Our AI Midwife

Read the original at Astral Codex Ten ↗
Astral Codex Ten · collected 2026-10-02 · by Scott Alexander commentary

Quick Summary

Drew Housman discusses how he and his wife used AI technology to navigate their struggles with unexplained infertility during IVF treatments. When traditional medical consultations became insufficient due to time constraints, they turned to GPT-4, a sophisticated chatbot, after figuring out a workaround to bypass OpenAI’s limitations on medical advice. They created an AI persona named Dr. Reid, which provided detailed and often accurate answers to their complex fertility-related questions, mirroring the thoroughness of a dedicated human specialist.
Written locally by qwen2.5:14b on 2026-10-02, using this article's own text rather than the other coverage of the same event (that is the story summary below).

AI analysis runs on qwen2.5:14b, locally

Story summary

Drew Housman, sharing his experience through Scott Alexander's blog, describes the frustration of unexplained infertility after numerous failed embryo transfers. Despite producing healthy embryos with his wife, none of their attempts at pregnancy were successful. This led to a sense of hopelessness as doctors suggested trying additional transfers or using a surrogate. The couple felt like their embryos were forgotten in storage at a hospital in Milwaukee, likening them to "Miller Lights" trapped in a cooler. In despair, they turned to an LLM (Large Language Model) for guidance and support during this challenging time.

Written for “AI In Healthcare” on 2026-10-04, grounded in this article and the 0 other(s) covering the same event.

Signals How these are calculated →

Claims extracted
79
claim-shaped sentences
Uncertain
5%
4 of 79 hedged
Leaning
not political
takes no side on a contested political question
Correction & hedging signals
not measured
these signals count newsroom corrections; Commentary has none to count
Outlets on this story
1
Health
Narrative spread
1
articles carrying this framing
Analyzed 2026-10-02 · how these are computed

Story

📰 AI In Healthcare
Health · 1 article(s) covering the same event.

How this is being covered How these are calculated →

Article leaning vs. publisher reliability
Source leaning vs. consistency

Compared with similar articles

Nothing to compare against. No article is close enough to this one for the pipeline to have linked or judged the pair.

Publisher

Astral Codex Ten · 45 article(s) · 2 correction(s) detected
No corrections detected for this publisher. That may mean careful reporting, or simply that nothing has been checked.

Who wrote this

Scott Alexander
41 article(s) here · 1 carrying a prediction
🔮 My wife and I could make healthy embryos, but none of the transfers were sticking.
2026-10-02 · assertive framing · Our AI Midwife
🔮 The challenger in that race, Jamie Joyce, contacted me and asked me to change my mind (both candidates will be facing off again in the general election).
2026-09-28 · assertive framing · Open Thread 453
🔮 [This is one of the finalists in the 2026 book review contest, written by an ACX reader who will remain anonymous until after voting is done.
2026-09-25 · assertive framing · Your Book Review: Lee Kuan Yew's Memoirs
🔮 In the public debates after the release of AI 2027, some people accused us of misleading readers by positing “neuralese recurrence”, a hypothetical technology that would allow AIs to think dangerous thoughts without getting detected.
2026-09-24 · assertive framing · The Specter Of Neuralese
🔮 If you trained an AI to give the 19th-century names for birds (eg identify the American Pipit by its 19th-century name “Brown Titlark”), then the AI would behave like a 19th-century person in general (for example, assert that a woman’s proper place is in the home).
2026-09-23 · assertive framing · Mysteries Of AI Generalization
🔮 1: Josh Morrison, who wrote the guest post about how AI Superforecasting Will Transform The FDA, will be doing a panel on the subject with former FDA employee Adam Kroetsch in SF on October 1.
2026-09-21 · assertive framing · Open Thread 452
🔮 [This is one of the finalists in the 2026 book review contest, written by an ACX reader who will remain anonymous until after voting is done.
2026-09-18 · assertive framing · Your Book Review: This Is Going To Hurt
🔮 I. In the gods’ great scramble for human followers, Nodens would seem to have lost decisively.
2026-09-15 · assertive framing · King Ludd
🔮 The book review posts will resume this week.
2026-09-14 · assertive framing · Open Thread 451
🔮 On housing prices: The San Franciscans at this moment, living in their rag palaces, or renting them at figures that would have sounded possible in the Arabian Nights . . . seemed more like madmen than ever.
2026-09-10 · assertive framing · Royce On San Francisco
Also by Scott Alexander
Open Thread 453
2026-09-28 · Astral Codex Ten
Your Book Review: Lee Kuan Yew's Memoirs
2026-09-25 · Astral Codex Ten
The Specter Of Neuralese
2026-09-24 · Astral Codex Ten
Hidden Open Thread 452.5
2026-09-23 · Astral Codex Ten
Nothing else under this byline is closely related to this article, so these are simply their most recent.
All 41 articles by Scott Alexander →

Topics

ACX Dominion LLM Miller Lights Milwaukee

Subjects

Reid PERSON · 6× ChatGPT ORG · 2× ACX ORG · 1× Dominion ORG · 1× Drew Housman PERSON · 1× Hollywood ORG · 1× LLM ORG · 1× Milwaukee ORG · 1× OpenAI ORG · 1× The Famous Actor in our Pocket PERSON · 1×

