Matt Zefferman
@mzefferman.bsky.social
Any views expressed are, at most, my own and do not reflect those of any organization.
DoggCatcher >> Spotify
November 6, 2025 at 6:58 AM
DoggCatcher >> Spotify
When I was in the Air Force, many of my troops were on SNAP.
October 27, 2025 at 2:39 AM
When I was in the Air Force, many of my troops were on SNAP.
"The toppings contain potassium benzoate." Whenever getting ice cream.
what's the stupidest/randomest Simpsons quote that lives in your head rent free? Mine is the urge to say "you said go to bread" every time I am about to head to bed.
October 2, 2025 at 2:13 AM
"The toppings contain potassium benzoate." Whenever getting ice cream.
Reposted by Matt Zefferman
**PhD POSITIONS** I am recruiting 1-2 PhD students to work on sociality and cognition in Monk Parakeets, starting Fall 2026. Full ad is here: hobsonresearch.com/index.php/20... (please help spread the word!)
September 23, 2025 at 6:58 PM
**PhD POSITIONS** I am recruiting 1-2 PhD students to work on sociality and cognition in Monk Parakeets, starting Fall 2026. Full ad is here: hobsonresearch.com/index.php/20... (please help spread the word!)
By my departmental colleague.
When Japan’s cruisers met China’s ironclads at the Yalu River, the outcome stunned the world. What are the implications today?
The Importance of the Battle of the Yalu
Editor’s Note: This is part of a new series of essays entitled “Battle Studies,” which seeks, through the study of military history, to demonstrate how
warontherocks.com
September 17, 2025 at 1:47 AM
By my departmental colleague.
This was the most played non-kid song on my playlist last year.
Lauryn Hill releases “The Miseducation of Lauryn Hill” this week in 1998 — making her not only the first hip-hop artist to win a Grammy for Album of the Year but also the first woman since Debbie Gibson to debut at #1 with a song she wrote, recorded and produced on her own.
“Doo Wop (That Thing)”
“Doo Wop (That Thing)”
August 26, 2025 at 4:29 PM
This was the most played non-kid song on my playlist last year.
I once wrote a sentence that stated that the iterative development of weapons technology was akin to what evolutionary biologists call an "arms race."
Canon
xkcd.com
August 25, 2025 at 8:26 PM
I once wrote a sentence that stated that the iterative development of weapons technology was akin to what evolutionary biologists call an "arms race."
48 - 50 (not counting airports or drive-bys). I'm pretty sure I have not visited Tucson or Tulsa as an adult, but maybe as a kid. Disappointing that this does not include cities in Alaska or the entire upper half of the country between St Paul and Portland.
July 31, 2025 at 3:32 AM
48 - 50 (not counting airports or drive-bys). I'm pretty sure I have not visited Tucson or Tulsa as an adult, but maybe as a kid. Disappointing that this does not include cities in Alaska or the entire upper half of the country between St Paul and Portland.
Reposted by Matt Zefferman
Applications are open for SFI's 2026 Complexity Postdoctoral Fellowships
If you’ve recently earned a Ph.D. in any scientific field and want to pursue independent, transdisciplinary research, consider applying.
Deadline: October 1, 2025
Apply here: santafe.edu/sfifellowship
If you’ve recently earned a Ph.D. in any scientific field and want to pursue independent, transdisciplinary research, consider applying.
Deadline: October 1, 2025
Apply here: santafe.edu/sfifellowship
July 28, 2025 at 5:50 PM
Applications are open for SFI's 2026 Complexity Postdoctoral Fellowships
If you’ve recently earned a Ph.D. in any scientific field and want to pursue independent, transdisciplinary research, consider applying.
Deadline: October 1, 2025
Apply here: santafe.edu/sfifellowship
If you’ve recently earned a Ph.D. in any scientific field and want to pursue independent, transdisciplinary research, consider applying.
Deadline: October 1, 2025
Apply here: santafe.edu/sfifellowship
Here are my two for the Double Feature.
July 21, 2025 at 10:01 PM
Here are my two for the Double Feature.
Reposted by Matt Zefferman
Dream internship for someone out there! Paid, Remote, and with @xercessociety.bsky.social 🐝 🪲 🦋 recruiting.paylocity.com/Recruiting/J...
