Steuard

I can't say what the reality looks like today, but when I was in college 25 years ago (Harvey Mudd), closed-book take-home exams were pretty common (often with a specific time limit: maybe 3/5/8 hours), and I would have been shocked and aghast at the thought of anyone cheating. We took our Honor Code deeply seriously. (I've given my own students take-home open-book exams with "no outside resources" rules regularly in the past, but I've pretty much concluded that isn't viable anymore. But then, we don't have a formal Honor Code here.)

One particularly striking example was a timed, closed-book math exam where the exam paper just listed four problem numbers from the book. We were expected to open the book, copy down those problems without looking at anything else, and then put the book away. Honestly, that one felt like just asking for trouble and now that I'm a professor myself I think that the prof was unreasonably lazy, but it certainly captured the spirit of the thing. (Technically, that wasn't a Harvey Mudd professor but rather a professor at the adjacent graduate university. Maybe that had something to do with it.)

Do you by chance have references to those conceptually cleaner bivector publications? I've spent a good bit of time looking for other people working in that space, and haven't found much other than an article by Jancewicz from 1980. (I like to imagine that my articles these past few years have been "conceptually clean", but they certainly aren't "previous" to much GA work.)

[Possible obscure spoiler]

A friend once showed me a post on rec.games.roguelike.nethack where someone was finally begging for a hint because they'd gone deep in the dungeon and couldn't figure out anything to do next. They couldn't find any staircases down, though they had found a weird vibrating square, and none of the many weird items they'd collected seemed to do anything to help.

If you come up with a process to do that efficiently, the helium will be a lovely bonus but not remotely the most important result. :D

In principle, flagrant abuse of the pardon power is blocked by Congress's ability to impeach and remove a President who engages in such abuse.

In practice, that has always been an ineffective threat against Presidents who are within days of leaving office anyway. And more importantly, the framers of the Constitution seemed to have entirely failed to imagine a party like today's Republicans who value strict personal loyalty to the President over every other principle of government.

All I have to say is that if one of my students turned in those curves as "best fits" to that data, I'd hand the paper back for a re-do. Those are garbage fits. To my eye, none of the very noisy data sets shown in the graph show clear enough trends to support one model over any other: are any of those hyperbolic curves convincingly better than even a linear fit? (No.) The "copilot code share" data can't possibly be described by a hyperbolic curve, because by definition it can't ever go over 100%. (A sigmoidal model might be plausible.) And even if you want to insist on a model that diverges at finite time, why fit 1/(t0-t) rather than 1/(t0-t)^2, or tan(t-t0), or anything else?

The author does in fact note that only the arXiv data fits this curve better than a line, and yeah: that's the one dataset that genuinely looks a little curved. But 1) it's a very noisy sort of curved, and 2) I'll bet it would fit a quadratic or an exponential or, heck, a sine function just as well. Introducing their process of doing the hyperbolic fit, they say, "The procedure is straightforward, which should concern you." And yeah, it does concern me: why does the author think that their standard-but-oversimplified attempt to fit a hand-chosen function to this mess is worth talking about? (And why put all of that analysis in the article, complete with fancy animated graph, when they knew that even their most determined attempt to find a signal failed to produce even a marginally supportive result 80% of the time?)

In short: none of the mathematical arguments used here to lead in to the article's discussion of "The Singularity" are worth listening to at all. They're pseudo-technical window dressing, meant to lend an undeserved air of rigor to whatever follows. So why should we pay attention to any of it?

How old are you? Because I promise you, that description was pretty much spot-on for most shows through most of the history of TV prior to the late 1990s. My memory is that the main exception was daytime soap operas, which did expect viewers to watch pretty much daily. (I recall a conversation explaining Babylon 5's ongoing plot arc to my parents, and one of them said, "You mean, sort of like a soap opera?") Those "Previously on ___" intro segments were quite rare (and usually a sign that you were in the middle of some Very Special 2-part story, as described in the previous comment).

