Home Creators Posts Import Register Favorites Logout

Content

Hi everyone, got a few things coming up:

I'm planning to shoot the next video next week and was thinking of recording an Office Hours at the same time. I have some questions collected already that people have already given me, but if you have one on your mind you'd like to add, please leave it in a comment below (reminder that this is for questions about actual things related to my content and not, e.g. what is my favorite food, thanksss)

And just as a reminder, book club for Pedagogy of the Oppressed starts Monday June 16 (meaning I'll post the reading schedule/pre-reading discussion that day, you don't have to have started before then)

Looking forward to reading with yall!

fatima

Comments

Ryan Freeman

Hello Dr. Fatima, I really enjoyed your latest video essay on Galileo and the scientific method. While watching, I kept expecting you to bring up Bayesian epistemology, and I was surprised when it never came up. It felt like it would have been a perfect fit for your discussion. Your criticism of the Popperian approach really resonated with me, especially the points about "falsifying" theories and how rigidly dividing beliefs into true/false categories can be limiting. It made me think about how Bayesian approaches handle this differently. Instead of simply disproving or failing to disprove theories, Bayesian methods let us calculate likelihood ratios, comparing how probable our evidence is under different theories. These ratios can be combined across experiments and updated with new evidence, allowing our beliefs to evolve more naturally. Different individuals can start with different priors and get different posteriors, but as more evidence accumulates, their beliefs converge. What I find elegant about this approach is how it addresses two of the three "commandments" you mentioned Galileo breaking. From a Bayesian viewpoint, hypotheses don't have to be grounded in established knowledge because you can start with any reasonable prior and let the evidence update your beliefs. Similarly, "falsified" theories don't have to be permanently discarded if new evidence appears that only they predicted - you can simply update your probability assessments. This framework seems to align well with your critique of overly rigid scientific methods, though I'm curious if I'm missing something important about Feyerabend's perspective that you discussed. I'd love to hear your thoughts on Bayesian epistemology and whether you considered including it in the video. Thanks for the thought-provoking content. Keep up the great work!

Javier F

I gotta know, what's your favorite ice cream? (Just kidding!) Nah but for real, what's your take on pop science news articles where they announce stuff like, "scientists invent crazy new material that will turn t-shirts into solar panels!!" ? There's always something that sounds exciting but then is never heard of again. Does astronomy also deal with hype articles like this?

AJ Biffl

You mentioned in the DOGE stream (I think) that you consider your work on physics education research as distinct from the channel, but I wanted to ask if there is any possible overlap between the two in the future, either videos about things your group has found or conversations with people you've worked with, so on. At a more philosophical level, how do you relate these roles of being a YT creator and being a physics education researcher? Obviously there's the element of making physics and science more accessible in general (which I think you've also talked about) but how do you relatively value building conceptual accessibility for the general public in a public forum (as a creator) vs helping build direct material accessibility for people already within or adjacent to academia from inside of it (as a researcher)? Would you characterize these roles differently?

Wami Ogunbi

This is a fascinating perspective. I am not familiar with Bayesian epistemology, but a thought that came to mind as I was reading your explanation was: how do you define "reasonable" prior? If this is a requirement for the Bayesian approach, this strikes me as being a limitation. Afterall, we don't know what we don't know, and something that seems "unreasonable" now may prove to be "reasonable" if we change our perspective. While the approach seems to try to correct for this by allowing the morphing of ideas based on new evidence, it still has an over reliance of the current state of what is known and the methodologies by which we find evidence. But again, this is outside of my field of expertise so I would be curious to hear yours or Dr. Fatima's thoughts on this!

Kevin Hengehold

What if Galileo was wrong? How many counter induction practitioners have been forgotten because they were just blowhards?

