About Vending-Bench

Recently, I read an AI/NLP paper by Andon Labs called “Vending-Bench: A Benchmark for Long-Term Coherence of Autonomous Agents.” Here’s the press release version, and here’s the actual paper itself. The idea is that we are hearing a lot about autonomous AI agents lately—so let’s create a benchmark test which can give us some comparisons on how different agents do on a simple autonomous task. In this case, it’s running a vending machine, which includes finding good products, ordering those products, stocking and restocking the actual machine, and trying to make as much money as possible. On their press release page, you can try playing the game yourself for a couple turns—it’s basically a more realistic version of those old browser games where you have to run an ice cream truck or pizza parlor.

The paper is definitely worth the time it takes to read the whole thing, but the basic result is that the agents are generally coherent and able to keep the business running, but when failure occurs, it is catastrophic. (This is probably not surprising.)

Fun Failures

( interesting examples of catastrophic failures )

Results, Memory, Context

( the results of the paper, paraphrased )

The Medium Article and Consciousness

( an off-topic tirade about whether chatbots are alive )

Conclusions

I think I went a bit off-topic here, haha. If you’re interesting in the Vending-Bench concept, Andon Labs also did an experiment with Anthropic AI where they had their AI, Claude, run a real vending machine inside their offices. It’s a good read, and covers many of the same topics as the original paper but in more of a blog post format.

I don’t think we’re close to having these tools actually do jobs just as well as humans do. The fear I have, instead, is that they will be able to replace a person in an office job up to a certain point, and the people in charge will see that, say “good enough”, and let the LLMs have control that they shouldn’t.

So FLETCHER has come out with a new song! For the uninitiated, FLETCHER is a solo artist who released an album last year called In Search of the Antidote which was one of the best gay/lesbian albums of the year (my opinion). FLETCHER is queer and her last two albums were (almost?) exclusively love/breakup songs about her female girlfriends/crushes etc. I really like FLETCHER’s music…I think she has great vocals and her lyrics, while not groundbreaking, are creative and so fun to listen to/sing along with.

This is my favorite song off her last album, but I really recommend the whole thing!

So two weeks ago FLETCHER released her first single off her upcoming album, Would You Still Love Me If You Really Knew Me? which is titled “Boy” and is about a recent or current relationship that she had with a man. When blueshiftofdeath told me about this I was immediately skeptical and shocked (FLETCHER has only ever written “lesbian” music before) but bsod pointed out to me that she’s never claimed to be a lesbian, which I had forgotten. Her music is so unapologetically horny for women that it’s easy to forget (haha).

So, here’s “Boy”…

I’m immediately interested. The song is not anything groundbreaking, but I like the country music vibe/slide guitar (very unusual for FLETCHER!) and I thought the lyrics were very open while also being tongue-in-cheek. It’s basically a coming-out anthem, but she’s coming out as being in love with a man. It’s a song for the opposite world where being gay is typical. (I thought the slide guitar/country vibes were on theme for this.) This is all pretty funny on its own, but it actually also rings true since she’s so beloved by lesbian listeners, you can see her actually being nervous to tell “us” that she’s dating a man.

Now personally, I like FLETCHER as much as I currently do because she writes music that I can really relate to since it’s about woman/woman relationships. I think this is probably very common for her fans. But I also think FLETCHER is special because she’s so messy and honest in her music, plus she’s a great artist who puts out whole albums worth of interesting tunes. In other words, I’m hooked and I’m along for the ride. I’m so curious to see how the rest of the album feels—I’m assuming it’ll be all about this relationship based on her previous albums. “Boy” is obviously about the titular “boy”, but it’s also very much about her relationship to “us” (her fans) and the expectations that we have of her as our favorite queer artist. I’m curious to see her explore both these avenues in her upcoming album and to see what else she touches on. (For the record, I can absolutely imagine this going into territory that I hate and find really irritating. But the album isn’t out yet!)

So when I listened to the song, I thought there would be more “cheer pressure” in the comments and people “validating” her “new direction”. I did not expect the comments section to be almost 100% people hating on the song/her! Wow! Let’s take a stroll through the comments section, shall we? (Line breaks indicate a new comment.)

