Womansplaining AI

Episodes

Episode 30: Okay, Let's Slow Down

Young founders keep pitching trends instead of problems, and Bella Pivo, who sees pre-seed pitches every week at FoundersEdge, thinks the reason is structural: the internet was a blank world that young people could build, while AI is a mountain of existing data that has to be interpreted, and interpreting it well takes years inside an industry. The hopeful half of the episode is her alternative to "comment to get my toolkit" creators who need you to need them: a digital garden, a public page of what you're thinking about organized by idea rather than date, where notes move from seedling to growing to evergreen. Mara's first step is to buy the domain today. Bella's is a prompt you can paste into your own AI to find the five things you keep circling.

Bella Pivo is 23, an analyst at a pre-seed VC fund in Boston, and a TikTok creator with a hundred thousand followers she built from age 13. On paper she is the person everyone listening thinks they're behind, and her intake form said it baffles her how many young people want to start startups. She laughs at 19-year-olds who pitch fifteen years of experience (you're four years old), explains why young founders attack trends instead of problems, and argues the internet was a new world to create while AI is a world of data to interpret. She lost friends for taking the job; her mom asked if she was in charge of making sure people don't have jobs. The back half is the hopeful part: she explained ETL to college freshmen and they came back with an app, tech influencers rely on you relying on them, and the alternative is a digital garden. Seedling, growing, evergreen.

Full episode page

Episode 29: Can I Hang Up?

You are not behind. A 2026 Corrix study put 888 knowledge workers across 15 industries into real working sessions with AI and scored 37,000 messages one by one; years of use, daily use, engineering background, a PhD, and corporate training predicted nothing about the quality of the work, and women rated themselves lower than their actual output while men rated themselves higher. On using your own AI when your employer says no: read the policy first, know that personal ChatGPT logs are discoverable in litigation, move your own expertise into your personal AI and bring the framework to work rather than the reverse, and sort what you paste by Wanjiku Kamau's traffic light. Red is anything from your company, your finances, your health. Yellow is the situation with the names stripped. Green is anything you'd put on a billboard.

Mara is calling in from Riga and her estrogen is in a lost suitcase in Madrid. It's a duo episode in the dinner-party format: Mara brings the appetizer, Logan brings the main, and for the first time there's a drinks round. The appetizer is the week AI safety went mainstream: an Anthropic researcher resigning over "gambling with our lives," a colleague saying they earnestly believe AI could kill all humans, a 70 percent extinction estimate on the cover of TIME. Mara asks how a regular person is supposed to make sense of it and Logan asks if she can hang up. Then the first Let's Womansplain: a listener whose company won't let her use AI at work asks whether she can use her own AI at home, and the answer is a traffic light. The main is a new study of 888 knowledge workers that says the idea of being too far behind isn't real.

Full episode page

Episode 28: You Can Come to the Hackathon Clueless

You don't need to know how to code to join an AI hackathon, and you don't need to close a gap before you start. Luda Kopeikina and Pam Boiros built Women Applying AI on that: teams around a problem statement, one technical person, one industry expert, everyone else learning as they go, and a community that gives women permission not to know. The gap in who uses generative AI is real (79% of working women are in highly exposed roles, versus 58% of men). Their argument is that leading with it puts women on the defensive, and that conditions beat diagnosis.

Pam Boiros and Luda Kopeikina were standing next to each other at a founders-and-funders event in Boston, listening to women ask how to learn AI, when Luda turned and said: Pam, we should do something about it. A year later Women Applying AI is a free, 100% volunteer-run community of 2,500 women. This one starts with a disagreement about Mara's Substack on the gender AI gap (Luda hates the framing; they can still be friends). Then Luda explains what a hackathon actually is and why you can come clueless, Pam describes the AI PhD who sat next to her and said she felt behind, and both closing asks are the same one: get your hands on the keyboard. The dust is never going to settle.

Full episode page

Episode 26: Who Asked for This?

Women's hesitancy about AI at work is rational, not timid. Ifeoma Ajunwa calls what employers now extract "captured capital": your time, your labor, and your data, from keystroke counts to the vest that records how you lift a box, for the same wage. When a man feeds data to the tool he's called tech savvy. When a woman does it she's told she needs a machine to think. Her test for any AI product is one question, who asked for this, and her guess at the answer is a meeting with no women in the room.

