Happy Monday! 

I watched a very timely play yesterday and really enjoyed it. It’s hard for me to feel moved by dystopian technology stories these days, but this one resonated. I think it helped that it felt less like a sci-fi future and more like an apt rumination on a current problem. A few more observations below.

Also in this newsletter: James Charles gets looped into a Filipino meme, Spotify Wrapped clubs, Fyre Fest (redux), and more.

Yesterday, I saw Marjorie Prime, the Pulitzer Prize-nominated play about aging, AI, loss, memory, and grief. Without giving too much away, the show’s premise is this: Set in the 2050s, an 86-year-old woman named Marjorie, who has some sort of dementia, is given an AI companion that looks like her late husband Walter in order to help her remember her life. 

The “Prime” bot is programmed with memories by Marjorie and her family, albeit selectively. Certain events are withheld from Walter Prime, so “he” gives an incomplete picture of Marjorie’s life — and his role in it — back to her. While the bot has Marjorie’s late husband’s likeness, how much of it is him? What kind of comfort can be found in this kind of technology?

I was not familiar with Marjorie Prime before walking into the theater yesterday. I was surprised to find out that the play is over a decade old at this point. It officially opens on Broadway today and is scheduled through Feb. 15, 2026. 

I was taken on a whim by Yap Year oomfie Moises Mendez II and I’m glad I saw it. I think it nails so much of what we’ve seen and discussed about AI this year, from AI psychosis to sycophancy to uncanny avatars of ourselves. There is now an AI product for any use case you can imagine. The tech is being touted as a solution to any discomfort, inconvenience, or problem a person may have, apparently including grief.

When I recounted Marjorie Prime to my boyfriend yesterday, he said that it was like an episode of Black Mirror (literally). I wish that was simply the case, but as we all know, Black Mirror has essentially become our reality. Watching Marjorie Prime, I was specifically reminded of a controversial app called 2wai that was announced just last month. (Sidenote: The app was co-founded by Austin & Ally actor Calum Worthy, who appears to be Bridgit Mendler’s Wario in the Disney-Channel-to-STEM-career pipeline.) 

2wai is an AI avatar app that allows you to create your own “digital twin” using a 3-minute video of yourself. When it was announced in November, it came with a promotional video demonstrating its potential to bring your loved ones back to life (kind of). 

The ad shows a pregnant woman speaking to her late mother’s 2wai avatar, which speaks to her as if it has her mom’s memories. When the woman’s child grows up to have a baby of his own, he also speaks to the 2wai avatar for advice, despite the fact that he never actually knew his real grandmother. (File that under unsettling marketing campaigns for AI products next to Friend.)

The app received a lot of backlash on X, where a lot of people questioned the ethics of keeping a version of a person alive this way. Is it right to have a bot speak as though they are the dead person? Or is that dishonoring a person’s memory? Can you program personality accurately? While this tech clearly has a lot of self-soothing potential for those in mourning, I wonder how much keeping a simulacrum of a lost loved one around sets back the grieving process. Will you ever reach acceptance if the virtual version of them is always around and speaking to you?

Then, of course, there is the issue of reality and the ways AI obscures it. People aren’t actually speaking to their dead relatives, but it could easily feel that way. The 2wai ad itself makes it seem like the users are speaking to and continuing to develop a relationship with the deceased person. We’ve already seen the ways that people can fall into AI psychosis through sycophantic LLMs. Now imagine hearing delusional or problematic advice coming from a bot with the face of a person you love and trust. 

Worthy went on a few news channels to promote this product and respond to criticism around it. Avatars can only be made of a person if they create one themselves. He said that these digital twins only know what the actual person tells them. At a certain point though, especially if the person passes on, isn’t the avatar just regurgitating the same few talking points that its original user fed them upon creation? What kind of relationship can families have with an entity like that?

There is a line toward the end of Marjorie Prime where a character, in a fit of grief, says to a Prime bot, “I’m talking to myself.” A decade old, yet this line encapsulates the unreality of AI perfectly. Despite what these companies try to sell us, these bots are what we make them and what we ask of them, not a continuation of the person themself. They simply cannot replace the real, imperfect, human loved ones we have and lose in our lives. 

If you can’t go see the play but you’d like to read it, I found a link to read Marjorie Prime here. 

Every Monday letter gets a playlist. I kept seeing videos about bringing back the art of the “dance floor makeout.” I made a playlist of club songs that I think could inspire a kiss or two.

In this week’s playlist:

Other things I want to share with you.