Amy J. Alexander

Purple/black "psychedelic"-style graphic with overlaid text that says: "Trust is a flimsy thing, especially when it's based on a stranger's words on the internet. Belief is often a choice, made in the absence of sufficient information. (Sounds like we're all just winging it, doesn't it?) "

Been working on Deep World — a custom LLM/chatbot I'm "training" on writers and thinkers who tried to make sense of difficult times.

Current installation version: you talk, it types back, and somewhere in there you both try to figure things out.

Videos + details:
amy-alexander.com/deep-world-i...

We could have had health care.
But instead we have this.

See you at NO KINGS in San Diego!

Today we said NO KINGS👑 loudly and clearly from Ocean Beach (San Diego, CA). #NoKings

Extra excited to be performing at, AFAIK, San Diego's first Algorave!
www.instagram.com/p/DKf1DGdyFk...
I (VJ Übergeek) will be performing a new live, custom LLM -- "Deep World -- a narrative reflection for 2025.

Honored to have the Plagiarist Manifesto as part of Panke Gallery’s “In Between – A NetArt Exhibition for Bright Night,” curated by the always amazing Sakrowski.
www.panke.gallery/exhibition/b...

Visual Arts, UC San Diego (@ucsdvisarts.bsky.social) The UC San Diego Department of Visual Arts encourages experimentation, innovation, and risk-taking in scholarly and artistic production.

Keith Obadike will join us in person; Mendi Obadike and Louis Chude-Sokei will participate via Zoom.
Co-presented by Gallery QI and ucsdvisarts.bsky.social UC San Diego Visual Arts Speaker Series.

Gallery QI Exhibit Opening Event: Mendi + Keith Obadike - The Skeuomorph - Qualcomm Institute Join us Thursday April 3rd for an opening event for Mendi + Keith Obadike’s installation of The Skeuomorph, exhibited and hosted at Gallery QI, and presented in collaboration with the Visual Arts Depa...

San Diego folks:
Please join us this Thursday April 3rd!
@mendiandkeith.bsky.social Mendi + Keith Obadike at Gallery QI. Their new work, The Skeuomorph, will be installed at Gallery QI Opening talk/panel/reception with guest panelist Louis Chude-Sokei.
qi.ucsd.edu/events/galle...

what_the_robot_saw - Twitch Invisible YouTube. Curated & animated live by contrarian algorithms. (No human is present.) what-the-robot-saw.com

BTW, "What the Robot Saw" 2.1 is now live!

www.twitch.tv/what_the_rob...
what-the-robot-saw.com

2025-03-21: AI and Artistic Autonomy Visit the post for more.

Excited to be on the panel: AI & Artistic Autonomy
This Friday, 21 March 9 AM PDT
Moderated by: Mauro Martino and Rebecca Ruige Xu and Gustavo Alfonso Rincon
Remote panel / Free registration!
Lots of cool panelists!

dac.siggraph.org/sparks/2025-...

Mahmoud Khalil’s Detention Is a Trial Run The pro-Palestine student was arrested without due process for exercising his right to free speech. He will not be the last.

Two things are true at once: 1) US campuses have had problems with actual antisemitism (not protests, not Israel criticism) they've been trying to brush under the rug for years. 2) Deporting/renditioning a green card holder for leading protests is McCarthyesque and worse. This does nobody any good.

The cover of the Elvis Costello album King of America, with Costello wearing a crown and looking stern.

This is the only King of America I recognize.

Hey! Lots going on in the world, and in the overlapping realms of social media, algorithms, and reality. “The algorithm” is more salient than ever. So, I decided it’d be a good time to rebuild What the Robot Saw with a 2025 edition.
amy-alexander.com/2025/02/what...
what-the-robot-saw.com

I'm obviously paying attention to the wrong stuff.

Tomorrow?? Philly on high alert right now! 😁😁

"enhanced standards of suitability and conduct" is the kind of phrase Orwell would pick a gun up to go shoot in a foreign war.

Exactly. Plus LinkedIn, Threads, and so on. Each with differing expectations. I’m in discussions with ChatGpt on coordinating all this…. Employing a bot as a concierge to all the other bot services seems fitting..

For those of you who want to help, here are some local Altadena organizations who were impacted.

These are all groups I have worked with personally and can vouch for the good work they do.

Thread. #EatonFire

I don’t so far, but I like that idea…hmmm.

This is what Grok suddenly decided i should generate of myself. They’re going to put drunk posting out of business at this rate.

(Including a number famously tasked with learning COBOL just for the occasion). Was weird when we did all that and then immediately everyone thought it was a big nothing burger….. The one thing I didn’t see coming was that!

Yyep! I only Y2K patched about 30 machines in my IT job. But there was a lot that had to get done, (specially on those SGIs!) Took the better part of a year, and I knew well that every sysadmin in the world was occupied with the same problem, often at a larger scale and more critical scope….

Very AI people who all look like slight variations of each other

I found a bot account on here and it has a great picture of his bot family

Nothing profound, but: When people use algorithms to analyze “the world,” they are analyzing that algorithm’s intervention into that world, not the world itself. This can be fine, but people neglect to acknowledge this when arriving at conclusions. There is an “unobserved observer” bias.

Yes, the algorithms are hiding the good stuff from you! This was (nearly) my point in making What the Robot Saw (2020). The algorithm that brings you the normal people you've been missing.
what-the-robot-saw.com

Microsoft Created a Twitter Bot to Learn From Users. It Quickly Became a Racist Jerk. (Published 2016) The bot, @TayandYou, was put on hiatus after making offensive statements based on users’ feedback, like disputing the existence of the Holocaust.

They're going to start training AI models on X? Where have I seen that movie before...
www.nytimes.com/2016/03/25/t...

Tidbits of new work
amy-alexander.com/2024/11/tidb...

Good on my name doppelganger for spotting this article! A great primer on how social biases (race, gender, and all sorts of others) get into generative AI training sets and thus amplified as new AI-generated content that is generated. As someone who generates the stuff: it really is this bad!