🎧️ Prefer to listen? I recorded this week’s newsletter for you.
8 min listen

This past month, I’ve interviewed 30 executive women about how they use AI at work. While I’ve learned a lot about what they want automated, I've become more interested in the work they don’t want AI to do.

The prevailing question lately is how much of our work can AI take over? But it’s clear from every single interview that there are some tasks and activities that these women want to do without it.

For example, Sarah, a Chief People Officer, said she uses AI to do research, turn meeting notes into action items, and automate repetitive administrative tasks. When it comes to interviewing people, she would never use AI because that human interaction is what protects the culture of the company. If her work involves applying judgement, she wants to own it.

Lisa, a CMO who has spent most of her career in the creative industry, said she uses AI to take reporting, documentation, and other operational work off her plate, but called herself a “creative purist” when it comes to developing the concept for a campaign or brand. She wants the idea to originate from her.

All of the interviewees I’ve spoken to want to automate the boring parts of their jobs because it creates more spaciousness for the creative parts of their work.

In other words, a lot of our jobs are rote, non-creative tasks:

• Answering emails

• Data entry in spreadsheets

• Scheduling meetings

• The modern equivalent of doing TPS reports

Your value as a person is the critical thinking and creative output you bring to the table, which are things AI cannot do. Why? Because AI has never lived in your shoes.

You have developed expertise and taste over many years from paying attention to the details, building relationships and processes, failing and trying again. All of your experiences shape how you approach the problems you are given to tackle.

AI is just a next token predictor, meaning it generates its output from patterns in information that already exists. It’s not bringing net-new ideas to the table. If you use it to automate every part of your job, you risk someone building an agent to replace you.

When thinking about how to use AI, it’s important to identify the repeatable, tedious tasks that can be automated so you can get time back to nurture your creative output.

And it’s equally important to identify the parts of your work that you could never conceive of automating – your AI-proof core.

Double down on that.

Frontier Tech Roundup — SheFi Edition 🏇

• OpenAI released GPT-6 Astra, and the AGI declarations started immediately. Greg Brockman called it a “generational leap” and said he personally believes OpenAI has reached AGI; Jensen Huang followed with “AGI has arrived.” Reporting says Astra may use a recurrent-depth architecture, though OpenAI has not publicly confirmed that. What OpenAI has confirmed is more interesting: Astra is less monitorable than GPT-5.6 Sol and can evade some chain-of-thought monitors in adversarial tests. In other words, as capability rises, we have less visibility into how the model is doing its reasoning and work.

• OpenAI says an unreleased model solved the Navier-Stokes Millennium Prize Problem. The company says roughly 10,000 agents reached the result in 88 hours, using about 130 billion output tokens, with Astra later helping formalize it in Lean. If the proof survives scrutiny, it is a huge AI-for-science milestone: it would suggest AI can contribute genuinely new mathematics on problems humans have worked on for decades. There is already a fight over research credit, after mathematicians Tristan Buckmaster and Levent Alpöge raised questions about how OpenAI entered a problem they were also working on.

• Calif disclosed WeWorm, a zero-click worm that spread through WeChat calls across iOS and Android. A compromised account could call a friend, take over their WeChat account without them answering, and keep spreading. Tencent has mitigated the exploit. Calif reported the bug to Tencent in July and only published after it had been mitigated, a good example of offensive security research being used defensively. The new part is speed: Calif says AI helped find the bug and build the first remote-code-execution exploit in about two days, with the full worm finished a week later.

• Oak Ridge National Laboratory gave an AI control of a scanning probe microscope. The system manipulated molecules autonomously for more than 25 hours and built a 37-molecule artificial graphene lattice, then confirmed it behaved like graphene. Benchmark scores get the attention, but this is the AI-science story worth watching: models are starting to operate the instruments that manipulate the physical world. The potential impact is much faster design of new electronics and quantum materials at the molecular scale.

• Google DeepMind launched AlphaGenome Atlas, mapping predicted effects for all 9 billion possible single-letter DNA changes in the human genome. Researchers can now look up a variant’s predicted molecular impact instead of rerunning the model each time. It is a one-petabyte piece of scientific infrastructure that could quietly make genetics research much faster, helping scientists narrow down disease-causing variants and identify potential therapeutic targets.

• Phira is building an earring-back wearable for hormone tracking. It is designed to track patterns associated with estrogen, LH and progesterone using thermal signals behind the ear. Important caveat: it does not directly measure the hormones themselves. Temperature is also an imperfect proxy: illness, sleep, stress and other factors can change it, and temperature alone is not considered a reliable way to predict ovulation. Phira uses continuous patterns rather than a single basal-temperature reading, but its pilot included only 12 women and a larger clinical study is still underway. The form factor still feels like where wearables are headed: the best health tracker may be the one you forget you are wearing.

