right now the plan looks like:
----
Trump is giving a speech tonight. The notes will be elsewhere.
https://www.readnaturally.com/article/hasbrouck-tindal-oral-reading-fluency-chart
The Hasbrouck-Tindal oral reading fluency charts show the oral reading fluency norms of students as determined by data collected by Jan Hasbrouck and Gerald Tindal. Teachers can use these tables to draw conclusions and make decisions about the oral reading fluency of their students.
Notes on "oral fluency WPM":
https://collegetowns.substack.com/p/no-you-and-a-couple-of-friends-shouldnt
No, You and a Couple of Friends Shouldn't Buy a Dead College
Of course a "couple of friends" shouldn't. You would need at least ... 40 people. A hundred would be better. 🔥 or, one billionaire who can hire 40 people.
In fact, one issue that caused colleges to end up closing in the first place is related to the maintenance of campus buildings and infrastructure. Administrations may put off deferred maintenance due to budget constraints, leaving repairs festering. The problems then exacerbate, with repairs becoming more expensive.
Old pipes can easily freeze or leak, causing flooding that ruins both the floor and ceiling. Roofs certainly will leak. Foundations of buildings can start to crack or crumble. Old wiring may corrode and become a fire hazard.
💡 if you allow your buildings to not have water and electricity, it becomes slightly easier.
🔥 it's also easier if you don't mind a bit of asbestos.
----
As always, it's also a "zoning" issue: 💡 I would recommend you negotiate with the local city council before you buy.
Let’s just say that you and your friends do not care about the maintenance or the upkeep; there is another blockage for your dreams of reliving Old School, Animal House, or Accepted: it’s called zoning.
The US is famous for its strict Euclidean zoning, which means that land is separated out by specific usages. Some low-density neighborhoods may be zoned for R1 residential, meaning only one- or two-story single-family homes, while others may be C4 for big box stores, with a range of other categories and codes.
Universities are typically coded in a special “institution” type of zoning category. This means that the university can operate some businesses on campuses like coffee shops, along with students in dorms, and other standard administration functions. It does not allow universities to do anything they want.
How to create a thematic worksheet: 💡 a theme could be "rhyming words", or "American history", or "basketball", or "cars"
The last one is where the commercial LLM tools struggle the most. But, they can largely be done by a Python script. 🔥 the machine can probably write the Python script
I'm not sure yet what evaluations I want to do for Claude 3.7 and/or ChatGPT 4.5.
Some questions I have been pondering asking it to evaluate:
One of the difficulties is that a truly "frontier" task cannot be effectively graded by the machine, or indeed by non-experts in the field.
----
education thoughts for the week:
----
other tasks will be minimal. it is a short week; I fly out very early on Thursday.
All I Really Need To Know I Learned in Kindergarten is a book of short essays by American minister and author Robert Fulghum. It was first published in 1986.
The title of the book is taken from the first essay in the volume, in which Fulghum lists lessons normally learned in American kindergarten classrooms and explains how the world would be improved if adults adhered to the same basic rules as children, i.e. sharing, being kind to one another, cleaning up after themselves, and living "a balanced life" of work, play, and learning.
At some point, school becomes entirely about reading, 'riting, and 'rithmetic. 💡 or, in today's age, Language Arts, STEM, and Social Studies.
Even the likes of Home Economics and Woodworking are often cut at the altar of college prep.
----
Unfortunately, the realities of teaching behavior to those over the age of nine cannot be discussed in this medium.
https://carlalallimusic.substack.com/p/the-true-costs-of-being-on-youtube
I started my YouTube channel in earnest in October, 2021, shortly before That Sounds So Good, my second cookbook, came out. A little over three years later, on January 29, 2025, I uploaded the 177th episode of Carla’s Cooking Show. That video, for a cheddar burger named after my mom, is my last for now, and possibly forever.
In this time, I logged 18 million views, grew my subscriber base to 231,503, and grossed a total of $187,997. Some of those numbers sound big, but the costs—financial, logistical, and personal—didn’t earn out. In this newsletter, I’ll explain the hard and soft costs of doing business on YouTube, why leaving is the right move for me, and what I’m doing (for money) next. ...
The original goal was to use my channel to drive sales for my second cookbook, That Sounds So Good, ideally earning the book a spot on the New York Times Bestseller List. I had been making videos for Patreon, where I had about 3000 paying subscribers. For the book, though, I needed to maximize reach. ...
If we roll with the average Adsense income, here’s the bottom line: $14k going out. $4k coming in. Net loss, month over month: TEN THOUSAND DOLLARS. That’s a lot to sink into a channel that is barely moving book sales and not getting me a TV deal. Simply put, it’s completely unsustainable from a business perspective.
