The article Function calling using LLMs by Kiran Prakash describes an approach that works very well when you don’t have a testing environment: build a plan, evaluate validity of the plan, and then apply the plan after verification. You should also be able to slice the work into sub-tasks to make verification more reliable.

This is the approach I took for a PowerShell script that runs against ADOS (Azure DevOps): it’s production data, so you really want to be sure what is going to be executed, but you have no... [More]

The article A Critical Look at MCP by Rasmus Holm (Raz Blog) discussing many of the drawbacks of MCP as it is currently conceived. One of them is the push to build everything in Python, which is a dynamic language that’s better-designed than JavaScript, but isn’t a lot better at helping users write maintainable code.

“Am I being pretentious/judgmental in thinking that people in AI only really know Python, and the “well, it works on my computer” approach is still considered acceptable? This should be glaringly obvious to... [More]”

This video is a wonderful discussion of what it will mean to offload knowledge and wisdom to machines. Professor Asma discusses how humans have always offloaded to the environment to a certain degree. He argues that offloading to LLMs is like “the man in Searle’s Chinese Room”. I think that this offloading of knowledge and still believing that it would be a path to wisdom already began with the “just Google it” generation.

AI and the Post-Knowledge World by Professor Asma (YouTube)

The trend toward offloading knowledge—a little something called... [More]

Almost as usual, this interview with Mark Blyth doesn’t exactly go where the interviewer thinks that it’s going to go. They discuss jobs, AI, and scams. They discuss Britain’s idiotic economic policy—but Blyth levels that accusation against most of Europe and the West (e.g., Germany and the U.S., the two countries with which he’s the most familiar).

The interviewer tries to steer things to tried and true liberal topics—I don’t know why; had he never seen Blyth speak before? Or didn’t he... [More]

The post A pronounced issue by the-mothermayhem (Reddit) is a Reddit repost of a Tumblr “essay” that describes the painful fallout of having taught an entire generation without phonetics, with only the “whole language” approach, which—checks notes—involves a whole lot of wild guessing because you have no tools with which to analyze—in the strictest sense of the word: i.e., “break down”, or “parse” in the case of sentences, words, and phonemes—unfamiliar words.

The title is play on the once-popular Bushism (Wikipedia): “Is our... [More]

Chana MessingerThis ~10-minute video discusses research about chain-of-thought LLMs that “show their work”. Chana points out that, once you can see what the machine says it’s doing, it’s actually openly discussing “cheating” to achieve the correct result. She says that, once you add penalties for “cheating”, the machine doesn’t stop cheating—it simply stops writing about it. While this feels hilarious because it really seems to be acting like a teenager, it’s exactly this kind of anthropomorphizing that is... [More]

I ran into a small problem while upgrading Visual Studio 2022 to 17.14.0, so I reported it (Visual Studio Developer Community) with the following text,

“The Microsoft.VisualStudio.Community.MSI component could not be installed. At one point, the installer told me that another installer was running, but there wasn’t any installer running. An installer had run before the Visual Studio upgrade: JetBrains Rider. It’s possible that this interfered?

“I am unsure how this problem will affect my work. I don’t really use the MSI... [More]”

This podcast episode Who’s Coding Now? AI and the Future of Software Development by AI + a16z (Apple Podcasts) was recommended to me by a colleague. These are my notes that I took (and later cleaned up) from listening to this single episode.

Send a check or money-order to…

Near the beginning, one of the hosts says,

AI Bubble

“There was a good blog debate about whether we’re overinvested in AI. I think the number was $200B annual investment. And I think the question was how we would recuperate it?

“Well, here we have a way to... [More]”

The article New Claude 4 AI model refactored code for 7 hours straight by Benj Edwards (Ars Technica) talks about how awesome Claude is but then when you look at all of the charts, you see that it’s data published by Anthropic about its own software, publishing impressive percentages indicating some performance in benchmarks that they made up. So, they’re telling you that their software is amazing according to measures that you only learned about from them. This is basically a press release.

Pyramid SchemeAnthropic wouldn’t lie to get... [More]

The article I really don’t like ChatGPT’s new memory dossier by Simon Willison describes a new feature that incorporates memories of context from prior queries to ChatGPT.

👽 Thanks to PKD for the title.

“I’m an LLM power-user. I’ve spent a couple of years now figuring out the best way to prompt these systems to give them exactly what I want.

“The entire game when it comes to prompting LLMs is to carefully control their context—the inputs (and subsequent outputs) that make it into the current... [More]”

Below are links to articles, highlighted passages[1], and occasional annotations[2] for the week ending on the date in the title, enriching the raw data from Instapaper Likes and Twitter. They are intentionally succinct, else they’d be articles and probably end up in the gigantic backlog of unpublished drafts. YMMV.

