Irregular Update #10

Hello Friends in the Computer,

Well, I can’t lie. I have had a supremely shitastic week. It started with monitoring my dad for signs of, well, imminent death last Friday1 and ended with the both of us riding out a tornado in my basement last night. In between my basement (which happens to be where my bedroom is) flooded three times (twice after the initial plumber visit where they assured me it was freak accident with float in my basement pump that wouldn’t happen again)(and yes the flood was into my bedroom), the brakes of my car stopped working, and I feel like there was something else but I can’t remember what that was now. So not much writing last week.

There is also still a lingering feeling of unpleasantness around the new LLM based tools like ChatGPT-X. It’s not the fact that it will end up like The Terminator or The Matrix (although…I’m not ruling it out), it’s not that it will be endlessly hyped by people that hype everything and make unattainable promises (they will), it’s not that people with no experience working in A2J will decide to throw it at the poor and disadvantaged with no real checks as to whether they are actually helping (ditto), and it’s not FOMO because I can dive in if I want, I’m…just not sure I want2

What’s the opposite of FOMO?

Maybe I’m a Luddite. No, the real Luddites, not the slur thrown at you by tech bros on Twitter when you tell them their new shiny tool sucks. Similarly I have never been able to muster excitement for self-driving cars because…I like to drive. I mean, automatic transmissions take away too much agency for my taste. I like to DRIVE. Yes I know they (eventually) will be safer. I don’t care.

So many of the use cases that I’ve seen for The Quattro are things I like to do3. Don’t get me wrong, I don’t worry about being replaced - I’m irreplaceable, baby! <finger guns> - it’s just that these solutions are not only NOT solving a problem for me, but they’re threatening to make it hard to justify my creative outlets. And they’re adding a middleman in between me and my output and I don’t think I like the way that feels.

So once again I’m thinking of Stickley and Hemingway and van de Rohe and Harper and Craftsman and Bauhaus and how they and my other artistic preferences may translate to my professional interests.

Link Round Up

I already wrote more than I intended, since I thought this would be more a link round up, but here’s some fun/interesting things I saw this week…

Let’s get the ChatGPT stuff out of the way…

Ryan McClead - who gave one of my favorite talks that I’ve ever seen at a Reinvent Law thing at Chicago Kent - has a good perspective on adoption of LLMs in law.

Also, about LLMs, this from John Grant on Mastodon

I don’t know that I care about the making money thing, but it’s good to remember that doing the AI work - which honestly doesn’t interest me - is not the only needed thing in the coming years and there’s lots of opportunities that will spring up around it.

Fun with History

• This listicle on 100 diagrams that changed the world is interesting, but what brought it across my desk is that apparently emoticons were first published in March 30, 1881

Of course, the best one remains ಠ_ಠ

• A fun rabbit hole thread on the dumbest turning points in History.

• And for the Cincinnati folks and/or the law nerd folks, here’s a fun (for varying definitions of fun) story about a Cincinnati court case conducted in German. Sometime I will need to scan and post the photo of my paternal grandfather with his mother. When I first saw this picture I was like “why am I in it and why I am I wearing a dirndl?” It’s hilarious- like a bad Lifetime movie where someone plays their own ancestor in a historical flashback.

Adversarial Fashion

Whether you’re fighting the police state or the AI overlords or a partnership between the two, here’s some fun fashion choices to throw some sand in the gears.

• Adversarial fashion

• Trick facial recognition software

I feel like I read something about people inserting metadata or other content into their materials to affect the AI harvesting it, but I can’t find it right now. But it makes sense that something like that would exist. Or should exist. Which reminds me, this is a good thread about the implications of how the ChatGPT magic gets made.

Well that’s all I have for now.

Be well,

Sarah

Thank you for reading Sarah's Commonplace Book. This post is public so feel free to share it.