Go back and watch any two episodes (maybe not the season finale) from the same season of Star Trek TOS or TNG, or Cheers, or MASH, or Friends, or any other prime time show at all prior to 1990. You won't be able to tell which came first, certainly not in any obvious way. (Networks didn't really even have the concept of specific episode orders in that era. Again looking back to Babylon 5 which was a pioneer in the "ongoing plot arc" space, the network deliberately shuffled around the order of a number of first-season episodes because they wanted to put stronger stories earlier to hook viewers, even though doing so left some character development a bit nonsensical. You can find websites today where fans debate whether it's best to watch the show in release order or production order or something else.)

By and large, we all just understood that "nothing ever happens" with long-term impact on a show, except maybe from season to season. (I think I even remember the standard "end of episode reset" being referenced in a comedy show as a breaking-the-fourth-wall joke.) Yes, you'd get character development in a particular episode, but it was more about the audience understanding the character better than about immediate, noticeable changes to their life and behavior. At best, the character beats from one season would add up to a meaningful change in the next season. At least that's my memory of how it tended to go. Maybe there were exceptions! But this really was the norm.

The article feels like LLM output, too. (And they don't actually credit an author in the byline.) Is there another source out there that this was based on? Can we read that instead, and skip the extra layer of interpretation/distortion?

I'm a physics professor who regularly teaches about special relativity in my Modern Physics course. I've made a web app for drawing spacetime diagrams (technically, two-observer Minkowski diagrams), which are one of the best ways I know for building intuition about how relativity works. The link points to an introduction to the diagrams, including a brief explanation of some key relativity concepts based on diagram illustrations. (It's meant to be at least halfway understandable to people who haven't studied physics before, though it'll be clearer the more you already know.)

Read through the linked page if you want the basics, or if you're eager to just jump straight in, follow the links to use the main app and play with that. (It has multiple predefined scenarios that you can load, each with a brief explanation, but you can design your own scenarios as well.)

[Aside: I feel really good about the UI I've got for this so far, but my last significant JavaScript work before this project was back in 2005 or so. I've had to learn a LOT.]

Ah, I see what you're getting at. My instinct here is that (exactly as you've pointed out) fields like E and B will fall off like 1/r instead of 1/r^2, but that all of the qualitative behavior will be basically the same. So I wouldn't trust this simulation to predict the precise behavior of a real circuit (even one whose shape was basically planar), but I suspect that it will behave more or less right.

Looking at the examples, it seems like you can make 1D and 2D strings/grids of resistors here in much the same way you would in a 3D model; you just can't make a 3D grid (or non-planar circuits). My general experience working with and teaching basic circuits is that it's rare that we consider current flow in a genuinely 3D medium: the vast majority of problem-solving examples approximate wires as simple 1D paths for charge to follow, and more careful treatments that talk about where charges accumulate to guide current flow around corners, etc. still almost always illustrate their points in 2D diagrams/examples.

So my impression is that this simulation is likely to give a pretty solid qualitative sense of how these systems work, despite its 2D framing.

I don't pretend to know what this simulation is doing, but for the record, electromagnetism works just fine in 2D. You might be thinking "but magnetic fields are intimately tied to cross products, which only work in three dimensions." But you can set up the equations of electromagnetism just fine either using differential forms or bivector magnetism (https://arxiv.org/abs/2309.02548), and it works in any dimension you'd like. (The cross product version is really a narrow and sometimes misleading special case.)

Possibly related: there are options to "View B" and "View H" in the scalar dropdown, not in the vector one. That may be closely related to the fact that in two dimensions, the magnetic field has just a single component. Whether you describe is as a 2-form or a bivector, the magnetic field is an antisymmetric rank-2 tensor: an antisymmetric matrix. In 3D, that means 3 independent components, and there's a one-to-one mapping to vectors (more or less). But in 2D, an antisymmetric matrix has just one independent component. (And in 4D, it's got six: this is precisely the relativistic electromagnetic field tensor, that in 3D splits into an electric part and a magnetic part. My paper has more details.)

I think that the previous poster's point is that historical photographs are not in-scope to be added to this project: for example, this project will never include the first known photo of a living platypus (or a living cat, as noted), because such photos existed before this project began. The project collects photos posted to iNaturalist that meet the specified criteria.

It's a cool collection of modern observations of rare or remote species! But the title could also describe an entirely different research project, focused on historical media rather than modern exploration. That could also be very cool.