Finally. A woman with a boyfriend being brave enough to sing about it. The world needs more of this kind of representation lmao.

Lots of comments are in this style—sarcastic comments about how the song is a “coming out” anthem but with a straight relationship. I get it, and I do actually think it’s funny, BUT—the fact that you all are coming here to comment this actually validates her writing those lyrics in the first place. LOL.

Choosing to drop this album right in Pride month shorten humility for the community who built who she is. I hope is all worth it at the end.

Not this being released during pride month Oh MY GOD

I celebrate Cari and her bisexuality, ofc. But it IS very odd to release a song highlighting a male love interest, and a heterosexual connection, during Pride month. Especially with the political climate we are in in the US. Ofc she is 100% valid in her queerness, even in a het relationship, but it leaves a strange taste in my mouth for her to choose to roll the album out like this. It reads as ragebait?

Lots of comments mentioning this being released during pride month. I know this is kind of a meme online, but some comments seem to take it seriously. Pride Month isn’t actually a sacred time for the community—and saying this implies that the song is somehow anti-gay (I did not get this impression).

Tbh I see why people in the queer community are mad, she built a brand on lesbians and she comes back with a song about a guy, it gives the ammunition to the you haven't found the right guy crowd

I realize I'm going to get shitcanned for my opinion, but in 10 years of her career there hasn't been a single video of her kissing a man. Yes, she didn't say she was a lesbian-ever, but she also didn't reflect her attraction to men in her music; she only mentioned one affair in one song where the guy turned out to be a mistake for her. And as a result, now she has unknowingly become a confirmation of the phrase “it's just a phase - it'll pass, you just haven't met a normal man, you'll fall in love again”. And we don't have the right to judge her personal life, but you could at least represent your queerness a little bit and release a couple of songs about love for guys, and not just slap to our faces....I'm sorry.

This is more of just a personal gripe, but it really irritates me when people cite “it will make men [or parents?] feel like they can say ‘you just haven’t found the right guy yet’” as a reason for not saying certain things or writing certain stories. I get it—the point is that an environment where people think that it’s possible for lesbians to “switch back” to wanting men is bad. If this song contributed to that environment, it would be actively harmful to gay women. But…DOES IT? I say no, not only because FLETCHER never called herself a lesbian or implied she wouldn’t date men so it’s not even relevant, but because that idea stems from such a deep seated social understanding of male/female relationships that little things like this could never have enough of an effect to make a difference. It’s a drop in the bucket if it’s anything at all.

I should make a researched post about this, since I’m relatively uninformed, but this is another example of something that can’t be solved directly, only via greater social change and feminism. You can’t solve people saying “you’ll find a good man eventually” by telling them that “no, I will never find a good man and no lesbian ever eventually marries a man” because it doesn’t ring true. You can solve it by supporting women’s right to live with or without a male partner, or any partner—then it doesn’t matter what people think about who you’re dating, because you are free to do what you want. We’ve already made a ton of progress in this direction but obviously there’s more to do. I get that it’s frustrating but this is just not FLETCHER’s responsibility.

I live in a country where homosexuality is not supported. I hide my relationship from my family. I live in constant fear of getting caught. Fletcher, announcing your heterosexual relationship during Pride month really seems like an artificial problem.

I do feel for this person, because obviously it’s not funny to them that it’s a reverse-coming-out song. But like…give her a break. She didn’t do it to hurt you. (I also can’t help but find it funny that this person with “real” problems STILL references pride month. Would it really be less painful to you if she had done it in July?)

Honestly, I don’t really care what gender someone is dating. But when someone builds their whole career on that, it’s kind of disappointing. I mean, I grew up with these songs that helped me through a difficult time in my life, because I believed there was an artist out there with a great voice and wasn’t afraid to say things that almost every lesbian feels or goes through. I feel some Tatu vibes here.

Citing TATU?? If you’re not aware, Tatu was a female Russian pop duo who performed as if they were really dating, but it turned out that they weren’t actually. So, implying that FLETCHER was pretending to be gay before, basically. I feel like this is really offensive LOL.

