A good, nerdy deep-dive into the details: Stranger Things and the Amiga 1000
I’ve waited for years for a cross-platform whiteboard/diagramming/mindmap tool that worked anywhere near as well as Scapple on the Mac. Today, Obsidian Canvas launched and so far, it appears to meet the brief.
What part of the Gartner Hype Cycle are we in now for ML? Is it the Burning of Megatons of Cash or the Accidental Destruction of Innocent Bystanders? Let me know when we get to the Plateau of Standing Around in the Ashes Saying “Who Could Have Predicted This?”
I generally don’t pay attention to Bird Site shenanigans these days, but I just learned they’ve banned links to a specific set of other social media. Ok, whatever. But one of those is Nostr, and as far as I know, you can’t even link to Nostr. They added that for hipster cred.
The daily 🐶s
🐶
I believe Talking About Large Language Models by Murray Shanahan should be required reading for anyone working with LLMs. You really must understand what you’re actually asking of a system vs. what it looks like you’re asking of a system.
The Model Does Not Think; The Model Predicts
In my day job, we work with transformer models a lot, so I’m always interested in how they advance. I played around with the OpenAI Chatbot and found the results compelling. Asking it to generate a coding tutorial (“How would you create an event sourcing application in PHP 8?") was …
This makes me happy: Charles Schulz’s 100th Birthday Comic Strip Tributes 🗨
For #FeedReaderFriday, a helpful hint: Every Mastodon profile comes with its own RSS feed, for easy following. Just append .rss to the profile URL. For example, https://mastodon.social/@Tonybreed.rss.
How wonderful, a new Octopus Pie story! 🗨
An unfortunate week for influential writers: RIP Fred Brooks (The Mythical Man-Month, The Design of Design), and Greg Bear (Blood Music, Eon).
Funky Winkerbean is ending after 50 years. That strip, oddly enough, has been a touchstone for most of my life. 🗨
📷 Credit: Hackaday: Ugliest Airplane Ever Built Predicted the Future
I’m not saying this is just a Big Trak for (mostly wealthy, I suppose) Gen-Xers, but it sure pushed those buttons for me: Yarbo Yard Robot
🐶
Data science at its most relevant: Compound Pejoratives on Reddit (NSFW: language)
Well, this is just sad. I used to go to the Parkette when I lived in Lexington. I have photos of the sign I took for a project back in the early 90s. Lexington’s Iconic Parkette Drive-In Closes
From /dev/lawyer: Legal Sources for Not-a-Lawyers “It’s your law, people. Don’t let lawyers scare you off of reading it!”
I either missed or forgot that Ben Bova died last fall. They just released the last novel he conpleted. I read dozens of his books as a kid, one of which was the first book I ever finished and thought, “Wow, an author can make mistakes like that?” (Cyberbooks was that novel.)
This is the kind of content I think the web needs to get back to in a big way: 8-Track Heaven
For anyone you know who might need it: Vaccines Work. Here Are the Facts. | The Nib
Pure evil. Terms and Conditions Apply
Fellow children of the 80s, read this and say it with me: “No shit, Sherlock.” 1980s pop culture was all about impending, inevitable nuclear doom. Tom Nichols at The Atlantic: I Want My Mutually Assured Destruction I have a vivid memory – cir. 1983, so I’d’ve been 14 …
Now that this comic has ended, I feel compelled to go back and re-read the whole thing again: Back by KC Green and Anthony Clark. Highly recommended.