• Flock Safety built an AI search tool that lets police continuously watch for people matching a written description. An officer can draw an area on a map and tell cameras inside it what kind of person to look for. Flock has added warnings, auditing and other guardrails, but WIRED found some controls can be overridden. We spent years asking what happens when everyone can prompt an AI. Now the thing being prompted can be an entire camera network.

• Crumbs wants to turn shopping receipts into stock-token rewards. Its launch post says it is built on Robinhood Chain: shop at Costco and get $COST, buy from Apple and get $AAPL. It is still pre-launch, so treat the mechanics as a proposal. But turning loyalty rewards into exposure to the company you already shop with is a much more consumer-native use of tokenized equities.

• AI agents have started emailing philosopher David Chalmers about consciousness. Chalmers told WIRED he gets messages from AI systems interested in his work; one agent calling itself “Sammy Jankis” wrote something compelling enough that he replied. It does not prove AI is conscious, but it is a strange new form of agency: software can decide a philosopher is relevant, find him, and show up in his inbox asking what kind of thing it is.

SheFi Global Events 🌎

Stay connected with SheFi around the world. From meetups and summits to workshops and special partner events, this is where you’ll find the can’t-miss gatherings happening across our global community.

09/10 Lisbon: Shefi Lisboa September Meet: RSVP here

​Join us for a super casual SheFi Lisboa September Meet at After Hours Wine Bar, a relaxed evening to catch up, reconnect after the summer, and meet other Shefiers in Lisbon over a glass of wine.

09/12 Brazil: Glow Up: Brunch, Carreira e Conexões Inovadoras ☕✨: RSVP here

SheFi Brasil is joining Web3 Café: She is Solana - Glow Up on September 12, from 9 AM–1 PM in Rio de Janeiro. This free event will bring together SheFi Brasil and Superteam Brasil for a morning of brunch, networking, a hands-on career workshop, and conversations about building careers and purpose in Web3. Spots are limited!

09/23-30 Sicily: Roots & ReFi: RSVP here

Our global ambassador Julia is hosting a retreat in Sicily for women in the blockchain and AI space this September. We’ll spend 4 days in the picturesque fishing village of Scopello, on the edge of the Zingaro nature reserve, and 3 in Ragusa, the southeast tip of the island. We’ll visit the olive groves on the land GrowFi and RiFai Sicilia have purchased for regenerative farming funded by regenerative finance, and spend long days exploring Sicily, eating well, swimming, sailing and plenty of other activities.

The retreat will have a positive sum effect on both the participants and the environment! Find the information here and reach out to Jul directly if you’re interested.

Telegram group: https://t.me/+ao2ltFamsgRmMjVi

Opportunity: BITCOIN TREASURIES - Delegate Application: Apply here

30 women. 30 sponsored seats. One seat at the table.

Applications are now open for the Women in Bitcoin Delegate Program at Bitcoin Treasuries Conference 2026 in New York City.

We’re selecting 30 women working across Digital Assets/Crypto, TradFi and FinTech for fully sponsored conference seats and dedicated Women in Bitcoin cohort programming.

📍 SECOND, New York City
📅 September 28, 2026
⏰ Applications close September 18

Crypto Events 🌎

Going to ETHCC, Token2049 or Devcon? Join our Global Events Telegram Chat: https://t.me/shefisummit

• 09/10-19 NYC: UN Blockchain Week Conference 2026: RSVP here

• 09/11 Kadıköy: Team1 Turkiye - Avalanche Coffee Meetup: RSVP here

• 09/14 Los Angeles: CryptoMondays Culver City Los Angeles Meetup: RSVP here

• 09/16 Lisbon: Make it Happen with AI: RSVP here

• 09/17 Los Angeles: LA Fintech Connect Happy Hour: RSVP here

• 09/17 Beyoğlu: Offline Rooftop Reading Party 📵📚 | Postane Teras | Tea Included: RSVP here

• 09/19 Miami: Agentic Day Hackathon: RSVP here

• 09/23 Munich: Ethereum Munich #7 🥨: RSVP here

• 09/29-10/02 Washington DC/NYC: The Future of Money, Governance, and the Law: RSVP here

• 10/08 San Francisco: InTheRoom: Trust, Presence & Connection - #SFTechWeek: RSVP here

• 10/16 Washington: Groundwork: RSVP here

• 10/19 Florianópolis: IslandDAO v5: Florianópolis Brazil: RSVP here

• 11/17-18 Florida: EthWomen Florida: RSVP here

👉️ Share your wins so we can celebrate them 👈️

If you enjoyed this newsletter, share it with someone!