Sometimes the monopolist does something better for everyone. I’m this case, it’s selfish, of course, because they want people to click the back button to get back into their advertising platform “ecosystem”, but it’s nice to catch a W from "Don’t be evil [when you’re building marketshare]” Google.
I wonder if that’s AI slop? Not saying the whole article is, mind, but if an LLM is used at any stage, it likes to inject superfluous, overly exact adjectives that don’t match the context.
I don’t have time to get into the full 13 (? iirc) steps of Liljedahl’s Thinking Classrooms approach, but it’s exactly designed to meet the needs of students like you. Some highlights:
It’s an incredibly effective teaching method for secondary math. And there’s clear motivation every step of the way for what you’re doing and why it matters.
And the teacher only explains about 5-10% of the material; everything else is explained by the students as the carefully curated progression of activities guides them through discovering the math themselves.
Totally agreed, but authors are straight fucked if they try. Popular authors in my genre of choice have tried, and they all say it was a financial disaster for them, and that they can’t afford to be a full time author without KU income. And readers will follow where authors are, since those are the books they want to read.
Amazon’s monopoly on self publishing is probably illegal, but until regulators notice, network effects and anticompetitive practices from Amazon reinforce their monopoly.
Like, my options are, literally:
Authors have also said that they’re so dependent on The Algorithm, that pirating their books hits them double, from the lost revenue and from the reduced visibility. So that’s a double dick move.
I hate it, but here we are.
At least I read so much that Amazon pays authors like 10× what I pay to subscribe, so that’s pretty cool. (~300-400 books/year adds up to a lot of KENP pages!) And I’m not paying $3-5K/year for books to buy them all, sorry. I can’t afford that!
Yes, examples like that are good, of course. But, frankly, abstract examples like that won’t do much to motivate the students who need the most help to get motivated learning math.
I like to interject little anecdotes like that, too. One of my “go tos” to “why are quadratics useful” goes something like “Well, they come up a fair bit, so I could give you some examples—and I will, as we with through the unit, but the real reason we teach quadratics is because they’re the simplest non-linear function. This is the first steps into looking at functions that aren’t a straight line. And the tools you use to work with quadratics are super important for understanding all the really cool functions you get to learn on the next couple of years…”
That’s basically your example, but one step lower and more directly applicable to students, imho. The Taylor Series thing I usually only drop in grade 11/12 (pre)calculus classes, mostly as a hook for the math nerds that they have really cool things to look forward to learning in post secondary. It’s a terrible application to use to try to motivate learning about polynomials for a student who couldn’t care less, lol.
Really, we need to intermix all approaches, depending on the students in the class. At private prep schools, leaning into academic needs works well. In a non-academic math stream, both your example and my examples will go over like a lead balloon.
But, regardless, motivating students to be excited for math, and the excitement of finally figuring out a tricky concept/problem? That’s what we need more of.
If by “practical application” you mean “motivation for learning the skill”, which is I think the way you’re using it, then yes. But that’s not the usual definition in math education, and not what most people mean by it.
Like, for example, to introduce quadratics, a good progression might be to challenge students to build a table of values and graphs for x², then x² + 3, then graph x² – 5 without a table of values, then 2x² vs. 5x² vs. ½x², –x², etc.
And if you have a Thinking Classroom, every student in the class is working on figuring out that progression collaboratively in small groups. The teacher guides students to discover the math themselves through a series of examples, and mostly interacts with the students by asking questions, never giving them the answers.
That’s not “a practical application of quadratics”—at least not in the usual definition—that’s a learning activity sequence (paired with a set of interrelated pedagogical practices).
A good, practical application of quadratics is more like a Dan Meyer “3 Act Math” lesson on predicting the trajectory of a basketball shot. Also cool, good teaching. But not a great way to introduce quadratics.
(P.S. Yes, I use and like em dashes. I’m not a robot.)
Citation needed.
Seriously, though, that’s not what the research is showing. Peter Liljedahl’s research, for example, supports that a very effective way to teach mathematics is by having students actually think about math, instead of just passively receiving info dumps (as is common in most traditional math classes). See Building Thinking Classrooms for details but, in short, it’s a method of getting students playing with math concepts for almost the entire class time every day.