Is there a place to submit an issue with the data? The final map makes it look like a bunch of properties around the neighborhood were purchased by the University between 2004 and 2005. But I recognize one of them (5125 S. Kenwood, just south of Hyde Park Blvd/51st St.) as my first grad school apartment: I lived there throughout the 1998-99 academic year, and it was definitely a university-owned/managed building at the time.

It's quite possible that there's a real effect here. But while I've only had time to skim parts of the paper, I don't see any indication of whether the authors have accounted for the different norms in different fields when analyzing their data for potentially fraudulent or deceptive behavior.

Just for example, physics papers produced by large international collaborations (e.g. every single paper from the Large Hadron Collider at CERN) routinely have hundreds of authors (e.g. https://www.nature.com/articles/nature.2015.17567): everyone who has made substantial contributions to the design and operation of the facility is listed, as is everyone on the data analysis teams. (My understanding is that people in those specific fields all recognize that "number of citations" is a mostly meaningless number for those involved, and other metrics for productivity are well-known in those communities and routinely used.) I hear that some genomics papers have broken 1000 authors as well.

I could easily imagine that the high end of observed publication numbers and coauthor counts would be dominated by those giant collaborations, even though there is absolutely no attempt to mislead anyone in the process. Can anyone tell from this article to what degree its conclusions might be influenced by this factor?

Sure! There are always people out there who obsess over linguistic and behavioral purity, and they tend to be very loud. I'm not certain that I've ever met one of them in person. (I've certainly met people who might say (e.g.) "Here's why I've started including pronouns in my email sig," but I think that's not what you're talking about.)

If the current "anti-woke" movement were just about shutting down those obnoxious purists, I think it would be a lot less controversial. But from what I've seen, the political rhetoric (and the associated policy positions now being implemented) strikes just as hard or harder at the folks actually trying to do the hard work.

Accessibility has always been part of DEI. (Certainly it's been a repeated topic of conversation in my campus's Diversity and Inclusion Advisory Board.) And yes, any effort that involves spending time or money to make sure that every person gets to "have some quality of life" falls under this same umbrella. That is literally what people are referring to when they call things "woke".

I don't know how old you are, or whether you ever really knew the web in the prior era that we're talking about. Forgive me if I'm making flawed guesses about where you're coming from.

Back in the day, if I wanted the answer to some specific question about, say, restaurants in Chicago, I'd search for it on Google. Even if I didn't know enough about the topic to recognize the highest quality sites, it was okay, because the sorts of people who spent time writing websites about the Chicago restaurant scene did know enough, and they mostly linked to the high-quality sites, and that was the basis of how Google formed its rankings. Word of mouth only had to spread among deeply-invested experts (which happens quite naturally), and that was enough to allow search engines to send the broader public to the best resources. So yeah, once upon a time, search engines were pretty darn good at pointing people to high quality sites, and a lot of those quality sites became well-known in exactly that way.

Reading through the report, the trouble turns out to be that "devastating disease for humans" is possibly the least of our problems. Even if we had a perfect stockpile of antibiotics that would protect humans against any possible mirror organism (we'd need a wide variety, right?), all of our crops and livestock would be wiped out. All the forests. All the plankton. It would be really bad.

There were easily visible stairs in my level 1 and level 3, but on level 2 I walked around the whole map twice without finding them. Possibly I was just being oblivious, but I eventually tried moving sideways onto what I thought was a slightly odd looking wall segment (I wish I remembered what exactly made me think it looked odd), and suddenly I was on the next level.

I don't know whether somehow the stairs icon just blended in with the walls unexpectedly well and I managed to not recognize it in that context (which seems possible but hard for me to believe, given how much time I spent walking in circles looking for stairs), or whether something odd was going on with the icon being different or distorted for some reason.

The impact on election news coverage may not be that serious. Quoting from a NYT newsroom person:

"NYT Games and Cooking are BEHIND THE PICKET LINE. Please don’t play or engage with Games or Cooking content while the strike lasts!

News coverage — including election coverage — is NOT behind the picket line. It’s okay to read and share that, though the site and app may very well have problems."

(https://bsky.app/profile/maggieastor.bsky.social/post/3la4qg...)

I see your point, that it's really just an overall normalization for the size rather than anything to do with the shape. I can accept that, and I'll grant them the "four non-zero parameters" claim.