The videos are good, but not great. She is pleasant and the recipes are competent, but I struggled to find a video that both piqued my interest 💡 for various reasons, I would only promote a vegetarian recipe here and was "good enough to share". Ultimately, the one in her blog post: https://www.youtube.com/watch?v=N9wfngvhV0E might be the best choice.
In the game of clicks, "good" isn't good enough. 🔥 there is a world market for approximately five celebrity chefs. ⚔️ well, actually, five American celebrity chefs
----
As far as the money side of things:
The most significant metric for earnings on YouTube is CPM, or the cost per thousand views that advertisers must pay to run spots against a channel’s content. To advertise on my channel, that number is about $29 per thousand views.
The other metric is RPM, revenue per thousand views, which is what the creator is paid. My RPM is around $10. Easy math: For a video with 30,000 views, I earn $300. People tend to assume that my take is much higher, in particular the viewers who can’t believe I have the audacity to put my recipes behind a paywall.
Again: It costs $29 per thousand to run an ad in my videos, and I get $10 per thousand. Where does the other $19 go? To YouTube, of course. That’s a 2:1 split in favor of the platform. Lord, give me strength.
This does not acknowledge the fact that YouTube will let you upload (and distribute) an unlimited number of videos, for free.
It does note that, if you "only" make $10k/month in advertising revenue, you are largely immaterial to YouTube.
----
And, the "production" side of things:
Releasing a video every week for years meant I was always navigating three overlapping schedules: pre-production for the videos we’ve yet to shoot; post-production for the videos going through their edit; and pub week for the video coming out. Honey, we’ve been grinding.
In 2023, when my marriage broke up and I had to stay in (goofy, upbeat) character for the sake of my and my family’s privacy—that was brutal. I was living a lie, projecting an inauthentic version of myself. The conflict between the person in reality and the performer ate away at me.
for grenora.
----
Create the "word banks".
----
Score the "word banks"
----
Score a text
There were a few issues with the c.2020 implementation of "proxy voting" for House members.
💭 Historically, for the Senate, there was a courtesy of "giving live pair", where a member opposed to a bill would decline to vote, to balance an absent member in support of the bill.
🔥 none of the Founding Fathers ever owned a telephone, or even used a telegraph.
----
There are issues with the unlimited use of proxy voting.
But, a limited form would be more feasible.
In particular, it should be optimized for cases other than a closely-divided House where every vote counts.
----
The initial idea was that remote voting would be permitted only from the State Capitol building of each state.
This has a few advantages:
----
Beyond that: there are distinctions between procedural votes, and substantive votes on legislation. 💡 even "substantive" is too broad a term; amendments are considered substantive, yet are meaningless unless-and-until the underlying bill passes
The first set of votes are often "contingent"; the nature of later votes are dependent on them. The infamous filibuster results from a failure to pass a procedural vote allowing for a substantive vote.
Yet: it is no great injustice if there are errors in procedure. Perhaps a minor injustice, but not a great one.
The impact of substantive votes is much larger; and these are generally not contingent on each other. It is possible to construct convoluted scenarios where one is dependent on another.
💡 But, a certain amount of a "multi-track" parliamentary process would be necessary to benefit from this distinction. And, unfortunately, a control-bloc of both parties is committed to the ability to kill bills procedurally that would otherwise pass.
1. https://ew.com/woman-suffered-crushed-spine-from-harry-potter-ride-awarded-7-25-million-11686593
A California federal court earlier this month ruled in favor of 74-year-old Pamela Morrison, awarding her $7.25 million in damages after she says she suffered a crushed spine after slipping getting off of a ride at the Wizarding World of Harry Potter at Universal Studios Hollywood.
According to court documents, per the LA Times, during a September 2022 trip to the theme park with her grandson, Morrison had been asked to get off the Harry Potter and the Forbidden Journey ride after her harness failed to buckle. She then slipped when attempting to step from the moving walkway and onto solid ground, resulting in her injuries, which included a lower back fracture and a significant tear in the muscles around her hip.
In said documents, Morrison described the fall as thus, per the report: “The belt was still moving, and so my foot went on that belt and then ... my other foot went on to the stationary floor, and it knocked me off my feet."
Per a scheduling report filed in the suit and obtained by Entertainment Weekly, the plaintiff sought monetary damages "for various acts of negligence (including premise liability and negligent hiring and supervision) resulting in multiple and severe injuries to plaintiff while she visited the premises at defendants’ amusement park – Universal Studios in Los Angeles, CA."
The coverage is minimally useful. The injury happened when stepping orthogonally off of a moving walkway. 💡 details will be forthcoming later, but this is less natural than stepping forward off a moving walkway But, the injury is described as a simple "fall". 🔥 it would be inappropriate to speculate about her pre-existing health conditions that might have increased her injuries from a fall 💡 it would be less inappropriate to note the Catch-22 that Universal Studios would also be sued if they didn't let her attempt to ride
----
2. https://gizmodo.com/chegg-sues-google-says-ai-search-results-are-killing-its-business-2000568061
Chegg is sort of on its last dying breath at this point, but it will go out fighting. The online education company, which started out renting textbooks and later expanded into online homework help, sued Google on Monday for anticompetitive practices, saying it is unfairly scraping material from Chegg for its AI-powered search results.