They gonna play the shit outta this at the next Republican convention

obviously we all knew she was bi and not a lesbian but why is she doing it while completely erasing her sapphic identity. like what is going on. it's such a weird conservative vibe.

Lots of comments citing like “conservative” or “Republican” here. I feel as though this is reading “around” the song itself—like these commenters are just nervous/thinking about conservatives so they attribute the “liking a boy” theme to that, even though it has nothing to do with it. I think this likely stems from the parts of the song that are like “I didn’t think this would ever happen…but I fell in love with a boy 😳” which I could see someone feeling like it aligns with conservative ideas about like…lesbians actually being straight. But you’re letting them live rent free in your head! It doesn’t have anything to do with that!

Let me finish by saying that I do understand why people would be annoyed or dislike this song. They relate to FLETCHER because she also loves women, and then she comes out with a new album (yippee!) but when they listen it’s like “I’m dating a boy…sowwy” so they get pissed off. I get it! They absolutely don’t have to listen—but they are not trusting her at all. And all this during Pride month!

The reason why the vibes are off in the MDR room after Petey left is because Petey was a strong yang presence and Mark, Irv, and Dylan are all yin. Without Petey to balance out the strong yin from the other three, everything feels off. (Helly is yang but her fear and apprehension cause her energy to be dulled.)

I want to do a more robust reading update (like actually writing my thoughts) but this may be all I have the energy for right now. My last update was in June last year. Let’s see how many more countries I’ve covered since then! (New countries are marked with an (!))

(!) Canada*

Emily of New Moon by L.M. Montgomery

China

• The Husky and His White Cat Shizun (5) by Rou Bao Bu Chi Rou
• The Husky and His White Cat Shizun (6) by Rou Bao Bu Chi Rou

United Kingdom

• The Comfort of Strangers by Alan McEwan
• Rebecca by Daphne du Maurier
• Pride and Prejudice by Jane Austen

Japan

• The Honjin Murders by Seishi Yokomizo
• The Decagon House Murders by Yukito Ayatsuji
• Slow Boat by Hideo Furukawa

(!) Mauritania

The Actual True Story of Ahmed & Zarga by Mohamedou Ould Slahi

(!) Mauritius

Riambel by Priya Hein

(!) Norway

The Red Handler by Johan Harstad

USA

• Tom Lake by Ann Patchett
• The Cloud Roads by Martha Wells
• Fleishman is in Trouble by Taffy Brodesser-Akner
• Player Piano by Kurt Vonnegut
• Brokeback Mountain by Annie Proulx *(not a novel)
• Erasure by Percival Everett
• The Friend by Sigrid Nunez
• A Tale for the Time Being by Ruth Ozeki
• Sharp Objects by Gillian Flynn

*I technically had Canada before, but this is the first book I’ve read that’s from a single Canadian author (other one was co-written by a Canadian author and an American author). So I feel like this deserves a bit of its own fanfare. I’m still so surprised that I’ve read so little from Canadian authors!

…The United States still overwhelms every other country! This is now 13 countries, 14 if I count China. I haven’t been, just because it doesn’t feel right to include the webnovels I’ve been reading, even if they are traditionally published. There’s no actual rule I’ve been enforcing about what types of novels count and which don’t, but maybe I should include one about intellectual rigor. It just doesn’t seem right to let myself not read an actual novel from China, especially when there’s options out there. So, let’s say 13.

I wanted to do my informal yearly round-up of what books I read this year, and realized that I never did one for last year! So here’s a double header, 2023 and 2024. This year, I’m not including movies—I’ll hopefully put that together sometime soon. I watched an unusually large number of movies this year (I think…we’ll see if the numbers agree).

2023 Books

It’s interesting how relatively many books I read in the first three months of 2023, and how relatively few I read the rest of the year. I can attribute the stoppage at the end of March to a particular reason, but I wonder why I stopped for so long and so completely. I guess I was busy with moving and decluttering the second half of that year…maybe I was just unusually brain-busy. A lot was happening at work at the time as well. I read almost nothing for a whole calendar year—weird to think about in retrospect.