Table of Contents

• Public Policy & Politics
• Journalism & Media
• Labor
• Economy... [More]

This is a nice explanation of how CSS is a declarative language, where you describe the metadata of your styles. The layout algorithm determines which property values affect the size and position of the element. Generally the properties position and display properties determine which layout algorithm is used for a given element. The layouts are,

CSS makes sense when you realize it's a collection of algorithms by Kevin Powell (YouTube)

• Normal flow layout (MDN) (selected by default)
• Inline layout (MDN) (selected by default for inline elements)
• Flexible box layout (MDN) (selected with display: flex)
• ... [More]

This is a clip from 20 years ago. It’s 3 a.m. and Private Equity is Extending an Invitation to “The Big Club” by Eric Salzman (Racket News) linked to it to point out that the vultures of Wall Street have been after Social Security for a long time.

The American Dream by George Carlin (YouTube)

🎩 h/t to George Carlin on the American Dream (with transcript) by Shoq (Shoqvalue) for the initial transcript. I’ve tweaked it a bit more, mostly for punctuation.

“But there’s a reason. There’s a reason. There’s a reason for this, there’s a reason education sucks, and it’s the same... [More]”

The article The Who Cares Era by Dan Sinker describes this era as a time when

“[…] completely disposable things are shoddily produced for people to mostly ignore.”

He writes further that,

“If you don’t care, [AI] is miraculous. If you do, the illusion falls apart pretty quickly.”

And that,

“Most people […] use it quickly and thoughtlessly to make more mediocrity.”

He gives what I consider to be good but probably career-killing advice in the our era. I really hope its not because I’m an optimist.

“As the... [More]”

This is an eviscerating satire of the nattering careerist nabobs, unprincipled hypocrites all of them.

Writer can no longer stay silent by Tadhg Hickey (YouTube)

“JanusJanice McUturn here, writer. Guys, I think we can all agree, the images coming out of Gaza this week, they’ve ripped my heart out and flung it against a wall.

“It’s unacceptable and I now—through enormous personal courage, actually—I’m ready to use that blasted G-word. It’s a [whispered] genocide guys. I’m ready to tell you that it’s a [whispered] genocide guys and I can no longer stay... [More]”

I’ve seen these before but this one seems legitimate. The article What happened to “All human beings are born free”? Reflections on a ChatGPT “experiment” by Mazen Baroudi, Shahreen Chowdhury, Farchanda Abdoel Wahid (IHP) asked ChatGPT two questions about human freedom.

The answers differed considerably, depending on the tribe. The freedom of Palestinians is “a matter of perspective,” which is, like, true, though? Like, a lot of people think Palestinians aren’t even human, so they correspondingly don’t think that they should be free. How else to... [More]

Scott Ritter is on fire in this interview about Donald Trump’s and Marco Rubio’s foolishness and evil.

Scott Ritter : Can Trump Bring Peace to Ukraine? by Judge Napolitano (YouTube)

“[…] what’s going on here is: Donald Trump is too stupid to live. I want him to succeed. I really do. I want every president to succeed but this is a man, and you just said it, ‘I don’t know if I will support the Constitution’—then get the hell out of the office! Because you took an oath to uphold and defend that Constitution and now you’re saying it’s too complicated for you?!? It’s too... [More]”

Saying “I want to make content every day” is shorthand for “I am remunerated for obtaining and holding attention, so I have to generate it. Content is a means to that end.”

I think very few people enjoy what they’re doing once they get on that treadmill. There’s one guy whose first couple of videos about “1 day in Germany vs. 10 years in Germany” were funny. He’s now produced dozens of them—the algorithm is diligent in surfacing them for me—and I’ve long since stopped watching them, though... [More]

This is also neat. I looked up “der Gruppe” vs. “die Gruppe” and learned that, while some people think they’re interchangeable, others think that they mean subtly different things.

Grammatik: …in die Gruppe… oder …in der Gruppe… ? (Narkive)

“ist es weniger klar als “in die Gruppe”. Denn die reine Ortsangabe kann auch bedeuten, daß er innerhalb der Gruppe in etwas Kleineres integriert ist.

““Ich habe mich in der Stadt eingelebt” ist schon erfüllt, wenn ich mich in meiner Wohnung wohl fühle.... [More]”

The title is clickbait but the content is nonetheless interesting. It discusses how to move processing from “system 2” (logical reasoning) to “system 1” (intuition). It’s how you get to a point where you understand a language without thinking about it. Or how you can just read music, or code, or vast swaths of text on economics or philosophy. Or how your body has learned to move in any sport or activity.

