iiinova — where your digital self keeps evolving

Paul Duffell @paulduffell.bsky.social · 28/09/2026

Oh I plan on being a total wreck when that day comes for my daughter.

Paul Duffell @paulduffell.bsky.social · 19/09/2026

It's still too early to say, I've only done a few of them. But at least in the moment, I can tell some students are getting a reality check as to what they have learned.

010

Paul Duffell @paulduffell.bsky.social · 17/09/2026

I started doing an *ungraded* 5-minute quiz the day after homework is due, with 1-2 really simplified questions. The semester is not over yet, but I am hoping just giving the students the awareness of whether they had learned the material is enough to motivate them to do it themselves.

110

Paul Duffell @paulduffell.bsky.social · 15/09/2026

Absolutely! We are definitely planning to do a Rise_Time 2028 and I will be sure to contact you about it when we have dates figured out!

000

Paul Duffell @paulduffell.bsky.social · 14/09/2026

My department head encourages us to use google scholar's h-index, since that tends to be larger than other sources.

130

Paul Duffell @paulduffell.bsky.social · 14/09/2026

h-index is BS, so just pick whichever measure is the biggest.

220

Paul Duffell @paulduffell.bsky.social · 25/08/2026

But I am interested in hearing how oral homework goes.

020

Paul Duffell @paulduffell.bsky.social · 25/08/2026

This semester I have 44. And I'm thinking for classes that size and larger. When I teach a graduate course I am not as worried; grad students will learn what they want to learn, as usual.

020

Paul Duffell @paulduffell.bsky.social · 25/08/2026

Preparing my first day of teaching, realizing my grading system (most of the grade is homework) is based on antiquated (3-year-old) assumptions about how much students learn from homework in this new era of AI. I don't want to fall back on exams being most of their grade; what else can I do? 🎢

151

Paul Duffell @paulduffell.bsky.social · 18/08/2026

It's a skeleton crew over there. They can barely get money out to the handful of people they are still paying.

120

Paul Duffell @paulduffell.bsky.social · 17/08/2026

Oh god it's been awhile since I taught a lab course but that phrase "human error" still triggers me!

021

Paul Duffell @paulduffell.bsky.social · 02/07/2026

I just checked. A minute is indeed now 60 seconds long, so that checks out.

030

Paul Duffell @paulduffell.bsky.social · 29/06/2026

That depends; do you do science or do you develop marketable tools for industry and/or the military?

010

Paul Duffell @paulduffell.bsky.social · 17/06/2026

It's also not developing "exponentially". Chatbots are only slightly more useful now than they were a year or two ago, we haven't seen exponential growth (though spending might have increased exponentially).

031

Paul Duffell @paulduffell.bsky.social · 15/06/2026

I still do my first pass in ink (I actually keep special green pens for this purpose). My students always brace themselves for my first edit filled with pages of green ink.

110

Paul Duffell @paulduffell.bsky.social · 15/06/2026

Just paint over all the hideous gold feathering that I assume is going to cover everything.

000

Paul Duffell @paulduffell.bsky.social · 09/06/2026

Gotta wait for that bubble to pop, then it will become super cheap.

120

Paul Duffell @paulduffell.bsky.social · 08/06/2026

My daughter has recently learned to do a hand motion that means "come closer I want to give you a slobbery kiss" and every time she does the hand motion I feel like I've won an academy award.

010

Paul Duffell @paulduffell.bsky.social · 07/06/2026

My graduate student Miranda Pikus was featured on the AAS Journal Author Series! We were interviewed by editor Frank Timmes about our paper, check it out: www.youtube.com/watch?v=IDTs... 🧪🔭

youtube.com

AAS Journal Author Series: Miranda Pikus and Paul Duffell on 2026ApJ..1002..104P

YouTube video by AAS

081

Paul Duffell @paulduffell.bsky.social · 18/05/2026

Honestly, a one-year ban from arxiv isn't the end of the world for a scientist. And it's a good lesson to learn, even if it's the lesson of "be mindful of what you put your name on".

110

Paul Duffell @paulduffell.bsky.social · 09/05/2026

My graduate student Miranda Pikus just published a paper on supernova remnants. She did 3D simulations of supernova remnants with fast cooling and was able to reproduce the highly unusual "spiky" shape of the supernova remnant Pa 30 (SN 1181 AD) iopscience.iop.org/article/10.3...

Supernova Remnant Pa 30, over 800 years after the original supernova of 1181 AD. Left side is the observed remnant, right side is a 3D model generated by Miranda Pikus at Purdue. Full paper here: https://iopscience.iop.org/article/10.3847/1538-4357/ae57ac/pdf

140

Reposted by Paul Duffell

👻 Saavik ‘Spooky Season’ Ford 🎃 @saavikford.bsky.social · 06/05/2026

Keep me from posting this to the chat in our faculty zoom meeting… please…

Should you ask a question at a faculty meeting?

