Bonus! Why stealing movies is good, actually

Wow, the whole opening theme song by Rita Moreno is great!

Oh, Matt, I seem to be the only one who's thought to mention this, but aren't you wearing a shirt (that I think I've seen you wear before) that says "I'd rather be watching the 1965 sitcom My Mother the Car", or am I losing my mind? Losing my mind... Losing my mind... (Ok random time for a Liza reference, but as soon as I typed 'losing my mind' the first time, she just popped in there. Haha.)

Let's not forget that Monty Python is "Monty Python" because Terry Gilliam stole and kept recordings of every episode. BBC notoriously wiped its tapes in that era, so a ton of equivalent sketch comedy shows were lost, but Monty Python remained long enough to find a secondary life through syndication and video sales. Because Terry Gilliam didn't want people to tape over his animation.

Kiss of the Spider Woman. Awesome. Very interested to see how that film would be received by contemporary audiences. I've considered recommending it to my daughter who is trans but not sure how it will land.

Newspaper! LOL

I really think you should have specified what particular parts of the "guiding principles" you are upset about, because otherwise it just sounds like you're saying "No, all AI is bad for history forever" since you are not pushing back on any specific points. (And you do literally say "AI will never be able to X" which I think would be foolish even for Computer Science experts to proclaim.) I personally am neither in the "OMG AI is horrible" nor the "OMG AI will fix everything" camp, but realistically (having looked up these "guiding principles" to try to figure out what you're talking about) I agree with their statement that "Banning generative AI is not a long-term solution; cultivating AI literacy is." And no, I did not figure out which parts of the guiding principles you might be so opposed to. It seems like they are saying AI use in an educational context should be OK if you're **not** relying on it as a source of truth. That it could be useful for things like summarization, brainstorming, proofreading, etc - things that more traditionally, students might turn to tutors, parents, or peers for. I don't think it would make any sense for them to say "no, nobody should ever use AI in history education for any purpose", because students WILL and DO use generative AI in every educational context. The thing that all educational fields are scrambling to figure out right now is how to channel that AI use into something that actually helps student develop the AI literacy they need, and understand what AI can and cannot be used for effectively. I also don't quite understand how you conduct your experiments to draw conclusions about the current state of AI. Of course an LLM by itself is going to be bad at pulling specific facts and pieces of information that are not common knowledge just out of thin air/its own training data. But why is that the criterion for whether AI is good/bad for history in general? LLMs **are** good at pulling things from texts you give them, as in the summarization examples in the Guiding Principles document. And increasingly, certain LLMs are able to consult or pull data from available public sources, and are getting good at making decisions about when to do that. A savvy user (with AI literacy they may have learned in an educational context!) is also able to explicitly direct an LLM to conduct "deep research", as in actually consulting and verifying data from real sources.

Okay, so this maybe going on with the color differences between your phone and your computer monitor. My husband works at a small video production studio. But since almost all their videos are for corporate audiences, he's familiar with how people view them. He's not an total expert on this, but he suggested it could depend on how your devices receive their information (cell? wifi?), and the compression rate needed for them. Not a total explanation, but maybe a starting point to solving the differences.

🏴‍☠️🏴‍☠️🏴‍☠️🏴‍☠️🏴‍☠️

I had always wondered what the deal with the Fleckman's was. It felt like a thing that everyone knew before I joined this patreon, and that it was just assumed that the rest of us would pick up through osmosis. Thanks for finally letting the rest of us in on the secret, and for the preemptive advice that a search on AI for Fleckman's would have invariable led us astray. You were right about the song too, it is pretty catchy, and the that operatic closing line is simultaneously jarring while also being a strangely perfect way to end such a strange song. It's too bad the show was never picked up to series. I think it would have been a hit on the opening song alone...

"Old people"? OLD PEOPLE???????

please take this as the good natured ribbing it's intended to be...Matt: "I hate what's been done to a movie for the Sphere." Also Matt: "I watch movies on my phone"

I have come to love my state library system and archive sites. Glad this was possible. Purely on the AI side. Thank you for saying AI is an imprecise term. I prefer to use AI only for things relying on ML models but they'll use it for anything these days. Gemini convinced me more than once I shouldn't use AI for anything more than OCR or assisted code auto-complete (never let it program for you). Regular OCR isn't perfect, AI isn't much better but I can train it to improve the results when it's failing. But Google's NotebookLM hallucinated a full 50 additional fields and people in a Census record. Gemini chat itself would mix up who/what we were talking about after just 1 interaction. In my opinion it's not just not ready for the public, it's very nature is not good for this use period.

I always appreciate folks preserving LGBTQ film, books, art, diaries. I donate funds (tiny) to LGBTQ archives in several different countries.

I would say it's less about relying on human expertise and more about being able to cite sources to verify. Humans are confidently incorrect and hallucinate information all the time. [citation needed] Out of interest, does Disney+ look different in Chrome versus Safari?

I love your commitment to accuracy! I'll see you at Fleckman's in the filtered pool. I'm the old lady with the flowery swim cap, you can't miss me.

So grateful for this one!!