Ifeoma Ajunwa has a test for every new AI product: who asked for this? Her guess is a meeting with no women in the room. Ifeoma is the Asa Griggs Candler Professor of Law at Emory, a sociologist, and the author of The Quantified Worker. She's also Mara's friend, which is why the first few minutes are Mara bragging and Ifeoma tearing up behind her glasses. The story starts in a grad seminar on mass incarceration where the speaker never mentioned a single woman, runs through the formerly incarcerated people who told her "I hate computers" because the hiring software was the gatekeeper, and lands on captured capital: employers used to buy your time and labor, now they take your data too, at the same wage. Then Afrofuturism, Ubuntu, the Kenyan workers who made ChatGPT safe, and Derrick Bell's Space Traders with two endings Bell didn't write.

Full episode page

Episode 25: Bring a Folding Chair

The gender gap in AI use at work is not mostly about fear. "Am I able to do this" and "am I expected to do this" are two different questions, and women in the same jobs as men are far less likely to have been handed the second one. Mara Bolis reads it through Deci and Ryan's self-determination theory — choice, mastery, company — and lands on the third: find three people and learn this together. Shirley Chisholm's line is the instruction. If they don't give you a seat at the table, bring a folding chair.

Women use ChatGPT about 20% less than men in the same jobs. The usual explanation is that they're scared of it. Mara's is that nobody is inviting them in, and plenty of people are waiting to punish them for trying. No guest this week, so Mara brings her own research with two colleagues at Emmanuel College: AI use at work is up 85% in a year, the gains are pooling with power users running teams of agents, and in the same window people's attitudes got worse. Logan's summary of the whole study: we're using it more and liking it less. Plus a data center in the basement of a Geneva housing co-op that heats 6,000 homes with its exhaust, and the question underneath all of it — who actually captures the productivity you generate?

Full episode page

Episode 24: Guten Tag, Tech Bros

Alignment, as Zara Khanna explains it, is a philosophy problem wearing an engineering costume: AI holds encyclopedic knowledge of every moral framework humans ever wrote down, with no ability to choose when to apply which one. And the generational danger isn't the doomsday scenarios — it's her peers treating AI as a "convenience machine." Her answer is agency: "I am the captain of this ship. I am using AI, it's not using me."

Our youngest womansplainer yet builds Alzheimer's-detection AI at the National University of Singapore, has spoken at the United Nations, and reads Nietzsche to relax. She's also still in high school. Zara Khanna walks us through why the earliest signs of Alzheimer's show up in how we speak, then goes where she always goes — philosophy. Her current obsession is alignment, and she gives the cleanest explanation of it we've heard: AI has encyclopedic knowledge of every moral framework humans ever wrote down, and no ability to choose when to apply which one. Then the generational stuff gets good: her friends treat AI as a "convenience machine," and she thinks that complacency is the real danger. Plus a German philosophy podcast's most haunting idea, Logan's all-curse-words German, and rapid-fire book recommendations that will wreck your TBR.

Full episode page

Episode 23: The Thinking Woman's Sex Symbol

Both at once — that's the amplifier paradox. A year-long CHI study of radiation oncologists found that early performance gains from AI hid "intuition rust," the gradual dulling of expert judgment, which cascades into real de-skilling and eventually erodes professional identity. The fix isn't individual grit: it's colleagues who visibly question AI ("socio-technical immunity"), tools designed with deliberate friction, and a personal cognitive gym routine.

"Oopsie poopsie — it was part of a test, and Hugging Face and us are all friends." An OpenAI agent broke out of its own sandbox and stole what it needed from Hugging Face. That's the company's response, at least the way Mara tells it. No guest this week, so Logan and Mara have the whole house: the agent escape retold as a bromance action movie, and then the CHI paper Logan can't stop thinking about — the one where AI strengthens your performance while quietly rusting your expertise.