Though in that case, I would have liked for them to make it explicit. Maybe normalize it to "1", and scale the other parameters appropriately. (Because as it stands, I don't think you can reproduce their figure from their paper.)

Sadly, the constant term (the average r_0) is never specified in the paper (it seems to be something in the neighborhood of 180?): getting that right is necessary to produce the image, and I can't see any way not to consider it a fifth necessary parameter. So I don't think they've genuinely accomplished their goal.

(Seriously, though, this was a lot of fun!)

You want to know why a monopoly is bad, using only evidence from when it was not yet a monopoly (or not quite)? That feels to me a bit like missing the point.

I think for a lot of us on the older end, we lived through the era of Microsoft Internet Explorer dominating the web and that experience informs our thinking. As long as there was competition between MSIE and Netscape, with each one trying to outdo the other, both browsers kept getting better and the web kept becoming a more and more capable platform. But quite soon after Netscape crumbled and stopped being a serious competitor, MSIE stagnated: development didn't just slow but halted for half a decade. The web stagnated, too, and Microsoft's dominance meant that a lot of what did get built was locked in to their platform. (Partly things like CSS quirks and nonstandard rendering behaviors, and just plain neglect of new possibilities in HTML, JS, and CSS. But more than that: how many companies built ActiveX controls in that era, which mostly required Windows to function? The entire internet infrastructure of South Korea got locked in to ActiveX by law from about 1999 to 2020.) So imagining an era of Chrome monoculture brings back some pretty negative memories.

I don't expect that Google would make the exact same mistakes that Microsoft made. But it would be awfully hard for them not to shape browser design around their own corporate interests if there were no competition driving innovation and no disincentive to shaping the entire future of the web platform in Google-friendly ways. I know that's not "things that have actually happened", but the whole point is that things change once an effective monopoly is achieved.

Just fetch it via http rather than https. Works fine.

I guess I'm not entirely clear on the background/context here. I gather that the tweet's author isn't a serious Nethack player himself, but he is trying to train a neural net to play the game, and his training system somehow takes a model created by someone else as a baseline and tries to fine tune it somehow? But despite Nethack being based on randomly generated dungeons, the other model gets a consistent score every time, somehow? But even though the reference system reliably gets the same score through all the randomization of the dungeon, the game's full moon mechanic somehow throws it off significantly.

I feel like I mostly understand most of the pieces of this story when taken individually, but I'm having trouble assembling most of them into a coherent whole.

It's fascinating how much of this reflects societal priorities in what to reward. That's made entirely obvious by "education" being one of the low ROI degree programs: teaching effectively genuinely does require a college degree, and society genuinely does need effective teachers! (The more students grow up with poor teachers, the more we waste our human potential.) But US society has chosen to pay teachers a low enough salary that it doesn't end up working out financially.

Absolutely none of this is about "students making poor choices"! Or at least, if every student made "smart" choices, we'd be left without anyone to teach our kids. (And in fact, that's happening, to some degree. My local school system has struggled to fill some vacancies for the past few years. And Florida, for example, has had such a teacher shortage that they've decided that military veterans and their spouses are mysteriously automatically skilled teachers even if they've had no training or college education at all. Paying higher wages doesn't seem to be considered an option.) I suspect that some of the other "low ROI" fields suffer in similar ways.

You're right: the "more distant galaxies are moving away faster" point is just Hubble's original observation of an expanding universe. It's not an argument for cosmic acceleration. (If you see people making that claim, they're probably either speaking carelessly or not experts themselves.)

The conclusion that the expansion is accelerating was a quite recent result: 1990s, I believe. It's based on careful measurements of supernova explosions of a type with computable intrinsic brightness in increasingly distant galaxies, and the exact pattern seen in their apparent redshifts vs. apparent brightness. It was a shocking discovery when it came out, with two separate teams announcing the result pretty much neck and neck. There's also independent and compatible evidence for acceleration from the exact pattern of variations in the temperature of the cosmic microwave background seen at different points in the sky.

I took two points from this part of the article. First, that the expected exchange for years was "users create content, the companies provide the platforms and show ads to pay for it". But over time, the companies have not just adopted new ways of profiting off of the users (which may feel more exploitative than simple advertising) but also reduced the functionality of the platforms.