( Read more... )

2024 Books

I didn’t read a huge amount this year either, but it was a more normal year as far as that goes. This is also when I began my Sovereign State Novel Challenge, so I tried to read a bit more internationally.

( Read more... )

Ten Arguments for Deleting Your Social Media Accounts Right Now by Jaron Lanier

( Open me! )

The Evening of the Holiday by Shirley Hazzard

( Open me! )

We by Yevgeny Zamyatin

( Open me! )

The Actual True Story of Ahmed & Zarga by Mohamedou Ould Slahi

( Open me! )

The Honjin Murders by Seishi Yokomizo

( Open me! )

Tom Lake by Ann Patchett

( Open me! )

I have a bit of reading to catch up on posting about! (The earliest here is from mid-February so it’s been a minute.)

The Glass Menagerie by Tennessee Williams

( Open me! )

Kim Jiyoung, Born 1982 by Cho Nam-Joo

( Open me! )

Demian by Hermann Hesse

( Open me! )

Ubik by Philip K Dick

( Open me! )

Woodcutters by Thomas Bernhard

( Open me! )

So my friends may know that I’m a dabbler in the “Booktube” community (Books + Youtube). One creator I watch often is named Jack Edwards. He reads mostly classics and literary fiction and I like his personality and the “click-baity” theme videos he does as well. He also makes a large effort to read fiction in translation.

This leads to a comment that was left on a video of his three months ago which read:

2024 challenge suggestion: reading one book from every country you haven't read a book from yet. You had a poll about video ideas I believe a year ago and I'd actually really love to see something like this. Only if you want to of course!

As a professional hater, I read this and was like “you’re not really thinking about how many countries there are…that would take so long” and scoffed at it. Then I thought about it for 30 seconds and thought that it sounded really fun so maybe I’d try it.

So the challenge as I’ve defined it myself is to read a novel from every sovereign state. This will have to be defined more as I continue, but I’m defining “from” to mean simply that the author is “from” that country (thinking that “born there” is not the best way to define this, but I also don’t think I want to include American expats living in Mauritius as Mauritian lit, for example). I also want the challenge to have as much work that’s in translation as possible. So far I’m operating off the “know it when I see it” rule and will sharpen my focus as I go.

So far I’m treating it like a completionism challenge, so the idea is to read from as many countries as possible. I’m not sure yet whether I’ll keep track of every book I read from each country as part of it (though I do write down every book I read so I could reverse engineer it after).

To start, I went through my reading list that I started back in February 2018 and marked off any country which I had read a book from. This ended up being only eight countries! I was surprised it was that low. This means that I have 197 sovereign states left to read from (assuming I’m able to find a novel from all of these, which will most likely not be the case); if I read one book a week, which is the fastest pace I’ve ever kept as an adult, and only read books from unique countries, it would take me four years to complete the challenge. Wow!

Well, that’s not going to happen, so we’ll see what number I’m at by the end of the year. I’m very excited to share more as I read :)

(During my research for this post, I learned that popular booktuber Emmie has been doing this challenge for a while now! It’s not totally clear to me how many countries she’s read from so far.)

Note: SORRY I never finished editing this!

I’m so excited to see what you create!

Music

Music DNWs: Taylor Swift

Other than my DNWs, I like all kinds of music in vids, so you can’t go wrong. I love pop, hip hop, country/honky tonk, and very rhythmic music in particular.

The Black Phone (2021) [SAFETY]

I went into this movie wanting to make fun of it (based on the trailer), but I ended up actually liking it quite a bit. This movie is based on a short story by Joe Hill, who is Stephen King’s son. I mention this because this book has a lot of King vibes. In one sentence, it’s kids swearing and fighting for their life against, in this case, a killer child abductor and using supernatural forces to do so. I loved the buildup and payoff was so sweet—-I watched this in the theater and screamed/gasped at the climactic scene. The kids’ performances are all good, and are much of the campiness that cuts through what is otherwise a fairly dark story.

This movie contains themes of alcoholism and child abuse.