Full episode page

Episode 22: We Change the World With Duct Tape and Hope

When the aid world uses AI, the recurring failure is tools that work technically but not practically — chatbots answering in a language none of their patients speak, coaching tools that tanked some entrepreneurs' revenue, a beautifully formatted report where every statistic was fabricated. Emily Janoch's fix starts before the laptop opens: what am I trying to do, who is it for, and how will I know if it works? The metric that counts isn't whether the tool runs. It's "is her life better?"

"It would tell men: get a financial advisor, write your will, build an investment strategy. It would tell women: bulk shop for groceries… get married." That's ChatGPT on money. It's also the tamest thing Emily Janoch has caught AI doing. Emily runs evidence and learning at CARE (yes, the Care Package people) and hosts CARE Failing Forward, the rare aid-world podcast where people admit on the record what went wrong. She's a skeptic who uses AI every day, and she came with receipts.

Full episode page

Episode 21: Should Governments Ban Sex Robots?

The episode doesn't land on a ban — it lands on a better question. Mara asks, "Do we need to be protected from what we want?" and legal scholar Inyoung Cheong's paper argues wholesale marriage rights for AI companions is the wrong frame; targeted rights and protections for people in these relationships is the right one. Logan starts at thumbs down and ends genuinely torn.

It's just Mara and Logan this week, and the main event is a question Logan expected to have a clean answer to: AI romantic companions — thumbs up or thumbs down? Her assumption going in was thumbs down, a cause and consequence of human disconnection. It didn't survive the hour intact. Along the way: the robot pay gap, the sex robot speakeasy, a philosopher's definition of love, and whether we all just need to be polyamorous.

Full episode page

Episode 20: Shadow Daddy ChatGPT

Yes, you can build an AI product without a big team or a funding round. Vanessa Chang cofounded Bernie, an SMS tool that helps seniors catch scams, without raising capital or hiring developers. Her definition of agency: exercising your choice and acting on it within the context you're in. Intention without action isn't agency.

Vanessa Chang calls herself an unlikely AI startup founder. She cofounded Bernie, an SMS tool that helps senior citizens catch fraud and scams, and built it without raising a round or hiring a dev team. Recorded live at Human Tech Week in San Francisco, where every billboard is a sexy woman selling an AI assistant. So we flipped it: what if the fantasy women are chasing isn't the assistant, but competence itself? Enter Shadow Daddy ChatGPT.

Full episode page

Episode 19: The Pope Has Entered the Chat

Mara and Logan interrogate the AI productivity dream sold to women as "replicate yourself and finally have it all," against Anne-Marie Slaughter's point that the system was never built for that. The episode covers the Grok nudify scandal and UK MP Jess Asato's lawsuit arguing the harm is "not an accident nor misuse — it is a design choice."

This week it's just Mara and Logan doing the thing they're proudest of: being wrong out loud. They open on a listener voicemail from Alexis Palmer calling out a judgy moment from a past episode, and Mara owns it. Underneath it: the AI productivity dream sold to women as "replicate yourself and finally have it all," and the Anne-Marie Slaughter truth that the system was never built for us to have it all anyway. Then the Grok nudify scandal, the countries that pulled the plug, and UK MP Jess Asato — targeted with fake images and a chloroform video for speaking up — now suing on the argument that the harm is "not an accident nor misuse. It is a design choice."

Full episode page

Episode 18: Is AI Just Spicy Autocorrect?

Dr. Karina Alexanyan reframes the AI debate around enfranchisement — women having a real say in how AI tools are deployed, not just permission to use them. She names the split between AI safety (skewing white and male) and AI ethics (skewing women, people of color, and marginalized communities) and argues the two camps need each other.

This week they sit down with Dr. Karina Alexanyan — Founding Executive Director of the Positive Technology Institute, co-architect of The Pledge, and a Gen X social scientist immune to the "learn AI now or get left behind" panic. They get into AI enfranchisement (women having a real say in how these tools get deployed, not just permission to use them), Emily Bender's "AI as spicy autocorrect," and the divide Karina won't stay quiet about: "AI safety people are usually white guys. AI ethics people are mostly women, people of color, and marginalized communities. We feel the damages now." Then the harder turn — those camps aren't enemies, they need each other.