And second, that if there are vast amounts of money to be made off of all this activity, it's fundamentally not just for all of that money to go to the people who run the infrastructure, rather than some of it also going to the people who create and moderate the communities that make the infrastructure worth money at all.

If most of the money flowing in to a company like Reddit went directly toward improving the user experience, then maybe the original deal that "the platform itself is your reward" would still be fine. But as soon as there are executives and investors getting big cash payouts, that means that a whole lot of user-created value is being siphoned off: the deal doesn't feel nearly as fair anymore.

I don't quite understand the words you're using in your "the logic is as follows" list. (What are "gaps that otherwise invalidates math and disqualifies observed measures"? Really, what are "observed measures"? What does it mean to "performed qualified measures"? etc.)

To a substantial extent, it kinda feels like what you're describing is literally the scientific method. We discover observations that are inconsistent with existing models, so we develop a new model that (within their realm of validity) correctly match the data, and when we continue to find that the new model successfully describes new observations our confidence in the model goes up (and eventually rises to the level of saying "this seems to be a correct description of reality").

Maybe if I understood the exact meaning of your words here, I'd understand how what you're talking about differs from that and what your concerns are. But as I've said elsewhere, I really don't see where this is a "confidence problem", and I've never had reason to believe that alternatives are "censored to hell". There are people who've built entire careers publishing papers studying MOND, after all. It just doesn't seem to work as well as dark matter (Lambda-CDM) to explain the copious data that we now have.

And I'll definitely push back (again) against calling the case for dark matter "an observational limitation". These are quantitative, positive observations that demand some model to explain them. The specific quantitative amounts and distributions of dark matter necessary to explain galaxy rotation curves also turned out to be the same specific details necessary to explain gravitational lensing observations. As new observations have come in over the years, the case for dark matter has gotten stronger. (That's why it's become a dominant belief among experts.) So I genuinely don't have any idea what you're getting at with those comments.

The accuracy of opinions/predictions is almost always inversely proportional to the confidence of the claimant.

I recognize the sort of phenomenon you're pushing back against here, and there's legitimate value in consciously trying to counterbalance the all too human tendency to nod along to overconfident proclamations of truth.

But even so, the claim you're making here is absurd when treated as a general rule. Expertise is a real thing that people can have. Yes, there are countless places where a genuine expert will express caution and give carefully qualified statements about what is "true", and that's good. It's especially common for topics close to the frontiers of human knowledge. But there are also a whole lot of places where an expert will be willing to say "it's basically safe to take ____ as true", and where if asked they'll be able to back that up with copious explanations and evidence.

If I tell you that tomorrow morning the sun will rise in the east rather than the west, I will say it very confidently, and I will be right. If I tell you as a physicist that this bowling ball I'm about to drop from the second floor window will accelerate toward the ground at a rate very close to 9.8 m/s^2 (maybe only a smidge less due to air resistance, given the short time involved), I will be right. You trust your life every day to countless engineers who have at some point said, "The design and fabrication process we have implemented has involved enough rigor and enough quality control and enough margin of safety that I'm willing to stake my career on it being safe." And the vast majority of the time, for each essential bolt in your car and each fuel line on your airplane and each support beam in your office building, they are right.

So here, in the case of dark matter, the community of astrophysicists and cosmologists has assembled over the past 50-ish years a tremendous array of evidence from multiple independent directions that all seems to paint a consistent picture of some type of matter with quite specific properties. Many of them were highly skeptical at first, and there are still some actively looking for radically different explanations. But the vast majority of those experts have considered all of this positive evidence for "dark matter" and have become convinced that something very close to the Lambda-CDM model must be more or less true.

It's not at all clear to me what more you want from them. I mean, heck, the gravitational lensing results (especially as applied to something like the Bullet Cluster) could be considered a successfully tested prediction of the dark matter model, since those measurements weren't even possible until well after most of the community had become reasonably convinced that dark matter was real. The fact that we haven't yet answered the question "how does dark matter fit with the rest of our understanding of particle physics and astronomy?" is fascinating, but the existence of open questions doesn't on its own mean that all that observational data somehow doesn't count!