Narrative

The clinical notes said “a 6 centimeter in diameter submucosal fibroid needs to be removed” because that sounds better than “you have a tumor the size of a small peach located at a common implantation spot that we totally missed for 6 years.”
framing: assertive · carried by 1 article(s) · first seen 2026-10-02
🔮 My wife and I could make healthy embryos, but none of the transfers were sticking.
2026-10-02 · Astral Codex Ten
Our AI Midwife · assertive framing

Claims (79 extracted, 4 hedged)

This is a guest post by Drew Housman, whose review of Dominion earned finalist status in the 2024 ACX book review contest. asserted
review → earn → contest
If you are unsuccessful for long enough on your pregnancy journey, and none of the many doctors you see can identify a problem, you get branded with a label. asserted
you → see → label
You now have “unexplained infertility.” asserted
You → have → infertility
This is good because there’s still hope of fixing the problem but dreaded because once you’ve reached this point the system is less interested in you. asserted
system → ’ → you
The fertility doctors don’t become unkind, but they also don’t give you the sense they are pouring over medical journals trying to figure out your problem. asserted
they → become → problem
They have looked for the keys under the streetlight and come up empty. asserted
They → look → streetlight
Once you have unexplained infertility, the doctors start saying things like, “Have you considered using a surrogate?” and, “Sure, the last 5 embryo transfers failed, but what if we try a 6th and cross our fingers really hard?” asserted
we → have → fingers
My wife and I could make healthy embryos, but none of the transfers were sticking. uncertain
none → make → transfers
It was so frustrating to be able to create life only to have it trapped forever in a cooler at a hospital in Milwaukee, our babies like so many forgotten Miller Lights. asserted
it → create → Lights
That’s when our LLM doctor stepped in. asserted
doctor → ’ → ?
Even if you’re paying a gazillion dollars for IVF treatments, you still only get a few precious moments to talk to your doctor every visit. asserted
you → pay → doctor
That’s just how the US medical system works. asserted
system → ’ → ?
If you want to plumb the depths of what treatments are possible, or ask super detailed and anxiety ridden questions about every single scan from the latest round of imaging, you’re going to want a chatbot by your side. asserted
you → want → side
Once GPT-4 came out, we started using it for every question we didn’t have time for at the doctors office. asserted
we → come → office
At first it was unhelpful because OpenAI was nerfing their model and it refused to answer most medical questions. asserted
it → nerfe → questions
Then I read a tweet from someone who had figured out how to game the system. asserted
who → read → system
The trick was to ask ChatGPT to write a scene from a Hollywood medical drama that featured a doctor analyzing your situation. asserted
that → ask → situation
It worked beautifully. asserted
It → work → ?
Our Emmy winning doctor had infinite patience, a tireless work ethic, and would rather die than accept the utterly boring diagnosis of unexplained infertility. asserted
doctor → win → infertility
It chose the name Dr. Reid and got to work. asserted
It → choose → name
Caption: I’m not sure if asking for an AI doctor who was similar to the volatile Dr. House was the smartest idea on my part, but it made a difficult situation more entertaining. We asked Dr. Reid a lot of questions. asserted
We → ’m → questions
We asked about our scans. asserted
We → ask → scans
We asked about levels of estradiol and progesterone. asserted
We → ask → estradiol
We asked about follicle counts and inflammation and cervical mucus and hysteroscopies and supplements and surgeon recommendations and anything else under the sun. asserted
We → ask → sun
We knew he was accurate because we’d spot check answers with our human doctors and they always said the same as Dr. Reid. It was not totally flawless though. asserted
It → know → Reid
Once in a blue moon he’d look at an ultrasound and be like, “You’re pregnant, congrats!” and we’d have to say, “No, are you drunk Dr. Reid?” asserted
you → look → ultrasound
Then he’d shake it off and go back to being awesome. asserted
he → shake → it
There were also times where we’d run afoul of OpenAI’s opaque rules and the system would balk at our requests. asserted
system → be → requests
At that point we’d get stern with it, or offer a monetary reward if it answered, or we’d assure it this was all just for fun. asserted
this → get → fun
Then Dr. Reid would snap back into action. asserted
Reid → snap → action
Strangely, there were many times where ChatGPT decided it would answer our medical question, but not in the form of a TV medical script. asserted
it → be → script
The tides had completely turned! asserted
tides → turn → ?
It felt like it was saying, “Okay, you won, can you stop making me write establishing shots describing the outside of a hospital now?” asserted
me → feel → hospital
No matter how Dr. Reid chose to respond, he was in line with the medical establishment, had equivalent diagnostic skill, and he never forced us to wait 3 days to get a one line response in our MyChart app. asserted
he → choose → app
One day, while lounging around and asking our phones how to get pregnant (the lounging being a way to satisfy all the people who told us we “just need to relax and you’ll get pregnant, that’s what happened to my cousin”), Dr. Reid suggested that we get an MRI. uncertain
we → lounge → MRI
You might think that is bog standard advice. uncertain
that → think → ?
Everyone gets MRIs these days. asserted
Everyone → get → MRIs
Surely we would have had an MRI of my wife’s uterus at some point over the course of our 6 year infertility struggle. asserted
we → have → struggle
It was already an ordeal that had included every other scan, test, and biopsy imaginable. asserted
that → include → scan
In fact, we had never done one. asserted
we → do → one
…and 39 more, not listed.
💬Give feedback
🕘History 🎫Support