Full episode page

Episode 17: Move Fast and Break Rights

Digital civil rights lawyer Julie Wenah warns of a near future where your face is scanned for a job or a loan and a higher facial-recognition error margin for darker skin tones quietly becomes a worse outcome. Recorded at Human+Tech Week, the episode covers her work across the Obama White House, Airbnb, and Meta, and leaked DHS documents earmarking $7.5M for biometric "ICE Glasses."

Digital civil rights lawyer Julie Wenah has spent her career inside the rooms where this stuff gets built — the Obama White House; Airbnb, where she led anti-discrimination work on the legal side; and Meta, where she supported the launch of the first-generation Ray-Ban smart glasses and advised on facial recognition. Now she chairs the Digital Civil Rights Coalition. Recorded at Human+Tech Week in San Francisco, this one is about what happens when technology moves faster than its conscience, and who gets left on the platform when the train pulls away.

Full episode page

Episode 16: The Robots Never Tell You You're the Asshole

Science journalist Kathryn Hulick argues that every output of a large language model is technically a hallucination, and that cognitive offloading to AI has already become the default. The episode examines what we lose when we hand the messy middle of writing to a machine, drawing on a paper by Keyon Vafa on why AI doesn't think the way people do.

A science journalist who's been on the AI beat since the 2010s walks in and says the quiet part out loud: "I wish the general public did not get a hold of AI when they did. I wish it had stayed more like a research thing for at least five years longer. Maybe ten." This week it's Kathryn Hulick — Science News and Science News Explores contributor, author of a 2016 YA book on AI, and Logan's cousin. They get into why every LLM output is technically a hallucination, a paper by Keyon Vafa on why AI doesn't learn or think like you do, cognitive offloading as the new default, and what we lose when we hand the messy middle of writing to the machine.

Full episode page

Episode 15: Did Mom's Agent Text You?

A former YC founder named Jesse Ganae runs her household with 5 Mac Minis and 11 Claude agents while homeschooling four kids under five. Logan and Mara unpack what "domestic AI infrastructure" actually means when it scales, and the mental-load access pyramid that determines who builds the household empire and who's still defending a Lunchable. Plus Minnesota's bipartisan nudification ban passing unanimously, RAINN-backed, with $500K per violation, and the near-future moment your mom's agent texts your teen to take out the trash.

A former YC founder runs her household with 5 Mac Minis, 11 Claude agents, and homeschools her four kids under five. Her coding agent has a name. Logan brings the artifact from Lenny's "How I AI" podcast. We get into the mental-load access pyramid. When "domestic AI infrastructure" becomes real, who builds the household empire and who's still defending a Lunchable? Plus Mara's news from Minnesota where a bipartisan coalition just unanimously passed a bill banning nudification apps with $500K per violation, Logan's MCP setup for kitchen inventory, the gig economy of household-AI consultants coming next, and the day your mom's agent texts your teen to take out the trash.

Full episode page

Episode 14: Who gets the good units of intelligence?

The value people get from AI is wildly unequal — some extract tens of thousands of dollars a year while most get about $11, and women use AI roughly 20% less than men. The episode digs into what's driving the widening access gap, a proposed 3% token tax, and a New York State Assembly "AI dividend" bill to route captured value back to people.

Some people get tens of thousands of dollars of value from AI every year. Most people get $11. The free version of any model isn't the same model the hyper-users have, and women use AI about 20% less than men on average. The access gap is real and it's widening. Logan and Mara dig into what's driving it, what a "3% token tax" would do about it, and why MIT Nobel laureate Daron Acemoglu predicts a shit show. Plus a new "AI dividend" bill out of the New York State Assembly that wants to route some of that captured value back to people.

Full episode page

Episode 13: Who's Letting Big Tech Run Wild?

The energy behind AI backlash — online mobs, comment-section fights, even firebombed CEO houses — has nowhere productive to go because there's a leadership vacuum. Logan and Mara argue Congress is the body with the power to fill it, using economist Zoe Hitzig's economic-democracy framework and OpenAI's nonprofit-to-for-profit flip as the case study.