There is no way around using familiarity and repetition to get to highly accurate and... [More]

Read the explanation of method, madness, and spoilers.[1]

1. Doubt (2008) — 8/10
2. Mission Impossible − Dead Reckoning Part One (2023) — 7/10
3. Der Morgen Stirbt Nie (Tomorrow Never Dies) (1997) — 7/10
4. Puss in Boots (2011) — 7/10
5. Independence Day (1995) — 9/10
6. The Mummy: Tomb of the Dragon Emperor (2008) — 6/10
7. Happy S01 (2017) — 9/10
8. Chronicles of Riddick (2004) — 8/10
9. Shane (1953) — 5/10
10. Pitch Black (2000) — 7/10

Doubt (2008) — 8/10

This movie combines two great things:... [More]

Standard disclaimer[1]

Dungeon Crawler Carl CoverThis is my first book in a genre that a very good friend of mine said was quite popular in Japan—Isekai (Wikipedia)—which is “[…] a sub-genre of fiction. […] that revolve around a displaced person or people who are transported to and have to survive in another world such as a fantasy world, game world, or parallel universe”. Closely related is a sub-group with a more western-sounding name LitRPG (Wikipedia), which is “short for literary role-playing game, is a literary genre combining the... [More]”

I was forwarded the article Major Warning Issued for Apple, Facebook, PayPal, and Google Users by Kevin Harrish, (Newsbreak / Men's Journal), which made me think about how you can keep yourself more safe online.

What happened?

The article is not very good, in that it makes a lot of extra noise to sound alarming but that doesn’t actually contribute to the conversation. For example, “database of 184,162,718 records across more than 47 GB of data” is good. Writing “massive trove” and “massive” in two consequent paragraphs, or writing “Apple,... [More]”

Real Fit Life (RFL) was a training program invented by a high-school friend that Kath and I did for years. I had forgotten the exact steps and exercises but then found that my past self had written everything down in a text file.

Warm-up Exercises

The warmup phase is timed. You should do each set of exercises above the same amount of time. You can also adjust the amount of time you pause between sets. If you’re just starting out or aren’t so fit, you should probably do: Sets of 30 seconds... [More]

Below are links to articles, highlighted passages[1], and occasional annotations[2] for the week ending on the date in the title, enriching the raw data from Instapaper Likes and Twitter. They are intentionally succinct, else they’d be articles and probably end up in the gigantic backlog of unpublished drafts. YMMV.

Table of Contents

• Public Policy & Politics
• Journalism & Media
• Economy... [More]

This is a nice ~13-minute explanation of how CSS is a declarative language, where you describe the metadata of your styles. The layout algorithm determines which property values affect the size and position of the element.

CSS makes sense when you realize it's a collection of algorithms by Kevin Powell (YouTube)

Generally the properties position and display properties determine which layout algorithm is used for a given element. The layouts are,

Kevin Powell

• Normal flow layout (MDN) (selected by default)
• Inline layout (MDN) (selected by default for inline elements)
• Flexible box layout (MDN) (selected with display:... [More]

The article Liberals Believe In Nothing And Remember Even Less by Caitlin Johnstone (Substack) writes about how most people don’t actually stand for anything. They don’t have principles; they root for a team. She writes,

“I saw a post on Twitter where a leftist responded to a liberal who was acting like ICE just suddenly transformed into a modern gestapo under Trump, saying, “Liberals believe in nothing and remember even less.”

“And it’s just so true. They don’t believe in anything. They don’t stand for anything.... [More]”

I snipped the following citation from the article “When The Banality Of Evil Becomes Normalized, It Grows Unchecked.” by Francesca Albanese (ZNetwork) about a month ago.

“[…] the situation in the West Bank is not fundamentally different from what is happening to the Palestinian people as a whole. In Gaza, the attack has been genocidal in its intensity, but the same logic of destruction is being applied in the West Bank — though in a way that garners less attention, with fewer visible explosions. Palestinian communities... [More]”

Standard disclaimer[1]

This book is a 200-page, tightly edited, tour-de-force summary of many of Noam Chomsky’s writings, liberally sprinkled with Vijay Prashad’s interpretations and some of his own writings. It is structured as a conversation between the two authors, with some parts of Noam’s conversation being new and other parts being citations from his incredibly voluminous past work.

Despite a deep familiarity with the material, I very much enjoyed this book and just couldn’t stop... [More]

This is a 210-minute video about LLMs are built and trained. What works? What doesn’t? The whole thing is well-worth your time if you’re at-all interested in learning about what the inherent limitations are, so you can better leverage these tools. For example, “models need tokens to think” was great.

Deep Dive into LLMs like ChatGPT by Andrej Karpathy (YouTube)

• 00:00:00 introduction
• 00:01:00 pretraining data (internet)
• 00:07:47 tokenization
• 00:14:27 neural network I/O
• 00:20:11 neural network internals
• 00:26:01 inference
• 00:31:09 GPT-2: training and... [More]