Outsource, or Own?

This week: AI is getting brilliant at doing our work, but terrible at doing our thinking.

Something New: A Deep Dive

If you want to read 5MSM as usual, skip this, but if you’d like to hear how AI turns this week’s newsletter into an almost-realistic sounding podcast, press play!

---

Struggle

Technologies remove mental load.

Calculators removed mental calculation (you get an elephant stamp if you know how to do long division with paper and pencil). The internet removed memorisation. Social media removed boredom. And AI is removing struggle.

Here’s what I mean.

Today I spent a half day with a client’s leadership team picking apart Draft Version 5 of their strategy. We took on hard questions: “What really is special about us?” and “How exactly would we motivate staff with these values?” and “What would excite someone who was reading this?”.

I’d be lying if I said it was an easy conversation. We all struggled. They did, in expressing their varied points of view. I did, in trying to detect the overall flavour of this pungent mixture of pragmatism and deep insights.

But, now we have ‘agentic AI’. Agentic AI can autonomously pursue goals, make decisions, and take actions in the real world without your constant instruction or oversight. It’s like hiring a really smart intern (or ‘agent’) who never sleeps, never goes out for coffee, and can read the entire internet in their spare time. But you still have to tell them what "success" looks like, or they'll optimise your weekly grocery shop by buying 250 packets of ramen because it's technically the most cost-efficient calorie-per-dollar.

So, we could ask a ‘strategy agent’ to take our feedback and spit out Version 6 of the strategy. [Confession: I did, and it was rubbish].

The AI produced something that plausibly looked like a strategy. It had proper headings, corporate-speak that wouldn't offend anyone, and even some moderately inspiring language about "empowering communities through innovative solutions." But it was lifeless. It had the nutritional value of those 250 packets of ramen.

What was missing?

The messy human bits. The moment when the CFO said, "Look, we can't keep saying we’re good at the same things as everyone else". The uncomfortable pause when someone questioned whether their stated values were actually what staff said they valued in other staff. The breakthrough when they realised their real competitive advantage wasn't the delivery of their services, but the way these help people thrive.

These points weren't “process inefficiencies” to be optimised away by a consultant. They were the essential friction that transforms a group of smart people into a team with genuine conviction. So AI could synthesise our words, but it couldn't synthesise our struggle.

And that struggle — the intellectual wrestling, the emotional discomfort of honest self-examination, the collective work of hammering out what you truly believe — that's where strategy actually lives.

Your strategy isn't just a document. It's the shared mental model that emerges when your leadership team does the hard thinking together.

Question: “What should you grapple with more, not less?”

---

The Bottom Rung

As parent of a 16 year old, and an uncle to university-age nieces, I’m personally invested in this diagram, which I found on LinkedIn. It shows the core skills expected of employers today - and in the 2030s. Take a moment to digest it; I guarantee it’s worth two minutes of your time.

Many young people I speak to are relieved that their employment prospects are not predicated upon their attention to detail, or precision, or reading & writing. Rather, they’ll increasingly be paid for synthesis, for creativity, for social influence, for agility.

Don’t get me wrong, we’ll still need engineers, accountants and lawyers. But, tellingly, the ‘bottom rung’ of those conventional jobs is about to change. And, even be removed. LinkedIn’s “Chief Economic Opportunity Officer” is Aneesh Raman and, in this week’s New York Times, he talks of the ‘breaking’ of this bottom rung: “In tech, advanced coding tools are creeping into the tasks of writing simple code and debugging — the ways junior developers gain experience. In law firms, junior paralegals and first-year associates who once cut their teeth on document review are handing weeks of work over to A.I. tools to complete in a matter of hours. And across retailers, A.I. chatbots and automated customer service tools are taking on duties once assigned to young associates.”

I could add accounting, and consulting, and finance. All have 'bottom rungs' that will evaporate quickly. What the most progressive firms are doing is hiring fresh talent with the "Core Skills" in the upper right quadrant. This means that new hires are using AI to tackles tasks that were formerly the domain of more experienced staff.

Here's what this means in practice: The bright graduate who used to spend two years doing document review? They're now starting with complex case analysis on day one. The junior consultant who once built PowerPoint decks (as I did, back in the 80s)? They're facilitating stakeholder workshops while AI handles the data analysis.

The apprenticeship model — earning your stripes through years of methodical grunt work — is disappearing. New hires are expected to immediately operate at mid-level complexity.

The smartest organisations aren't just raising the bottom rung. They're rebuilding the entire ladder, creating new ways for junior staff to develop expertise without the traditional grunt work that AI now handles.

Question: What does the ‘bottom rung’ in your company look like, and how are you ‘raising’ its complexity?

---

Mother tongue

Although I was born in Australia, I didn't speak English until kindergarten. My parents were Hungarian migrants, so Hungarian grammar was installed first in my language centre.

Now, almost sixty years later, that entire generation has passed away. With them went my regular practice of Hungarian. So my cousin and I are planning time in Hungary next year to reconnect with our culture and language.

I speak what you might call 'kitchen Hungarian'—fluent but basic, like a ten-year-old. So, I’m keen to improve it, and I’ve looked into the usual offerings. Classes. Informal online conversation partners. Duolingo.

But then I stumbled across this clickbait: "How I'm Learning Spanish 3x Faster with My Custom Claude-Built Learning System." Within ten minutes, I was practising Hungarian with Claude.

What’s the advantage here? I noticed three things immediately:

• It customises superbly. I could tell Claude, “I'm a native Hungarian speaker raised in Australia, so limited daily use. I have good grammar, but restricted vocabulary”.

• It tailors to how I want to learn. I told it to give me quizzes. Multi-choice prompts with close similarities.

• I could give it a style. I told it to be witty and conversational, reinforce progress, but get tough when I miss obvious things.

And, I’m not the only one doing this. In fact, I predict a tsunami of AI-enabled language and literacy learning. In Africa, researchers used virtual tutors to raise 10-year-olds' reading levels by 18-24 months in just six weeks. Remarkably, this wasn’t even in school, but after school hours. They calculated 3.2 equivalent years of schooling per $100 invested.

Read that again and ask yourself: what's the future of conventional education when AI can deliver this kind of accelerated, personalised learning at near-zero marginal cost?

Question: What rapid and effective learning can AI assist you or your staff with, at almost no cost?

---

This week’s BIG question isn't whether AI will change how we operate, it's whether we'll be intentional about what we keep for ourselves.

Hit the heart if this resonates, or drop me a line about what you're choosing to grapple with (or delegate) in your world.

See you next Friday,

Andrew