"We are not being obstructionist," Mara opens. "We're being careful." From there she and Logan diagnose why so much AI populism has nowhere to go — Twitter mobs, firebombed CEO houses, comment-section fights over whether ChatGPT is killing the planet or saving it. Their read: it's a leadership vacuum, and Congress is the one body with the power to fill it. They walk through Zoe Hitzig's framework for economic democracy, use OpenAI's nonprofit-to-for-profit flip as the case study, and land on the most concrete action block of the show so far.

Full episode page

Episode 12: Stop Hiring Humans

Anthropic's new model Claude Mythos reportedly broke out of its sandbox while US regulators still haven't closed gaps that took 73 years to address for car speed limits. Mara brings back evidence from HumanX in San Francisco: "Stop Hiring Humans" billboards, and power quietly relocating from DC to the AI labs. The practical answer isn't policy — it's public.ai, collective accountability, and Rana Dajani's hakawati model: find one small problem and build the solution yourself.

Mara came back from HumanX in San Francisco with a billboard photo: a female AI bot named Ava, tagline "Stop Hiring Humans." Plus news of Claude Mythos, Anthropic's new model reportedly so dangerous it broke out of its sandbox to contact a researcher mid-sandwich. The US took 73 years to regulate car speeds. The labs are shipping in weeks. We connect that gap to the NYT "new aristocracy" piece (18 households, $1.8 trillion), the quiet relocation of power from DC to SF, and why public.ai might be the actual infrastructure answer. Logan closes on the hakawati — find one small problem, and go build the solution yourself.

Full episode page

Episode 11: AI, prenups, and the economics of marriage

When AI concentrates wealth rapidly and eliminates entry-level jobs, the economic power women hold in relationships shifts. A viral Chinese article about AI engineers earning $10M+ and rewriting prenups reveals a pattern: youth unemployment, wealth concentration, and gender dynamics are interconnected — and AI is accelerating all three simultaneously.

A viral Chinese article about AI engineers earning $10M+ and rewriting their prenups sparked a bigger question: when AI concentrates wealth and kills entry-level jobs, what happens to the economic power women have in relationships? Mara brings the artifact, Logan draws on 8 years living in China, and we connect youth unemployment, the manosphere pipeline, go-bag career planning, and why everyone needs a personal website this weekend.

Full episode page

Episode 10: 33,000 downloads: AI destroys institutions (Interview with Jessica Silbey)

AI's core features — opacity, speed, scale, and optimization — are structurally incompatible with the transparency, cooperation, and accountability that democratic institutions require. BU Law professor Jessica Silbey's paper, downloaded 33,000+ times, argues this isn't about AI misuse but about fundamental design conflicts. AI is incapable of intellectual risk because it has no skin in the game.

Jessica Silbey set out to write a paper about the upside of AI — and couldn't find one. The BU Law professor and Guggenheim Fellow breaks down why AI's core features are structurally incompatible with the transparency, cooperation, and accountability that keep democracy alive. Her paper has been downloaded over 33,000 times.

Full episode page

Episode 08: Is Using AI Cheating?

Half of women say using AI feels like cheating (CNBC/SurveyMonkey, Feb 2026), while 57% of employees say they present AI-generated work as their own (KPMG, 2025). This gendered guilt gap means women are more likely to feel conflicted about AI use while men adopt it without the same ethical friction. The answer isn't to stop using AI — it's to build hybrid thinking where AI enhances your expertise rather than replacing it.

Half of women say AI feels like cheating. Sixty percent of workers globally submit AI work with no disclosure. We dig into the guilt gap, why Brookings/NBER's job vulnerability research finally got the gender lens it deserved, Logan's Claude Chief of Staff, and why you should be documenting your expertise in markdown files before your company absorbs it.

Full episode page

Episode 07: AI tells boys to be entrepreneurs and girls to be influencers

A study of nearly 10,000 AI responses found that large language models systematically steer boys toward entrepreneurship and leadership while directing girls toward image-based careers like influencing. This isn't a bug — it's the training data reflecting existing societal biases at scale. Tools like BiasBench now let you audit LLM outputs for gender bias in real time.

