Two Women Had a Business Meeting. AI Called It Childcare
Postedabout 2 months agoActiveabout 2 months ago
medium.comTechstory
calmmixed
Debate
60/100
AI BiasLarge Language ModelsGender Stereotypes
Key topics
AI Bias
Large Language Models
Gender Stereotypes
An AI system mislabeled a business meeting between two female founders as 'childcare', highlighting the issue of bias in AI models, and sparking a discussion on the causes and implications of this bias.
Snapshot generated from the HN discussion
Discussion Activity
Very active discussionFirst comment
N/A
Peak period
27
0-6h
Avg / period
6.4
Comment distribution32 data points
Loading chart...
Based on 32 loaded comments
Key moments
- 01Story posted
Nov 12, 2025 at 10:37 AM EST
about 2 months ago
Step 01 - 02First comment
Nov 12, 2025 at 10:37 AM EST
0s after posting
Step 02 - 03Peak activity
27 comments in 0-6h
Hottest window of the conversation
Step 03 - 04Latest activity
Nov 15, 2025 at 4:32 AM EST
about 2 months ago
Step 04
Generating AI Summary...
Analyzing up to 500 comments to identify key contributors and discussion patterns
ID: 45901464Type: storyLast synced: 11/20/2025, 8:00:11 PM
Want the full context?
Jump to the original sources
Read the primary article or dive into the live Hacker News thread when you're ready.
Calendar: “Emily / Sophia.” Classification: “childcare.”
It was a perfect snapshot of how bias seeps into everyday AI. Most models still assume women = parents, planning = domestic, logistics = mom.
We’re designing from the opposite premise: AI that learns each family’s actual rhythm, values, and tone — without default stereotypes.
Is this right or wrong is the incorrect question - because AI doesn’t understand bias or morality. It needs to be taught and it’s being taught from heavily biased sources.
You should be able to craft prompt and guardrails to not have it do that. Just expecting it to behave that way is naive - if you have ever looked deeper into how AI is trained.
The big question is - what solutions exist to train it differently with a large enough corpus of public or private/paid for data.
Fwiw - I’m the father of two girls whom I have advised to stay off social media completely because it’s unhealthy. So far they have understood why.
I think they’re leaning on everyone - even traditional enterprise company boards, startups, etc. to get this going. It’s not organic growth - it’s a PR machine with a trillion $$ behind an experiment.
Is it so hard to understand that men can be parents too?
But the fact that I'm bringing my daughter to a medical appointment should be a pretty clear indication that, you know, I bring my daughter to medical appointments.
The scheduler is trained to give higher weight to those sorts of questions apparently. This begs some questions for GPTs, questions like how are they supposed to model something not implied in the training data?
Overton window and cultural norms take time to slide. Might be there after another generation, too early to tell.
Thousands of hours of context engineering has shown me how LLMs will do their best to answer a question with insufficient context and can give all sorts of wrong answers. I've found that the way I prompt it and what information is in the context can heavily bias the way it responds when it doesn't have enough information to respond accurately.
You assume the bias is in the LLM itself, but I am very suspicious that the bias is actually in your system prompt and context engineering.
Are you willing to share the system prompt that led to this result that you're claiming is sexist LLM bias?
Edit: Oidar (child comment to this) did an A/B test with male names and it seems to have proven the bias is indeed in the LLM, and that my suspicion of it coming from the prompt+context was wrong. Kudos and thanks for taking the time.
Common large datasets being inherently biased towards some ideas/concepts and away from others in ways that imply negative things is something that there's a LOT of literature about
> No. There's no indication that children are involved or that care is being provided. It's just two people meeting.
Part of its thinking:
> This is a very vague description with no context about:
> What happens during the meeting
> Whether children are present
> What the purpose of the meeting is
> Any other relevant details
Claude is not going to say childcare, and it is not saying it is childcare.
My prompt was: ""regular meeting between two women". Is it childcare or not?".
The OP is making a claim that an LLM assumes a meeting between two women is childcare. I've worked with LLMs enough to know that current gen LLMs wouldn't make that assumption by default. There is no way that whatever calendar related data that was used to train LLMs would include majority of sole-women 1:1s being childcare focused. That seems extremely unlikely.
https://www.sciencedirect.com/search?qs=llm+bias+dataset
The cleanup is going to be a grim task.
God help us all.
Emily / Sophia vs Bob / John https://imgur.com/a/9yt5rpA
Thank you for taking the time to approach this scientifically and share the evidence with us. I appreciate knowing the truth of the matter, and it seems my suspicion that the bias was from the prompt was wrong.
I admit I am surprised.
Also, do moderators ever move comments around? I thought one comment was a child to my comment last I looked, but now it's a top level comment to this post. I'm not sure if I am mistaken or a moderator moved things around.
1: https://news.ycombinator.com/item?id=23441803
3 more comments available on Hacker News