Wami Ogunbi

I know there might be other questions in the queue, so feel free to bump this to later office hours if needed. My question is: what’s your take on the “trust the science” rhetoric? This phrase became especially prominent during the pandemic, initially encouraging openness to scientific guidance. But over time, it seemed to spark resentment and distrust—perhaps because many people expected science to deliver clear, unchanging answers, when in reality, it’s an iterative, sometimes messy process of discovery and revision. How do you think scientists should navigate this tension between public expectations and the true nature of scientific work—especially in the context of your work on education? As a professor, I often find myself in a similar position. For example, I have to deliver hard truths to my students, like the fact that coding isn’t always straightforward. It takes time, persistence, and often a shift in perspective to identify bugs—and sometimes the best way forward is to rewrite the code entirely. Helping students embrace this messiness can be challenging, so I’d love to hear your thoughts on how we can better prepare people—whether students or the broader public—for the reality of how science (and learning) actually works.

AJ Biffl

This is very cool! It definitely sounds much more philosophically in line with Feyerabend's position than Popper, though I viscerally react against this very buttoned-up-sounding idea of objective Likelihood (also code I'm working on for maximizing a binned likelihood function is not working great right now so I'm a tad salty). What you describe rejects objective falsification, but it IS a supposedly objective method for assessing competing theories, which whether or not it's rationally justifiable (it undoubtedly is) will end up being restrictive and will prioritize its own internal "rational" compass over any other considerations (democracy, sustainability, humanity)

AJ Biffl

Wait this is such a good question! I would love a thoughtful recap of the pandemic madness from a science-communication perspective

Brendon Elmore

Any chance you’ll discuss Israel’s unjust strikes on Iran, I feel like you’d have unique and compelling insights on it since so few people understand the history of the region as well as you do

Victor Steuck

Hi Dr. Fatima! Hi everyone! This is a bit of a playful question, but also a curious one. As a former astronomer, what’s your relationship to astrology, if any? I’m a Gemini (my birthday's is tomorrow, on June 14 🎉), so I was wondering. In Brazil recently, there was a bit of controversy when a Public University planned to include a single lecture on astrology (given by an astrologer and mediated by an anthropologist) in a nursing course. Some people got upset, saying it was trying to put astrology on the same level as astronomy. There was even a campaign against it, and in the end the lecture was canceled. So I’m curious... as an astrophysicist, do you see astrology as nonsense, or do you think it has symbolic or cultural value? Do you get irritated when people ask about your sign, or do you secretly enjoy it? Just asking for fun (and maybe a little provocation). Thanks!

Ryan Freeman

@Wami Ogunbi That's such a good question! It makes me really happy to see people thoughtfully engaging with my comment. The idea of 'reasonable' priors is something that people spend a lot of time debating. First of all, if someone has already seen evidence that isn't accounted for in the likelihood function, that evidence is often incorporated into their prior. For the sake of this conversation though, let's assume we're talking about priors formed before seeing any relevant evidence. So basically, priors can be thought of as predictions about the world. If someone claims their prior ratio is 1:99, they're essentially saying they could make 100 similar predictions and expect to be right 99 times. Given that we're talking about predictions made before seeing any evidence, that level of confidence is pretty hard to believe - and it would be even harder to believe someone claiming a prior ratio of 1:999 or 1:9999. One occasion where unequal priors make sense is when comparing hypotheses of different complexity. Due to Occam's razor, we might reasonably favor simpler explanations. So for instance, I think it would make sense to give heliocentrism and geocentrism roughly even odds, but epicycles lower odds due to their added complexity. There have been attempts to quantify this using measures like Kolmogorov complexity, though in general I think it comes down to trusting your gut and letting others trust theirs. If two people choose priors that genuinely feel right to them and still disagree after seeing evidence, I think they should go out and get more evidence rather than arguing about whose prior was better. I once heard an anecdote about someone being told there was a 1:10^50 ratio of evidence against their position, and they responded that it didn't matter because their prior was 10^100:1 in favor of it. I think most people would recognize such extreme confidence as unreasonable, but in my experience, if people bother to quantify their beliefs at all, their priors are pretty reasonable.

32 Emilia 32

happy birthday, just came back from pride protest in vienna :))

Victor Steuck

Thank you! Nice to meet you, Emilia!