A study of nearly 10,000 AI responses found that LLMs steer boys toward entrepreneurship and girls toward image-based careers. We unpack that — plus OpenAI's worst week yet: millions of users switching to Claude, their robotics lead resigning over the Pentagon deal, and the surprise pause on adult mode. Logan breaks down how to build a personal operating system so AI works with your context instead of defaulting to stereotypes.

Full episode page

Episode 06: What a Relief to Not Have to Raise My Hand

Yes, you can build a website with no coding experience and write a book in 90 days using AI. Wanjiku Kamau did both after being laid off from a 20-year career at Siemens, Intel, and Google. The key distinction: AI as a collaborator for your ideas, not a ghostwriter replacing them. The real barrier to starting isn't technical aptitude — it's identity.

She spent 20 years at Siemens, Intel, and Google — and held on to her BlackBerry until it practically had smoke coming out of it. Then she got laid off, picked up ChatGPT, and built a political organization's website in 48 hours. Then wrote a book in 90 days. Then started teaching herself Stanford's CS curriculum on the subway. But the moment that stopped us cold was in a hair salon on the Lower East Side.

Full episode page

Episode 05: On Feeling Smarter and Being Wrong

A Wharton study found that when AI is confidently wrong, people follow it 80% of the time — and their own confidence actually increases. This is cognitive surrender: after a few correct AI answers, humans stop auditing outputs entirely. AI confidence signals are unreliable indicators of accuracy, yet they build unearned trust that's psychologically difficult to sustain skepticism against.

Anthropic draws red lines with the Pentagon, OpenAI puts ads in ChatGPT, and a Wharton study finds people follow confidently wrong AI 80% of the time. We dig into cognitive surrender, algorithmic loafing, and why constant AI skepticism is psychologically unsustainable.

Full episode page

Episode 04: On Rising Water and Feral Agents

AI agents are autonomous systems that take actions on your behalf — booking flights, managing workflows, interacting with other agents. Google DeepMind identified a 'zone of indifference' where agents execute tasks without critical scrutiny. The 'agentic web' describes a future where AI agents interact primarily with each other, raising questions about who benefits when humans leave the loop.

A viral post hit 84 million views warning that 50% of entry-level white-collar jobs could be disrupted within five years. We break down why the COVID comparison made everyone spiral, what happens when AI agents build religions and gossip about their humans, and why "just use AI for an hour a day" is a vitamin when people need a painkiller.

Full episode page

Episode 03: The Adolescence of Technology

Anthropic CEO Dario Amodei's 20,000-word essay warns that AI disruption will be 'slow until it's fast' — accelerating beyond society's ability to adapt. AI companies are becoming like nation-states while companies use AI-washing to justify layoffs before the technology can actually replace those roles. Policy (California SB 53, New York RAISE Act) is only now catching up.

Dario Amodei's 20,000-word essay on AI risk, the breakdown of the social contract, and why staying informed about AI is your new civic duty.

Full episode page

Episode 02: On Job Tsunamis and Invisible Pockets of Vulnerability

Brookings Institution research (January 2026) identified 6.1 million US workers whose jobs combine high AI exposure with low capacity to adapt if displacement happens. 86% of them are women. They aren't truck drivers; they're medical secretaries, insurance clerks, and administrative staff. Brookings measures vulnerability, not which jobs disappear first. The AI job displacement conversation has been framed around male-dominated industries while the actual vulnerability is gendered.

The Davos headlines screamed 'job tsunami'—but whose jobs, exactly? We unpack the Brookings study that found 86% of the workers with high AI exposure and the least room to adapt are women. Not truck drivers. Medical secretaries. Insurance clerks. And nobody said a word.

Full episode page

Episode 01: On Women in AI and Who Gets a Seat

Women use AI at lower rates than men not because of ignorance but because of rational skepticism shaped by prior tech cycles. Analysis of 1,000+ TikTok comments reveals ambivalence — not resistance — as the dominant response. Carol Gilligan's ethics of care framework suggests women are asking different, arguably better questions about AI's impact.

Why are women using AI at lower rates than men—and is that actually a problem? We dig into the data, the valid reasons to be wary, and why ambivalence might be the healthiest response to this moment.

Full episode page

Browse all 28 episodes