Artificial intelligence is totally changing how we think, learn, and tackle problems. This post pulls together Vivienne Ming’s thoughts on shifting parenting and education from just passing down facts to building up the skills that AI can’t easily copy: creativity, curiosity, and gritty problem-solving.
Ming says deep learning models thrive on exploration and making mistakes, but schools and culture often treat errors like the enemy and reward only getting things right. So what can parents and teachers do if they want humans to stay essential, even when machines can spit out answers in seconds?
Rethinking learning in the AI era
AI now handles a lot of routine mental work, so education needs to move past memorization. We should focus on teaching minds to design, critique, and improvise.
This shift takes real effort. Kids need environments that reward risk-taking and encourage them to challenge tech, not just nod along with whatever a machine spits out.
If we start seeing success as growth, students are way more ready to guide AI tools instead of getting pushed around by them.
From correctness to growth
Research on deep learning links trying new things with real learning and what Ming calls “productive failure.”
But schools usually treat mistakes as failures, not as useful data. Growth-focused teaching means letting students try, reflect, and get feedback that sharpens their judgment.
Ming wants classrooms to feel more like labs. Curiosity should spark questions, and mistakes should show effort and resilience, not just error.
Failure as fuel: the Failure Resume
The Failure Resume is a monthly family habit that flips mistakes into something valuable. By writing down their challenges, kids learn to explain what they tried, what went wrong, and what they’d do differently next time.
This isn’t about wallowing in errors—it’s about building a story of learning when things are uncertain. Over time, risk-taking becomes a positive, trackable thing, not something to hide.
Key elements of a Failure Resume include:
- Dates and contexts for each stumble or dead end
- What the failure reveals about assumptions or gaps in understanding
- Concrete next steps or experiments to test improved approaches
- Evidence of progress over subsequent attempts
Engineering serendipity at home
Ming points out that elite universities offer something special: random conversations, a mix of perspectives, and tough, messy problems. She suggests families can create similar environments at home.
Set up spaces and routines that let kids bump into new ideas and open-ended challenges. The goal? To help them enjoy exploring and get comfortable with not knowing all the answers.
Practical steps to engineer serendipity at home include:
- Set aside a dedicated tinkering corner where curiosity can roam freely
- Provide mixed magazines, diverse books, and project materials to broaden inputs
- Offer broken objects and unsolved problems that invite hands-on investigation instead of quick fixes
AI as a partner, not a substitute
Ming warns against letting large language models do all the hard thinking for kids. If they just use these tools passively, their independent thinking can fade.
Instead, students should act as active critics, treating AI models as “brilliant but naive” partners who need to be questioned and challenged. It’s not about beating AI—it’s about sharpening human judgment by working alongside it.
Nemesis Prompts and first-draft workflows
Ming suggests a workflow where a child writes a rough draft before asking AI for help. Then, they use a tough, critical prompt—the Nemesis Prompt—to push the model’s answers and force deeper thinking.
This helps learners move beyond just accepting machine output. They start assessing credibility, relevance, and bias.
In the end, you get a group of Chief AI Critics—people who use AI to refine their own ideas instead of just taking whatever the machine says as gospel. Isn’t that what we really want for the next generation?
What this means for families and educators
The message here is pretty practical: create spaces and habits that reward risk, curiosity, and critical engagement with AI. Kids need to stay indispensable in a world where information’s just a click away.
When we foster failure-friendly learning and give kids real chances to mess up, they start to develop real resilience. Deliberate exposure to all sorts of new things and the habit of questioning AI helps learners steer technology with judgment and a bit of imagination.
Takeaways for parents and educators include:
- Design spaces and routines that encourage hands-on tinkering and open-ended questions.
- Normalize recording and learning from mistakes through a Failure Resume.
- Expose children to diverse inputs and ill-posed problems to build resilience and adaptability.
- Use AI as a tool for enhancement, not replacement—challenge models with Nemesis Prompts and require a human first draft.
- Nurture Chief AI Critics who actively interrogate and refine AI outputs.
Here is the source article for this story: Stop teaching kids skills that will be obsolete in 20 years—how I’m raising kids AI can’t replace: They ‘have an advantage,’ says expert