08 20:42:43 UTC

- reply

Yes. This os work worth celebrating. Gatekeeping Nobels is foolish. I hope people are enjoying themselves.

2024-10-03 20:24:25 UTC

The Diesel2p system (25 mm^2 field-of-view) has an 8 mm working distance air objective.
That big air working distance is nice, it also lets a lot of light in from the sides. And we care about that because we use visual stimulation in our projects.
Fortunately, PhD student Joe Canzano made a nice 3D printed cap for the objective that fits to the head plates that we use. You can customize it for yours.
More pictures and photos are on the Nemonic discussion board: https://neuromatch.social/@sls/112629868836444772
https://media.neuromatch.social/media_attachments/files/113/245/355/623/017/884/original/6b336eea53058aae.png