Wami Ogunbi

Hi Ryan, Thank you for such a thoughtful and generous response to my question. I’ll be honest—some of it went over my head at first! I had to do a bit of Googling to unpack some of the technical terms you used, so I appreciate the opportunity to learn something new. That said, forgive me if this follow-up question touches on something you already addressed. From what I understand, the process of assigning priors—whether grounded in past evidence or intuition—still seems to rely heavily on the imagination (and limitations) of the person making the prediction. Even with mechanisms to adjust based on new data, isn’t the initial framing still constrained by the scope of what someone can conceive? In that way, I’m still wondering what makes Bayesian epistemology fundamentally different from the traditional scientific method. Both approaches seem to evolve theories over time based on evidence—just through different lenses or computational frameworks. So maybe my question is: do you see the Bayesian approach as offering something qualitatively new to how we understand and revise scientific knowledge, or is it more of a refined tool within the same broader philosophy of evidence-based reasoning? Thanks again for the stimulating conversation—this has been really fascinating to think through!

Not a Lady Maria, She-Bear of the Global South

Hi Victor! I'm also a Brazilian, but I'll sitck to writing in english so that everyone can follow the conversation. First, I want to say a appreciate your question and hope to see it answered! Aslo, as it is realted to Dr. Fatima last video, I wanna recommend the podcast Vinte Mil Leguas, their last season was all about Galileo and they have some interesting insights about his own relationship with astology. On another note, I have a lot of interest in university curriculum and in the presence (or absence) of the so called "pseudo-sciences" in higher education. When I was in resicency, I had lectures in several alternative health practices ("práticas integrativas e complementares") and it was a mixed bag. I felt very uncomfortable during most of them, but some proved quite valuable as well. There are diffent ways of going about it, I think. How you approach and teach such subjects is integral to their belonging -- or not -- in the curriculum.

Not a Lady Maria, She-Bear of the Global South

This conversation is so interesting and enriching. I want to acknowledge all the contributions so far as thought-provoking and challenging! @Rayan Freeman I wanna share that while watching, I also expected Bayes to show up at some point in the video essay! @Wami Ogunbi your questions are so rich and provoking. I started to code about a year ago and I want you for a teacher. Now for my contribution: I've never worked with Bayesian statistics, but I do want to add that I feel like calling what Bayes proposes an "epistemology" might be a stretch. I know it is reffered as such, but I think we are working with different definitions of the word "epistemology". As far as I have dipped into these waters (very little, mind you) Bayesian statistics are an alternative to frequentist statistics. They are radically different in how they demonstrate evidence for or against a hypothesis, and I'll agree that shifting from a frequentist perspective to a Bayesian requires some rethinking of what we understand as evidence, demonstration and proof. In the health sciences, Bayesian reasoning introduced some interesting concepts and provided new ways forward for scientists and practitioners. Bayesian re-analyses of clinical trials and observational data have contributed to important advancements in the field. That said, at the bottom of it, I feel like it's a different methodology which serves the same ultimate goal. I struggle to think of a "Bayesian epistemology" as something radically different when it comes to thinking about "science" as a whole (if such a thing even exists). The "point" of Bayesian analyses and frequentist analyses is not that dissimilar. I’d argue that the kinds of questions Feyerabend poses go deeper than a mere shift from frequentism to Bayesian methods.

Not a Lady Maria, She-Bear of the Global South

I feel like health sciences, because they are by nature so prescriptive (and therefore have a potential for greatly disrupting people's lives), require a special case of "trust". It is one thing to "trust" that the Earth is round when I feel it flat under my feet, or that galaxies host solar systems and planets I've never seen, or that evolution is a natural process that acounts for the diversity of life I see today and also for the fact that the life-forms I see today are not all life-forms that ever existed. These are things one can commit to believing without changing one thing about their lives. It is a different thing to trust that certain behaviour will lead to disease, and therefore should be avoided. To trust that certain substances should be introduced into my body to prevent illness. To trust eating eggs (doesn't) raise my cholesterol. Because to trust such sciece is to allow it to dictate one's choices, and that is very scary is some contexts.