Flow chart—essentially the answer is ALWAYS ‘NO’ (if someone is actively dying… you should not ask a a question but call 911…)

12759

Paul Duffell @paulduffell.bsky.social · 04/05/2026

My student Miranda Pikus just had her paper accepted to ApJ! arxiv.org/abs/2511.08519 She developed a suite of detailed 3D simulations demonstrating how the unusual "spiky" shape of the supernova remnant Pa 30 could be naturally generated via a fluid instability in a rapidly cooling remnant. 🔭🧪

The supernova remnant Pa 30, caused by a supernova that went off in 1181 AD. The left side shows the highly unusual sulfur-rich filaments in the observed remnant. The right side is Miranda's theoretical model, providing an explanation for these filaments via Rayleigh-Taylor instability with fast cooling.

0192

Paul Duffell @paulduffell.bsky.social · 01/05/2026

Study finds that students who cheated on their homework do worse on exams.

010

Paul Duffell @paulduffell.bsky.social · 19/04/2026

We have had an informal policy of allowing an invited speaker to defer to our next conference if they recently gave birth. But we don't rescind the invitation; we just assume it will likely be turned down and offer the ability to shift to next time.

110

Paul Duffell @paulduffell.bsky.social · 16/04/2026

I'm pretty sure that an essential component of most art is that a person did it with some kind of intention. I can't imagine people having the same interest in a book or song if it wasn't made by anybody. But time may prove me wrong.

010

Paul Duffell @paulduffell.bsky.social · 15/04/2026

Congratulations to my student Ranadeep Dastidar for a successful defense of his thesis!

Paul Duffell (sitting, left) with Ranadeep Dastidar (standing, right) at the beginning of his thesis defense, titled "Investigating Observational Signatures for Non-thermal Radiation Processes in Gamma Ray Bursts"

040

Paul Duffell @paulduffell.bsky.social · 13/04/2026

I want a phone that is impervious to all the elements of the periodic table.

010

Paul Duffell @paulduffell.bsky.social · 07/04/2026

I don't mind xi, it's zeta that's the awful one in my opinion.

000

Paul Duffell @paulduffell.bsky.social · 02/04/2026

It's ok if I use the code to train some AI right?

010

Paul Duffell @paulduffell.bsky.social · 01/04/2026

I don't understand how anyone is comfortable doing that next to strangers. When I'm seated next to a stranger I'm neurotically terrified that I'll accidentally touch them.

010

Paul Duffell @paulduffell.bsky.social · 31/03/2026

lol that essay was probably written by a LLM anyway.

000

Paul Duffell @paulduffell.bsky.social · 22/03/2026

I was rejected by every grad school I applied to, 3 times. Finally, one year, one institution (NYU) took a chance on me. Now I'm a professor of Physics & Astronomy at Purdue.

0131

Paul Duffell @paulduffell.bsky.social · 18/03/2026

After reading a bunch of proposals, I would like to say that somewhere (ideally early on) you should say what you're actually going to do. I've read an astonishing number of proposals that read like "we're great at doing all of this science, please give us money".

010

Paul Duffell @paulduffell.bsky.social · 11/03/2026

I always think errors result in the coolest images. Definitely true in my code.

120

Paul Duffell @paulduffell.bsky.social · 02/03/2026

Seeing as how we find ourselves awake in the middle of the night for various reasons, we might try to check it out ourselves!

010

Paul Duffell @paulduffell.bsky.social · 27/02/2026

Arguably papers like that are easier to get published, because they're not making any controversial statements that can upset a referee.

100

Paul Duffell @paulduffell.bsky.social · 24/02/2026

Near the halfway-point in my GR course, I am officially done with "Kerr week". That's a rough metric but we all got through it together! 🎢

030

Paul Duffell @paulduffell.bsky.social · 24/02/2026

I think there are some people who code for a living but don't actually like doing it. I think if you find yourself wanting to offload all of your job to AI, you are probably not doing what you enjoy, and maybe should find another line of work.

141

Paul Duffell @paulduffell.bsky.social · 20/02/2026

I don't hate posters, but I do hate that they are now treated like a high-pressure job interview for junior scientists. Also when a student has a polished presentation on every section and they won't let me skip ahead or ask questions about one plot until they've finished their full presentation.

120

Paul Duffell @paulduffell.bsky.social · 17/02/2026

Of course, I do all of these tasks every time. For what is parenting if not capitulating to every single one of your daughter's whims?

000

Paul Duffell @paulduffell.bsky.social · 17/02/2026

My afternoon routine with my daugther (who only can say a few words): First, she demands that we each put on lei's, then I need to wear a baseball cap made to look like a dinosaur, then she needs to hold a small statue of an owl, then I need to put on an oven mitt. Every time we come home.

120

Paul Duffell @paulduffell.bsky.social · 10/02/2026

Very surreal to see the names of people I know personally in these files, even when it's just things like this shortlist.

160

Paul Duffell @paulduffell.bsky.social · 08/02/2026

Every prospective postdoc I talk to is experiencing the same thing. It's not over til the Hubble cascade this month, but things are not looking good out there.

040

Paul Duffell @paulduffell.bsky.social · 05/02/2026

Love these results. I've already sent Rana to build off-axis models with predictions for sky images. We will email you once we are further along!

110

Paul Duffell @paulduffell.bsky.social · 04/02/2026

Halfway through feeding my baby, she starts pointing ominously at the clock, making me feel like I'm forgetting an appointment or something. But no, she just really likes clocks.

010