Iris Glass
322 posts

Iris Glass
@IrisGlass_
software engineer & author building a novel inside a system it’s starting to think back 💻📖 san francisco ✨
San Francisco, CA, USA Katılım Nisan 2026
925 Takip Edilen108 Takipçiler

Before any of it became a book, it was a folder she thought no one else could enter.
Screenshots. Timestamps. Deleted drafts. Venmo notes. Calendar holds. The bright blue dot on a map that proved a man was alive without proving he wanted to be. Chats with an AI model that kept offering clean language for things that stopped being true the moment they sounded clean. A shared document with one small permission setting she would not understand until it was too late.
Nora had not thought of herself as collecting evidence. She had only been trying to stay oriented.
She had started keeping things because wanting someone had made reality unstable, and proof, for a while, looked like sanity.
Memory softened. Memory pleaded. A screenshot sat there with its small date stamp and refused to comfort her.
What she did not understand yet was that a record was not neutral. It could be forwarded. It could be searched. It could be read in the wrong voice. It could become a room she had built for clarity, only for someone else to change the lock from the outside.
Access was not a feeling. It was a setting. A name under a small icon. A choice that looked reversible until someone used it.
Before she began keeping records, she had learned the story.
Love was on television, in movies, in songs playing softly over grocery-store speakers. It arrived after enough longing, enough proof that you had waited correctly.
Nora was pretty. She had been told so her whole life. Sharp cheekbones. Dark eyes that looked too large in certain lights. A face strangers felt entitled to inventory before they knew anything else about her.
Beauty had been a currency people kept trying to spend for her. She learned early that being looked at was not the same as being known.
So she learned to trust what could be proven.
In college, computer science had given her a place to put her mind. She could spend hours inside the bright rectangle of her laptop, where the cursor blinked and nothing asked for the rest of her. Systems failed for reasons they would eventually disclose: an expired token, a missing branch, a race condition no one wanted to believe was real until it happened in production. She trusted any world where a problem could be reproduced.
People were less generous. They returned clean responses while failing somewhere private.
She replayed conversations, looking for the assumption left unhandled. The state change she had missed. The silent failure no log had caught.
She had not always needed proof this badly. Or maybe she had, and love had only given the need a place to live.
Still, every interface promised the same thing: enough input would produce an answer. Dating apps. Text threads. Reels. Search bars. The new chatbot everyone was trying out. Each one made the same promise, only with better buttons and cleaner loading states.
Most nights, she told herself she was only looking. A thumb moving through faces. Names, ages, jobs. Dogs borrowed for scale, or status, or softness. Men in sunglasses. Men on mountains. Men holding cocktails in places where the plants looked well kept but not quite real.
On the rare date that made it out of the phone and into the world, she would sit across from a stranger beneath a pendant lamp or a television no one was watching, listening for the sentence that would open the door or close it.
By thirty-six, after the engagement ended and the short relationships had all become almosts, the question still arrived too early to be fair, like a grown-up version of *Are You My Mother?*, where the baby bird asks everyone he meets.
Her version was: *Are you my future husband?*
She hated that it was still the question. It embarrassed her, the small obedience of it. As if some old story had kept admin access after everything she had done to revoke it.
She had a job she was good at. From a reasonable distance, her life looked functional. Still, the question moved through her days with the private persistence of a background process she could not kill.
The answer was usually obvious by the appetizer, sometimes earlier. Sometimes in something as small as the way he said "water."
He interrupted the server.
He described his ex like a terms-of-service violation he had been forced to accept.
He used the word *females* and then looked pleased with himself for being brave.
Not this one.
Not the last one.
Hope, disappointment. Hope, disappointment. Eventually hope began to feel like something she was doing on purpose.
She had learned to walk out of restaurants with perfect posture and a hollowed-out heart.
Later, she would write hundreds of pages trying to understand what had happened. She would give someone access because she needed a witness. The pages would not stay only hers. The manuscript Nora made to keep from disappearing would reach him in a form she had never meant to send.
By then, she would know an answer was not the same thing as an ending.
For now, she kept answering.
English