Not a Lady Maria, She-Bear of the Global South

Hi Dr. Fatima! I'm a long-time follower and a new Patreon. I want to ask your opinion on physics possibly extending its scope to other disciplines -- if it even is that what I'm witnessing, it might not be. I'm not sure, because I'm certainly not a physicist. I remember a while ago, I think in 2023, I got a small snippet in the Nature Briefing newsletter about a "theory of life" arising from physics, and it immediately caught my interest. The thing is, I can only read the dumbed-down version of the work being done, and muse about the philosophical questions it supposedly poses. But I don't understand the work itself, and I was wondering if you might be able to comment on it? I know you are an *astro*physicist, and I don't know how much specialization might hinder you from appreciating this work from a technical standpoint. But I'd love to hear your thoughts on it, especially your considerations on how physics can be interdisciplinary and how it benefits (or doesn't benefit) other sciences and society as a whole. I think it's worth interrogating the challenge of being interdisciplinary without colonizing other disciplines with our field's ways of thinking. Personally, I found the way the assembly theory researchers presented their findings irksome -- exactly this: "From the first sentence of the abstract, the paper seems to imply that the authors have cracked a foundational problem for biology: ‘Scientists have grappled with reconciling biological evolution with the immutable laws of the Universe defined by physics.’ No we haven’t, biologists respond – we have never found the slightest contradiction between them". Quoted from this source: https://www.chemistryworld.com/opinion/will-the-assembly-theory-imbroglio-do-anything-for-evolution/4018232.article . On the other hand, proponents of assembly theory raise interesting questions, such as: "Modern physics can explain everything from the spin of the tiniest particle to the behaviour of entire galaxy clusters. But it can’t explain life. There’s simply no formula to explain the difference between a living lump of matter and a dead one. Life seems to just mysteriously “emerge” from non-living parts, such as elementary particles." Quoted from this source: https://theconversation.com/physics-has-long-failed-to-explain-life-but-were-testing-a-groundbreaking-new-theory-in-the-lab-215636 My question is: should physics try to explain life? Is it relevant for the field of physics that we are able to differentiate between a living lump of matter from a non-living one -- using a formula? After all, we can certainly tell them apart already, using tools from other scientific disciplines (also how would once-living now-dead things fit into this formula-differentiated model?). Mostly what I think I want to know (and sorry for being so confusing, I'm actually confused): how do you feel about physics contributing to questions such as the emergence of life?

Victor Steuck

Hi Not a Lady Maria, thank you for your thoughtful comment and for the podcast recommendation, I’ll definitely check Vinte Mil Léguas! The topic sounds really interesting. Personally, I find astrology to be a fascinating form of culture. I've actually written quite a bit about it. I see astrology, along with things like tarot, mysticism, and even the so-called “pseudo-sciences” you mentioned, as something that holds value in people’s lives. It’s a form of knowledge and playful practice, much like religion. And even though it’s playful, it’s also taken seriously by many. It's hard do explain. I tend to believe that nothing is truly useless, and that even what seems useless has value, at least symbolically. After reading A Vida Não É Útil by Ailton Krenak, I started seeing many things in a different light. I’ve been writing a lot about this on my blog, The Brazilian Mutt, and conversations like this help to me to get inspired. So thanks again for the great input and the suggestion!

Not a Lady Maria, She-Bear of the Global South

My oh my that name. I'm actually called Maria but this penname was originally for another Patreon I subscribe to which is a re-reading of a fantasy book series called 'NotACast', so it seemed fitting. I feel a little silly now, specially because I had not realized the name carried between patronages 😅

Dr. Fatima

hi, i recorded the office hours just before you posted this, so I'll add "how do you feel about physics contributing to questions such as the emergence of life?" to the list for the next one (i generally don't do readings for q&a's so let me know if you'd rather i summarized your question differently)