Broad City

This show is the best example of female friendship on television. I love them, and a vid about any of the characters/relationships would be perfect.

Four Weddings (US TV)

I always have to nominate a reality TV show for Festivids, and this one is this year’s. The basic premise is that four brides who don’t know each other all attend each other’s weddings and judge them. Unsurprisingly, there is some cattiness, but I don’t think the show is actually all that negative (or at least, the stakes are low, because any negativity is not directed towards the bride herself, just to the camera). The show is very lighthearted and a majority of the brides are very sweet and funny. It has EXTREME 2010s vibes, which I absolutely love. It’s like a time capsule.

I think a comedic vid would be how I’d approach this source—-some sort of meta maybe.

Peep Show (UK TV)

I started watching this show after finishing Succession this year (the showrunner (?) is the same on both, vibes are totally different). Peep Show is about two men who are roommates and “failures” in their lives (through every fault of their own). The show is somewhat similar to The Office and Curb Your Enthusiasm (cringey moments, no laugh track, a commitment to “realism” in the videography/style). The conceit with Peep Show is that every shot is from the viewpoint of an actual character in the scene, and we get the main character’s thoughts as voiceover as well.

There are lots of content warnings for Peep Show; I can’t reasonably list them here, but proceed with caution.

She Loves to Cook, She Loves to Eat [SAFETY] The story is about a woman who loves to cook, who makes friends with her butch neighbor who eats a lot. They form a fast friendship which quickly develops to a mutual crush. There are lots of cozy food vibes, and some sexuality discovery in a fun and I thought fresh way.

So…the last time I posted about a Tom Ford perfume, I ended by thoughts by saying that I really enjoyed the perfume but probably wouldn’t have more than one in my collection, due to the price. Well…I went back on that pretty quickly. I was gifted this perfume for Christmas of last year by my parents. I first tried it during a trip to Sephora with my friend silenthilltwo, where I tried a ton of different Tom Ford fragrances and other perfumes. The Neroli Portofino really stood out amongst all those as something memorable that I really enjoyed. The description from the Tom Ford website is:

VIBRANT. SPARKLING. TRANSPORTIVE. TO TOM FORD, THIS SCENT PERFECTLY CAPTURES THE COOL BREEZES, SPARKLING CLEAR WATER AND LUSH FOLIAGE OF THE ITALIAN RIVERA. HIS REINVENTION OF A CLASSIC EAU DE COLOGNE FEATURES CRISP CITRUS OILS, SURPRISING FLORAL NOTES AND AMBER UNDERTONES TO LEAVE A SPLASHY YET SUBSTANTIVE IMPRESSION.

This fragrance really reminds me of another fragrance I own, the Acqua di Parma Colonia. I didn’t even realize they were so similar until I smelled the Neroli Portofino at home (not in the context of perfume shopping)…then I realized that I definitely have a “type” haha. They’re both very fresh, citrus-forward smells that I find enjoyable and that sort of blend-into-the-background after the top notes have dissipated (in a good way; they’re still there but not overwhelmingly). This is an eau de parfum, unlike the Colonia which is an eau de cologne, so unsurprisingly this is stronger and lasts much longer.

Comparing to my other Tom Ford perfume, the Tobacco Vanille, this is very similar in how strong they feel. This is difficult for me to describe, but both TFs have this fullness in body that the other perfumes I own/have tried do not. They feel layered and complex in a strong sense. I think I like the Tobacco Vanille more as a rule, but I think that is mainly just because I enjoy those warm smells more than I enjoy the sharp citrus ones (which is funny, because I own two citrus perfumes and only one warm/woody one).

I really like the Neroli Portofino. I am not sure whether I’ll replace this bottle once I’m done with it. I like the Colonia a bit more overall, and they fit the exact same category (similar scents, would wear in the same kind of weather or to the same sort of event, etc.). The Neroli Portofino definitely wins out in terms of strength and it smells more “expensive” (maybe a better word for this is “quality”?), but I prefer the actual scent of the Colonia.

The cost for a 100 ml bottle is $390 (pictured is a 30 ml bottle).