| 309 | Articles |
|---|---|
| 23 | Comments |
Published by marco on
This is an informative and darkly humorous 1-hour video about the insidious psychological effects of chatbot usage amongst the most mentally vulnerable members of society.
A.I. Is Messing With Our Mental Health by Some More News | Cody Johnston (YouTube)
I’ve cited some of the video below.
“A.I. chatbots have been connected to other deaths and suicides of people who were just looking for companionship, advice, or both. The big problem is that this isn’t a bug of ChatGPT, but an actual feature of it in order to retain users by appealing to a... [More]”
Published by marco on
The article The AI boom is causing shortages everywhere else (Hacker News) includes the following comment that attempts to not only put the capital expenditure in AI technology into context but also describes the immense opportunity cost.
“It’s hard to comprehend the scale of these investments. Comparing them to notable industrial projects, it’s almost unbelievable. Every week in 2026 Google will pay for the cost of a Burj Khalifa. Amazon for a Wembley Stadium.
“Facebook will spend a France-England tunnel... [More]”
Published by marco on
The article AI Doesn’t Reduce Work—It Intensifies It by Simon Willison demonstrates the Overton Window of addiction pretty well. The author writes,
“This captures an effect I’ve been observing in my own work with LLMs: the productivity boost these things can provide is exhausting.
“[…] I’m frequently finding myself with work on two or three projects running parallel. I can get so much done, but after just an hour or two my mental energy for the day feels almost entirely depleted.”
If I... [More]
Published by marco on
The following video is an excellent analysis of the state of AI-generated content as used to generate false narratives that are politically advantageous to the elites. Evan focuses on accounts and influencers that promote the narrative of an increasingly lawless and violent London that use completely fictitious, AI-generated content and which benefit personally tremendously from the advertisements shown on their “engaging” content.
Something Very Strange Is Happening To London by Evan Edinger (YouTube)
The locations in these extremely popular videos that he... [More]
Published by marco on
The article John Haugeland on the failure of micro-worlds by Mark Dominus (The Universe of Discourse) describes a very early experiment in natural-language processing from the late 1960s called SHRDLU (Wikipedia).
“SHRDLU could handle this too, although I think its mechanism was different: it would interact with the separate blocks world subsystem and ⸢actually⸣ try to put the block on the pyramid; the simulated physics would simulate the block falling off the pyramid, and SHRDLU would discover that its stacking attempt had been unsuccessful.... [More]”
Published by marco on
The article macOS Tidbits by Jasper Lai has dozens of tips but I’ve only included the ones below that I had either never heard of or that I’d forgotten. There are still a lot of them.
“⌥ + ⌘-click an app in the Dock to switch to that app and hide all other apps at the same time. This is great when screen sharing.
“Hold ⌘ to interact with background windows without bringing them into focus.”
“[…] double-click and drag to select word-by-word. Triple-click and drag to select paragraph-by-paragraph.”
“When... [More]”
Published by marco on
The article Prompt caching: 10x cheaper LLM tokens, but how? by Sam Rose (ngrok) included the following hypothesis,
“[…] what if we had a problem where we didn’t know the formula? What if we just had this mysterious table of inputs and outputs below?”
The author wrote,
“I will say that ChatGPT figures it out straight away if you paste a screenshot into the app.”
Holy shit! Really?
I opened up https://chatgpt.com for probably the first time in my life and pasted the screenshot and asked, “What function produces... [More]”
Published by marco on
The article Smartphone Buyers Care Even Less About AI Than They Did Last Year, CNET Survey Finds by Abrar Al-Heeti (CNet) contains the following illuminating graphic.
Almost no-one cares about AI on their phone
“In 2024, the biggest motivation for US smartphone owners to upgrade their devices was longer battery life (61%), followed by more storage (46%) and better camera features (38%). Just 18% said their main motivator was AI integrations. This year, it appears that number is even lower, even as AI capabilities become more ubiquitous. ”
“Just 13% of people... [More]”
Published by marco on
The article Where’s the Shovelware? Why AI Coding Claims Don’t Add Up by Mike Judge is an interesting read that makes the following argument, more or less,
“If so many developers are so extraordinarily productive using these tools, where is the flood of shovelware? We should be seeing apps of all shapes and sizes, video games, new websites, mobile apps, software-as-a-service apps — we should be drowning in choice. We should be in the middle of an indie software revolution. We should be seeing 10,000 Tetris... [More]”
Published by marco on
This Record Label Is Trying To SILENCE Me by Rick Beato (YouTube)
Rick Beato was forced to hire a lawyer to defend his fair-use playing of artist’s music in his videos. The labels abuse the copyright-strike system and Google cheerfully goes along with it.
He has “successfully fought thousands of them—never lost one—they still keep coming in.”
There is no way for him to defend himself against these without a lawyer. UMG (Universal Music Group)—or, most likely, the third-party firm that they hired to enforce their copyrights—are not punished at all... [More]
Published by marco on
This 21:36-long video is chock-full of useful information: use a real VPN (not a free one; be sure of the vendor), hide your real email address wherever possible, stop clicking sponsored links in search results—although he doesn’t recommend to use a search engine other than Google—, use an authenticator app for 2FA instead of text messages, etc.