No “practical applications” needed. Counterintuitive, but it’s a highly effective practice.
What’s core to practical applications working is student motivation, and practical applications are one way to induce motivation. But it’s often not the best option, especially for inherently abstract skills.
That kinda breaks down in practice, though. Math is hard for a lot of students. Adding an extra layer of domain-specific application on top of an already confusing topic just makes it worse.
Like, we need polynomials for huge swathes of higher-level math. My favourite application of polynomials is that most continuous functions can be approximated by a Taylor series, which makes some functions that are otherwise impossible to calculate a derivative or integral trivially easy. It’s elegant, beautiful, and deeply practical.
And completely useless for a grade 8 student learning about polynomials for the first time.
Sure, there’s lower-hanging fruit for practical uses for polynomials, but they’re either similarly abstract (albeit simpler) or contrived. Ain’t nobody making a sandbox with length (3x + 5) and width (2x – 7), eh?
I could go on. At length.
Point being, yes, practical applications are better. BUT (and this is a big but) only when there are simple practical applications.
Instead, recent math education research supports teaching fluency through playing with math concepts and exploring things in many ways: symbolically, graphically, forwards and backwards, extending iteratively with increasing complexity, etc. This helps students develop intuition for math concepts and deeper understanding. Then, and only then, teach the standard algorithms and methods, as students will appreciate the efficiency of the tool and understand what they’re doing and why they’re doing it.
Thank you for listening to my TED Talk.
The challenge is the monopolistic death grip Amazon has on self publishing.
For many genres, authors get almost all of their income from Kindle Unlimited. KU requires exclusivity. The result is entire genres of books that are almost entirely Kindle exclusive.
So, the only real options for readers of these genes is either Kindle Unlimited (or buying on Kindle, I suppose) or piracy.
Some authors release serialized content on Patreon or a similar paid or free platform, but those platforms often only get first drafts, are difficult to navigate to get full books, and only cover a subset of authors anyway. And books get “stubbed”, which means everything past the 10% mark gets deleted to comply with Amazon exclusivity, so this is only even an option if you read the whole thing as it is being written. (FWIW, it’s also crazy expensive if you want to support authors; it can easily cost hundreds of dollars monthly with all the subscriptions.)
So, if you want authors to get paid for their work, then, realistically, you’re stuck using Kindle.
It sucks, but that’s the reality until regulators prevent Amazon from forcing exclusivity for inclusion in the KU program.
It’s still easy/cheap to buy new or used legacy USB data cables.
One wonders whether these old devices just don’t have enough telemetry built in for Amazon’s liking.
I think it’s likely more about DRM.
Old Kindles are incompatible with Amazon’s .kfx format ebooks and newer, stronger DRM. With an old Kindle, it was trivially easy to rip Kindle books to retail-quality epubs.
With these devices ceasing to work with Kindle books starting next month, that loophole closes.
Also, old Kindles will continue to work with already-downloaded Kindle books and DRM-free books, but new files can only be added by USB cable, not using Amazon’s services.
The newer DRM also has working exploits, but it’s not nearly as easy, and they’ve indirectly hinted that one of the remaining methods may be closing soon. But, fundamentally, static media DRM (books, music, movies) is inherently beatable; the full content gets displayed to the user, so it can be intercepted and ripped. Worst case, someone will make a screen-capture app that uses perfect OCR to recreate the book. That’s already a solved problem, basically, it’s just horribly inefficient.
So Amazon will continue to play whack-a-mole, turning millions of devices into e-waste, without even causing a blip for book pirates and those needing format shifting for accessibility.
OctoPi is included by default for a GUI app, iirc.
100% agree. Israel has been committing war crimes, and the West (broadly) is thereby complicit. By continuing to support Israel’s genocide for 2½ years, Western democracies have eroded international law to the point of uselessness.
The Geneva conventions are dead in the water, and good luck to anyone who is ever in a war zone in the 21st century. The destruction of civilian infrastructure and disregard for civilian life is now, de facto, acceptable.