The lawsuit is not anything groundbreaking or novel—Google has been attacked over the years by the likes of Yelp over claims it uses its search dominance to copy products and keep users on its own website. Chegg argues that by scraping websites and surfacing snippets directly in the search results page, Google is killing demand for original content and eroding the financial incentives for companies like Chegg to actually invest in producing the material that powers AI programs. It will be all AI models of dubious quality citing other AI generated content of dubious quality.
While big tech companies have lined up in support of President Trump, some of the new administration remains skeptical of the industry, so it is not a guarantee that Google will be able to dodge the case. Though, it is notable that Chegg’s stock is down 30% the day after the lawsuit was filed.
💡 not even their own shareholders believe it. This is a dying gasp of a dying company.
The claim isn't that Google is using Chegg's content. No, they are simply complaining that Google is no longer giving them traffic.
Chegg cannot have a property-right to search-engine traffic as they are demanding.
💭 the story of the first Chegg (an Iowa State-based Craigslist-like site) is not one I remember well enough to tell right now. The Wikipedia article has some revisionist history and advertorial content.
----
3. the Shrimp Wars continue.
The argument is, roughly, that preventing cruelty to small animals is more important than large animals, because there are more small animals. This ... is mind-bogglingly stupid.
Many bloggers have suggested that the primary advocates (one "Bentham's Bulldog", and the infamous Richard Hanania) are doing it as deliberate trolling or psy-ops.
And now it is in "respectable" outlets like Asterisk Magazine ⚙️ it's unclear whether there is an actual magazine, or just their substack: https://asteriskmag.substack.com/p/yes-shrimp-matter
I left private equity to work on shrimp welfare. When I tell anyone this, they usually think I've lost my mind. I know the feeling — I’ve been there. When I first read Charity Entrepreneurship's proposal for a shrimp welfare charity, I thought: “Effective altruists have gone mad — who cares about shrimp?” ...
These numbers wouldn’t matter if shrimp didn’t have the internal experience associated with suffering, but a growing body of evidence suggests that they do. A comprehensive review commissioned by the U.K. government found strong evidence of sentience in decapods, which includes shrimp, lobsters, prawns, and crabs. Evidence was particularly strong in true crabs.8 Crabs can learn. They can make trade-offs and act to protect themselves in flexible, complex ways. Everything we know about the structure of their brains suggests that they feel pain. While shrimp have been studied less extensively, the evidence suggests that they have similar capabilities.
🔥 the evidence does not suggest this.
But it is also a way to get other people to beclown themselves. Lyman Stone, for one: https://substack.com/home/post/p-157972723 . The post summarizes as I surveyed 450 people and none of them agreed with the shrimp advocates. Which doesn't prove anything.
----
4. https://www.nbcnews.com/politics/doge/elon-musk-joe-rogan-doge-rcna194294 💡 the full story would be too politics, but one detail is relevant here
Musk had Rogan ask Grok, the AI chatbot built by Musk’s company, xAI, whether “all the gold was in Fort Knox.”
“Are you a f------ conspiracy theorist?” Grok responded, drawing laughs from Musk and Rogan.
When even Musk's own version of the machine points out that he is being stupid, you would think he would listen. But ... no.
----
5. https://kinder.rice.edu/urbanedge/what-if-we-spent-billions-improve-access-instead-gridlock
Between 1993 and 2017, roughly 30,500 lane miles of freeway were added in the largest 100 urbanized areas of the U.S. — an increase of 42%. The expansion outpaced the 32% population growth of those areas during that time, yet traffic delays grew by 144%.
In Houston, where the population increased 77% from 1993 – 2017, freeway lane miles were expanded by 28% and delays increased by 221%. An even more extreme example is Brownsville, which saw a 73% increase in population, a 287% increase in freeway lane miles and a 1230% increase in delays.
💡 with numbers like that, you know something other than induced demand is at-play.
⚙️ one should expect traffic to be O(n log n). there are various intuitive explanations (either just "sprawl", or the size of a Clos network)
And, the cause of the percentage gap is the "low baseline" fallacy: https://www.bts.gov/archive/publications/national_transportation_statistics/table_01_69 shows that delays increased from 2 HPY ⚙️ Hours per Person per Year to 24 HPY over a similar timeframe. Meanwhile, Los Angeles went from 50 HPY to 80 HPY. ... most cities went up by about 20 HPY.
This is a field filled with bad data, especially from the "transit activists". But they don't matter; the Trump policy is going to presumably be roads roads roads.