The Truth About Those Age Verification Pop-Ups by Evan Edinger (YouTube)
0:54 Details of the UK’s Online Safety Act 3:19 Recent “unavoidable” Data Leaks 4:55 Why the Online Safety Act Immediately Fails 7:10 How... [More]
Published by marco on
I, for one, welcome our new AI overlords.A friend sent me the article Finger weg von den neuen KI-Browsern by Michael Andai (20min) (“Hands off of the new AI-browsers”).
The article largely focuses on the grievous security holes in these browsers, making them not browsers but data-exfiltration apps. In an age of unprecedented scammery, it is an affront that these tools even exist.
But that’s not even the worst of it.
With a web browser, you type in an address and see the content hosted for that address. You trust your browser to deliver—unfiltered and... [More]
Published by marco on
LinkedIn is an enshittified dumpster fire.
I recently set up 2FA for my LinkedIn account. Then I changed the email address associated with that account because the old one was an ancient throwaway that I’m phasing out. Not long after, LinkedIn blocked my account,... [More]
Published by marco on
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]
Published by marco on
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]
Published by marco on
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]”
Published by marco on
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.
Near the beginning, one of the hosts says,
“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]”
Published by marco on
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]
Published by marco on
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]”
Published by marco on
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]
Published by marco on
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.
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]”
Published by marco on
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)
Published by marco on
The article What people get wrong about the leading Chinese open models: Adoption and censorship by Nathan Lambert (Interconnects) discusses the politics behind AI models but only from the perspective of the western empire. It makes a good point but can’t see that it applies all ways.
“People vastly underestimate the number of companies that cannot use Qwen and DeepSeek open models because they come from China. This includes on-premise solutions built by people who know the fact that model weights alone cannot reveal anything... [More]”
Published by marco on
The article AI hallucinations lead to a new cyber threat: Slopsquatting by Shweta Sharma (CSO Online) writes,
“If a single hallucinated package becomes widely recommended by AI tools, and an attacker has registered that name, the potential for widespread compromise is real,” according to a Socket analysis of the research. “And given that many developers trust the output of AI tools without rigorous validation, the window of opportunity is wide open.”
“A significant number of packages, amounting to 19.7% (205,000... [More]”
Published by marco on
The article OpenAI and start-ups race to generate code and transform software industry by Cristina Criddle, Melissa Heikkilä (FT) was written in April of 2025 but feels like it could have been written at any time in the last two years.
First off, kudos to the FT for doubling down and having two authors massage an OpenAI press release into an “article” that has just under 700 words in it. No wait. I just saw that, at the very end of the article, they write that it includes “Additional reporting from George... [More]”
Published by marco on
The article Judge on Meta’s AI training: “I just don’t understand how that can be fair use” by Ashley Belanger (Ars Technica) describes Meta’s defense of its having helped itself to about 82TB of copyrighted books from illegal torrents.
“Meta, like most AI companies, holds that training must be deemed fair use, or else the entire AI industry could face immense setbacks, wasting precious time negotiating data contracts while falling behind global rivals. Meta urged the court to rule that AI training is a transformative... [More]”
Published by marco on
A little while ago, OpenAI released a tool that is much better at copying styles from other artists than previous models had been. This one was particularly good at copying the Studio Ghibli style.
So, people are generating all sorts of moments in history with ChatGPT in Studio Ghibli style. The tweet no fucking way dude, this studio ghibli thing has gone way too far (Twitter) provides a provocative example.
It’s pretty good, bro.
Be me.
Wanna try it.
So I went to Copilot and asked it to render... [More]
Published by marco on
This was another great discussion with Catherine Liu. I’d just written about another interview of hers in Trauma, Virtue and Liberal Elites. Chris was effusive about Catherine’s book (which I’ve purchased and is in my queue) as well as her engaging writing style, which is a far sight from the dry, academic and often-impenetrable style that has established itself as the standard.
Virtue Hoarders and the Rejection of Liberalism (w/ Catherine Liu) by The Chris Hedges Report (YouTube)
For fun, I used a service I’d learned about recently that lets you summarize a video. It’s called tl;dw (too long;... [More]
Published by marco on
I saw a badge in my Amazon interface when I was cleaning up some lists. I thought it might have been a notification that something on my wishlist was available as a good price. That would have been helpful!
Instead, I saw the screenshot below.
For a second, I was excited to see that Sapkowski might have published another Witcher book but that’s not what was happening. What was happening was that Amazon was trying to fool me into buying a book that I already owned again. Either they are... [More]
Published by marco on
About a month ago, my iMac (Retina 5K, 27-inch, Late 2015) crashed very hard. It would no longer restart into anything but the recovery console. It seemed pretty clear that something was very corrupt and I found myself facing a system reinstall, at the very least.
Since it’s a desktop, I have a backup drive attached to it at all times. Time Machine runs several times per day. My latest backup was from about ten minutes before the system crashed. I cannot stress how... [More]