What the AI slop did I just read? That wasn’t even edited.
The infill selections for comparison don’t make a lick of sense, and one of the tables doesn’t even have content that matches its introduction.
What a waste of time.
Edit: 57 upvotes? What the hell. Does nobody read the article?
The last sentence is kinda core to why parent poster’s comment isn’t really helpful.
But here we are talking about it, so more evidence that engagement hacking works, I guess?
The article mentions that there are performance penalties to running hypervisor cracks in a VM with GPU passthrough, but if you’re dual-booting barebones Arch (just enough to get a VM running), is there really a significant performance loss?
And once you have that set up on a separate boot, then it should be safe and fairly efficient to set up a bunch of these hypervisor cracks, I’d expect.
I don’t care enough or have time to tinker with this myself, but that sounds like a more fun project than actually playing one of these AAA games, lol.
The article highlights how the UK is moving to ban infinite scrolling access autoplay videos. So, thankfully, those changes are coming in at least some jurisdictions.
That said, the article also helpfully points out that the Republican administration has stuffed their science & tech advisory panel with Meta and Google execs, so I’m doubtful that the US will regulate anything reasonable.
I’d like a ban in effect for children below 16, but enforcement should be a misdemeanor on the parent. It should be a social worker coming to discuss with the parents the known harms of the platforms and let them get away with a warning, but that there will be fines if this damaging behaviour continues with an automatic 1-year (or whatever) follow-up. Basically, treat it the way it’s treated if parents are giving cigarettes to their children.
It’s coming, but it takes time. Business revenue is largely based on existing products, procedures, brand, contracts/business relationships, etc. It will take time for AI slop to reduce the quality of their offerings to the point that they’re no longer competitive.
In the case of AA/AAA games, AI hasn’t been around long enough to get a full production cycle using it. We’re already seeing AI slop in the indie games space, but it’s not really making waves because most indie games never do well anyway. As far as I know, there isn’t a single “very successful” game released with heavy AI use.
I know someone who works for one of the huge game companies (on the “live service” side) and he’s seeing it. (You’ve heard of them.) But he’s needing to walk a fine line with his team: he can’t reprimand them for using AI to make the slop he’s sent to review, because upper management is pushing for more AI use, but he’s getting plans & proposals that don’t make a lick of sense. On the surface level, they read fine, but on a deeper read, you realize the solutions don’t line up with the included examples, and don’t make sense in the context of their existing tech stack. It sounds really good, but it’s just garbage.
It’s only a matter of time before slop like that ends up costing them, likely in both product delays and worse performance and stability. Hopefully, they don’t have slop in their database/server security, as that will hurt their users not just the company going to shit.
The future is looking rocky for high-budget games. I’m lucky I almost exclusively prefer and play smaller indie games.
It’s a win, regardless, but the response is important.
The response should be:
Like, there’s nothing wrong with social media as a concept, it’s that profit seeking + network effects + regulatory capture have incentivized harmful social media.
Great article, and the author brings receipts.
I want to preface this by saying I’m not an AI doomer. I fundamentally disagree with the premise that a word prediction machine (LLM) is capable of intelligence. We’re no closer to AGI with LLMs than we ever were.
I also think AI has its uses; it’s a great tool, for narrow, constrained use cases. Editing text and vibe coding simple scripts, for example—but even in incredibly simple cases, it gets shit wildly wrong very frequently.
But the benefits are massively outweighed by the harms. Coaching suicide. Filling the web with AI slop. Reputational harm from not catching hallucinations. Semantic ablation.
We’re not getting rid of AI; the models are here to stay, and anyone with $2K of hardware can run a decent model at home. But that’s also going to be the end of the AI bubble. There are no natural moats to protect a monopoly. OpenAI will never be profitable since the value they create is less than their operational costs. It’s a money pit.
So, in a sense, I guess I am an AI doomer —the inevitable collapse of the AI bubble is going to cause a major recession, at least as big as the '08 financial crash, and these tech bros are doing massive harm both now and when the economic fallout lands. No surprise people want them dead.
But I’m not worried about LLMs turning into SkyNet.