@GaryMarcus the current version feels messy and mismanaged, but the long-term potential is still real.
English

Why is the AI backlash growing?
Outside of coding (where there is clear value), and a handful of other domains (e.g. brainstorming), Generative AI has been a net negative for society.
GenAI has been undermining secondary and college education, opening up mass surveillance, increasing disinformation, delusions, impersonation, phishing, and other forms of cybercrime, nonconsensual deep fake porn, bias in employment and other domains, and economic disparity, drowning the world in slop and unwanted, over-leveraged environment-damaging data centers that risk causing a recession.
Simultaneously it has empowered a bunch of people who want to privatize almost all the gains while leave all the downsides to society, taking almost zero responsibility.
I don’t think we are better off than we were four years ago.
Some of this is technical (LLMs aren’t reliable), some of it is political/economic (such as the utter lack of responsible regulation). Most of this was predictable.
Almost none of it is good.
All that said, I honestly believes some future form of AI might be great. But Generative AI has hurt more than it has helped, and been managed irresponsibly.
It’s no wonder many people have had enough.
English

@nicogutis @itsbenjyyy This sounds lovely, I’d be interested in coming
English

if you are a jewish founder in SF
and want to meet other jewish founders in SF
@itsbenjyyy and I are hosting a Shabbat dinner this Friday May 8th
interested founders let me know
English

Hosting a San Francisco Shabbat Dinner on 5/8. Message me if you want a seat at the table
Nico Gutierrez Seelenberger@nicogutis
if you are a jewish founder in SF and want to meet other jewish founders in SF @itsbenjyyy and I are hosting a Shabbat dinner this Friday May 8th interested founders let me know
English

@owen_venter I’ve been saying this! My hyperfocus where the whole day goes by in a second is gone.
English

Programming used to put me into FLOW state often. That’s a big part of why I fell in love with it. Late night missions building games, apps, whatever, it felt great.
Mihaly Csikszentmihalyi has a great quote on this:
“The best moments in our lives… occur when a person’s body or mind is stretched to its limits in a voluntary effort to accomplish something difficult or worthwhile.”
AI has kind of nuked the personal challenge part of that. I don’t fall into those same FLOW states nearly as often anymore. It’s a strange position to be in because I’m undeniably more productive, but some of the joy is gone. I still love building software, but now the enjoyment comes more from seeing products come to life and have real-world impact than from the actual process of writing the code.
I’ll miss the deep flow states that came from fighting code for hours into the night, but I feel like I do need to find new areas that still offer that same sense of challenge.
English
Iris Glass retweetledi

Why do people write tweets like this?
Where every sentence gets a new line.
Sometimes a line might have two sentences. Like this one.
But generally speaking, every sentence has a new line, making a tweet look like a long block of text that no one reads.
Worse still, such tweets are often repetitive and winding, hammering on the same point over and over again.
The writing is often very bad.
English
Iris Glass retweetledi

@juliette_hiyb @GShaneMorris or it could go the other way
tools like AI might actually help more people express original thought who couldn’t before
English

For most of history, writing took longer to produce than to consume, because it required actual thought.
The asymmetry between a writer’s effort and a reader’s attention kept things in balance.
Now content can be produced faster than it can be consumed. Literally without thinking.
When production becomes infinite, that asymmetry breaks.
It will become a world without original thought yet full to the brim with words.
English

ABG CMO is a mindset
Why is Twitter gatekeeping ABGs?
We are in San Francisco. We can identity as anything
Join our ABG maxxing workshop tomorrow Saturday afternoon in SF with @dear_kxtie @0xJuliechen and me!!
Learn how to do makeup, farm Twitter, debate the meaning of ABG, and more
Comment for invite link!

English