The Xerces Society - Science Illustration Intern
JOB TITLE: Science Illustration InternLOCATION: Remote position. Candidates must have a personal office location suitable for work purposes, with broadband internet access, and...
recruiting.paylocity.com
July 9, 2025 at 2:14 PM
Dream internship for someone out there! Paid, Remote, and with @xercessociety.bsky.social 🐝 🪲 🦋 recruiting.paylocity.com/Recruiting/J...
Reposted by Matt Zefferman
In non-lab fields, faculty typically help MA/PhD students develop solo projects that aren't part of the faculty member's research, and that the faculty won't be a coauthor on. Faculty in non-lab fields also typically increase their advising load post-tenure, limiting their own time for research. 2/
July 8, 2025 at 12:48 AM
In non-lab fields, faculty typically help MA/PhD students develop solo projects that aren't part of the faculty member's research, and that the faculty won't be a coauthor on. Faculty in non-lab fields also typically increase their advising load post-tenure, limiting their own time for research. 2/
I watched the whole Frozen movie with my daughter, and not once does Elsa wear a dress with her own picture on it.
July 5, 2025 at 11:43 PM
I watched the whole Frozen movie with my daughter, and not once does Elsa wear a dress with her own picture on it.
Reposted by Matt Zefferman
'Meredith,' some guys ask, 'why won't you shove AI into Signal?'
Because we love privacy, and we love you, and this shit is predictable and unacceptable. Use Signal ❤️
Because we love privacy, and we love you, and this shit is predictable and unacceptable. Use Signal ❤️
June 19, 2025 at 7:59 AM
'Meredith,' some guys ask, 'why won't you shove AI into Signal?'
Because we love privacy, and we love you, and this shit is predictable and unacceptable. Use Signal ❤️
Because we love privacy, and we love you, and this shit is predictable and unacceptable. Use Signal ❤️
In the four-year-old's process of sorting out "real" from "pretend," my favorite part is discovering the fantastical, yet very real, things that she insists are pretend.
This morning's examples are "the Taj Mahal" and "bats."
The world is full of wonders.
This morning's examples are "the Taj Mahal" and "bats."
The world is full of wonders.
June 13, 2025 at 8:40 PM
In the four-year-old's process of sorting out "real" from "pretend," my favorite part is discovering the fantastical, yet very real, things that she insists are pretend.
This morning's examples are "the Taj Mahal" and "bats."
The world is full of wonders.
This morning's examples are "the Taj Mahal" and "bats."
The world is full of wonders.
Reposted by Matt Zefferman
“The gender-equality paradox entails that countries with more equality have larger gender differences in behavior. Positive associations between equality & gender differences disappear or reverse when accounting for cultural clusters of countries. The paradox appears to be a methodological artifact”
Simpson’s gender-equality paradox | PNAS
Several cross-country examinations have found larger gender differences in Western
countries. More recently, it has been argued, from an evolutiona...
www.pnas.org
June 11, 2025 at 5:33 AM
“The gender-equality paradox entails that countries with more equality have larger gender differences in behavior. Positive associations between equality & gender differences disappear or reverse when accounting for cultural clusters of countries. The paradox appears to be a methodological artifact”
Reposted by Matt Zefferman
New 2 year post-doc position on cultural adaptation in complex environmental management.
Post-Doctoral Position on Cultural Adaptation in Forest Management
The University of Maine seeks a post-doctoral researcher for a two-year project on human cultural adaptation in forest management, requiring a strong quantitative background and experience in modeling.
timwaring.info
June 9, 2025 at 3:35 PM
New 2 year post-doc position on cultural adaptation in complex environmental management.
Grammerly is now telling me to replace all instances of "use" with "utilize."
It used to be pretty good at finding typos, now it is trying to get me to write like a stochastic Internet parrot, I guess.
It used to be pretty good at finding typos, now it is trying to get me to write like a stochastic Internet parrot, I guess.
June 8, 2025 at 9:59 PM
Grammerly is now telling me to replace all instances of "use" with "utilize."
It used to be pretty good at finding typos, now it is trying to get me to write like a stochastic Internet parrot, I guess.
It used to be pretty good at finding typos, now it is trying to get me to write like a stochastic Internet parrot, I guess.
