Dear friends,
This week marks my natural move into Deloitte’s Technology & Transformation team. I’ve been impressed by the company’s resources and our internal AI tools. Of course, it also means my free time for self-study has shrunk since returning to full-time corporate life. These days, I average about roughly 4 to 5 hours a week learning, researching, and writing these reflections.
It’s not a huge amount of time, but it’s consistent. And I can already feel the difference: I’m building more confidence in AI and technology, I’m wandering less, and I feel less anxious about IT than when I started. I’m still figuring things out, but I feel more grounded now.
I hope some of you who’ve followed me from the beginning are also practicing in your own way — whether quietly on the side or more actively. The point is not the hours, but the consistency. If we keep showing up, even with small steps, we grow together. That was one of the original purposes I shared in my first letter, and I want to keep encouraging you: stay curious, keep practicing, and don’t give up.
Excel’s =COPILOT(): impressive but limited
Last week I read two articles back-to-back. One said AI cannot replace Excel skills. A few days later, another trainer was promoting Excel’s new embedded =COPILOT() formula. I was impressed and excited and also shocked by the speed of AI’s evolution. At the same time, I reminded myself: How exactly does this work? What are the limits?
So I did what I now try to practice as a habit: validate. Don’t just accept what you read online — test it. Microsoft’s own notes confirmed it:
- You need a Copilot license, not just Office 365.
- It’s still in the Beta Channel (think of it as Microsoft’s “test kitchen,” not widely used by enterprises yet).
- There’s a usage cap (about 300 calls per hour — fine for normal analysis, but don’t expect to fill 5,000 rows at once).
- And most importantly: Microsoft itself warns not to use it for high-accuracy reports like financial statements.
👉 So the trick is not just to know the new features, but to know how to use them wisely and correctly. I have to admit that I am so impressed with this evolutionary change - Build AI directly into Excel. Definitely in my watch list.
Using AI wisely: 5 wise ways
One of the biggest limits of AI is what the experts call hallucinations — when it confidently makes something up. I found that Stanford professor Jeremy Utley’s talk on Stanford's Practical Guide to 10x Your AI Productivity is the perfect antidote.
Here’s my digest of his five wise ways:
- Treat AI as a teammate, not just a tool. Don’t just give it orders — coach it, critique it, collaborate with it.
- Let AI ask you questions. Invite it to clarify your goals, constraints, and audience before answering.
- Provide strong context & role assignment. Share your role, purpose, and style — sometimes even ask AI to “act as” a specific reviewer or coach.
- Go beyond the first idea. Don’t stop at “good enough.” Ask for multiple variations, push for more creative options.
- Use reflective critique. Ask AI to spot weaknesses in its own answers, then refine and iterate.
I couldn’t agree more. These match what I’ve been practicing: providing context, checking, adjusting, asking more. They help reducing hallucinations and making AI as a our assistant and coach instead of a replacement tool. You don’t want to be replaced, do you?
Staying human in a digital world
I also feel privileged to have recently attended a lecture by Dr Fiona Kerr, founder & director of The NeuroTech Institute. Her research spans neuroscience, psychology, engineering, and anthropology, and explores how human-to-human interaction changes when so much of our connection is mediated through screens and devices.
Her reminder was powerful: digital tools offer many benefits, but overuse can reduce the richness of human contact, raise stress levels, and subtly weaken the judgment that comes from real, embodied interaction.
That struck me as especially relevant for AI. For example, we now have AI drafting our emails or Teams messages or CVs. It’s efficient, but if we just let it do the talking for us, we lose the nuance, tone, and trust that come from genuine human conversation. Digital tools are exciting, but without our judgement and intentional human connection, we risk losing more than we gain.
What this means for careers
- For employers: Don’t just ask “Are you good at Excel?” Technical skills are still useful, but many are now automated or outsourced. Focus instead on the harder-to-teach skills: adaptability, transferable thinking, curiosity, creativity, and problem-solving. With those as a foundation, people can build business context quickly.
- For employees: If you’re mid-career and worried about being replaced, don’t be. Highlight your growth mindset, adaptability, and judgment. These remain your edge.
- For young professionals: Take advantage of your capacity to learn fast. AI helps you skip repetitive admin, but use that time to build the judgment and experience that will matter most later.
Remote work has changed how we communicate. Digital tools make us efficient, but efficiency doesn’t mean isolation. The winners will be those who adapt their way of thinking, embrace new collaboration styles, and continually upgrade both their technical skills and their mindset.
Takeaway
AI is not just about coding, prompts — it’s about context, curiosity, and conversation. Use it wisely, validate what you read, asking for the logic, the source data, and never forget that judgment (wear your auditor’s hat) and human connection are what make us valuable.
If you’re looking for powerful resources to grow your mindset in this space, I highly recommend following:
- Jeremy Utley (Stanford professor on AI-powered creativity & problem-solving)
- Dr Fiona Kerr (NeuroTech Institute, human-technology interaction & wellbeing)
Both bring unique perspectives on how AI and technology shape our work and our humanity.
📖 References for those curious:
- Microsoft’s official documentation on the COPILOT() function
- Stanford’s Jeremy Utley on Stanford's Practical Guide to 10x Your AI Productivity
- Profile of Dr Fiona Kerr, The NeuroTech Institute
✨ Friends, I’d love to hear from you: how are you figuring out your own AI journey? Have you practiced, experimented, or even struggled along the way? Please share your experience — we can all learn from each other.
Until next time —
Lydia
P.S. If this letter found you at just the right moment, I’d love to hear about it. Join my weekly letter list and let’s figure it out together — one AI-shaped step at a time. Join the weekly letter list.40+ and Figuring It Out
☕💌 If you’d like to fuel my next cup of coffee and keep this journey going, you can:
Your support keeps the ideas flowing and the coffee brewing ✨