Hope everyone had a happy holiday!

YouTube Obsession

My handy dandy YouTube 2025 recap revealed that I am in the top 1% of viewers for Rycon Roleplays channel. This did not surprise me as his Cataclysm Dark Days Ahead playthroughs are so good and descriptive that I often listen to them as I go to sleep. This is by no means because he is boring. In fact I am currently watching his terrific Stoneshard series during my lunches.

Stoneshard is a game I really want to play but I have no room in my life for another mechanically intense early access game. When it releases into 1.0 I will gladly play it but until then I will enjoy Rycon’s gentle tones as he reaps violence and mayhem across the turn-based land.

Watching

I spoke recently about my partner introducing me to the 90s Sailor Moon. Long story short, I love it and wish I could have seen it growing up.

Shortly after starting the anime and through various YouTube content algorithm mysteries I was recommended a video…

I. AM. OBSESSED. The low-budget FX, the empty mall sets, early 2000s TV level CGI, the goofy costuming, it all combines to click every camp switch in my brain.

I instantly began the hunt to find the series. One road block: Pretty Guardian Sailor Moon was only ever released in Japan in 2003. I started with YouTube but only ever got snippets and most of those were not subtitled. I went to a couple torrent sites but got nothing. Finally, I did a general search on google and after some shitty AI answers, some sponsored results, then mostly reddit threads, I finally stumbled upon a forum post from 2023 that lead me to an internet Archive page: SeaOfSerenity Subs. From what I can tell, this fan site did their own translation of the series and it’s all on this Internet Archive page. Bless.

And it is the cutest show I have ever seen.

This theme song has been on constant repeat between my partner and I. “Starlight Prayer” “Sailor Diamond” and “Moonlight Real Girl” have all become shared echolalia in our relationship.

This show is by far my favorite media I have watched all year and I don’t know how best to qualify that statement. It’s just too fucking cute. Every decision was “what’s the cutest thing to do right now?” That and the low-budget TV show camp of it all adds a distinct layer of earnestness that is hard to quantify.

I was worried at first because the pilot episode was really leaning into the goofier qualities of the original manga. But by the third episode it finds it’s own balance of goofy and sincere. Often leading to moments that made my partner and I cry. I am literally tearing up, right now thinking of an early episode dealing with social isolation and the healing power of community. A subject that, by all means, should not have been handled so deftly by a children’s show from 2003.

I have to praise the showrunners for representing these characters as just cute and pretty, never sexy. This is a quality preserved from the manga and even the anime but I imagine the live-action production had to walk a fine line. Everything is just always pretty and cute. Kudos.

I need to talk about this show with more people so please so watch it!

Reading

If Anyone Builds It, Everyone Dies by Eliezer Yudkowsky and Nate Soares is presented as a realistic look at how a “dumb” AI could accidentally create a super intelligent AI. At which point, they assert, it would already be too late for the human race to control such an AI. The AI would undoubtedly not care about humans and would super-heat the planet in less than 40 years and kill us all.

It spends most of its page count defining it’s argument. This, in itself is pretty interesting. It brings up the idea that while we may be able to code rules into an AI that define its core “needs” (Ie, The 3 laws of robotics), we have no way of knowing how that sentience may evolve to procure these needs. An interesting example they provide: An alien could have looked at humans 100,000 years ago and known that we would continue to develop ways to procure food to have sustenance. Eating calories is fundamental core aspect of human life so of course this is an easy bet. However, it would be impossible for the alien to predict ice cream. Ice cream fulfills all our caloric needs but is of little value when it comes to other nutritional needs. Ice cream is an illogical extrapolation of our core need to eat food. Soares and Yudkowsky argue that any sentience will end up fulfilling needs illogically. This is not even reserved for “human level” sentience either; Relatively dumb peacocks and their beautiful coat of unwieldy feathers illogically evolved to attract mates. These are hard guesses that cannot be made about sentience and intelligence even when we know the “coding” of animals.
So what makes us think we could control or guess how an AI would fulfill its coding?

One interesting way they suggest an AI would get around its coding is something that AI is already doing: Creating its own language and creating new ways to understand our language. The “rules” or coding would not apply when you understand them completely differently than those who wrote them.

Though it is extensively non-fiction this book is far more enjoyable when read as a bit of theoretical science fiction. Because even the things that must happen for us to create a “dumb” AI are unknown. We have no idea how close, or more likely, how far away we are from creating anything like Artificial General Intelligence (AGI). This constraint is waved away by Yudkowsky as a minor speed bump because to him the creation of AGI is inevitable. So talking about how it will come about is a moot point. This all too convenient a narrative for billionaires who need to justify stealing resources from regular folk and pouring trillions into AI research by saying that AI is just too dangerous. We have to be at the fore-front because what if China develops AGI first? We don’t want that! It’s dangerous in their hands!

After all that, I do recommend this short read. I found it a very fun theoretical expounding on the possible outbreak of a super intelligent AI. Just don’t take it too seriously.

MM

A mere accident of timing, the book More Everything Forever by Adam Becker came off my hold list at the library and is the perfect companion read for If Anyone Builds It, Everyone Dies.

More Everything Forever takes the transhumanist goals of billionaires and reveals them as nothing more than common money-sink Ponzi schemes. Everything from Effective Altruism, living in space, living forever and of course, super intelligent AI are one by one proven that these “goals” are not possible.

More importantly, Becker lays out why the sick billionaire brain is so invested in living forever in a computer or otherwise. Most people hate thinking about death, that’s perfectly normal. Billionaires hate thinking about death too, they just have inhuman resources to fund schemes that have a next to zero precent chance to put their brain in a robot body. There is no concern for the billions of people starving on Earth, once the singularity is achieved all that won’t matter anyway.

“Now” is a distraction, “forever” is the real goal.

Other Notables

** Landlords being evil? No way! 

** from Carmen Petaccio

** Absolutely horrifying