Our in-house journal (Countering Threats Exchange) has a call for a special issue on Unidentified Anomalous Phenomena. They are looking for submissions from various angles.
nps.edu/web/ccht/con...
nps.edu/web/ccht/con...
Contribute to Upcoming Publication - Center on Combating Hybrid Threats - Naval Postgraduate School
nps.edu
June 4, 2025 at 5:23 PM
Our in-house journal (Countering Threats Exchange) has a call for a special issue on Unidentified Anomalous Phenomena. They are looking for submissions from various angles.
nps.edu/web/ccht/con...
nps.edu/web/ccht/con...
I'm presenting virtually at HBES this year. Thanks to @humbehevosoc.bsky.social for adding us stranded folks to the program. You can see our talks online or come to ACA 314 on Friday at 5:50 EST / 2:50 PST to see us on the big screen.
June 3, 2025 at 9:26 PM
I'm presenting virtually at HBES this year. Thanks to @humbehevosoc.bsky.social for adding us stranded folks to the program. You can see our talks online or come to ACA 314 on Friday at 5:50 EST / 2:50 PST to see us on the big screen.
Does anyone have an accessible guide for best practices in presenting DAGs and associated regression estimates in a paper?
This is for a student thesis project (and, frankly, for myself, as I have done it admittedly ad hoc, although no reviewers have so far complained.)
This is for a student thesis project (and, frankly, for myself, as I have done it admittedly ad hoc, although no reviewers have so far complained.)
June 3, 2025 at 5:22 PM
Does anyone have an accessible guide for best practices in presenting DAGs and associated regression estimates in a paper?
This is for a student thesis project (and, frankly, for myself, as I have done it admittedly ad hoc, although no reviewers have so far complained.)
This is for a student thesis project (and, frankly, for myself, as I have done it admittedly ad hoc, although no reviewers have so far complained.)
Reposted by Matt Zefferman
Our #MBARIFromTheVault series comes to a close with an animal that needs no introduction. MBARI researchers last encountered the giant phantom jelly in November 2021 during a dive in Monterey Bay. While exploring the midnight zone, this ghostly giant emerged from the darkness.
May 12, 2025 at 6:19 PM
Our #MBARIFromTheVault series comes to a close with an animal that needs no introduction. MBARI researchers last encountered the giant phantom jelly in November 2021 during a dive in Monterey Bay. While exploring the midnight zone, this ghostly giant emerged from the darkness.
Reposted by Matt Zefferman
To reiterate one of my academic themes for 2025: It’s a real problem that students feel incapable of operating without genAI, particularly given that they do worse with it. I don’t care if slackers slack, but if insecurity x manipulation shields well-meaning students from their education, I care.
Every commercial for AI is basically “look, you’re a fucking moron, but we’ll help you fake it!” and it feels like they really do get their target audience.
January 27, 2025 at 8:02 AM
To reiterate one of my academic themes for 2025: It’s a real problem that students feel incapable of operating without genAI, particularly given that they do worse with it. I don’t care if slackers slack, but if insecurity x manipulation shields well-meaning students from their education, I care.
Congrats to Rob Boyd, @scottepage.bsky.social, @alisongopnik.bsky.social, and others on election to NAS!
April 30, 2025 at 6:08 AM
Congrats to Rob Boyd, @scottepage.bsky.social, @alisongopnik.bsky.social, and others on election to NAS!
Reposted by Matt Zefferman
LLMs are nothing more than models of the distribution of the word forms in their training data, with weights modified by post-training to produce somewhat different distributions. Unless your use case requires a model of a distribution of word forms in text, indeed, they suck and aren't useful.
There are a lot of critiques of LLMs that I agree with but "they suck and aren't useful" doesn't really hold water.
I understand people not using them because of social, economic, and environmental concerns. And I also understand people using them because they can be very useful.
Thoughts?
I understand people not using them because of social, economic, and environmental concerns. And I also understand people using them because they can be very useful.
Thoughts?
April 24, 2025 at 4:54 PM
LLMs are nothing more than models of the distribution of the word forms in their training data, with weights modified by post-training to produce somewhat different distributions. Unless your use case requires a model of a distribution of word forms in text, indeed, they suck and aren't useful.