The Transition to
Intelligent Business
How to transform people and companies to lead and thrive, not just survive or be replaced by AI
The blacksmith
“Is AI going to replace me?”
What do I bring that
the machine does not have?
What makes us different makes us better.
AI does three things at once
tasks
people
work
Simultaneous — even within a single job, even within a single day. All three camps are right at the same time.
Real, but concentrated
- ~40% of global employment exposed to AI (IMF). Amazon, Microsoft, Salesforce, IBM already cite it in layoffs.
- And yet: “no discernible disruption” 33 months after ChatGPT (Yale Budget Lab).
- The pain is real, but concentrated in specific places — not spread evenly.
AI compresses the skill gap: it gives someone junior a senior-level output. Brynjolfsson, Stanford
Most reinvested — in new capabilities, cybersecurity, R&D and reskilling their people. EY, AI Pulse Survey
We didn't lose work; it changed shape. But progress is neither automatic nor even. David Autor
It's not in the destination.
It's in the transition.
The question is not which camp is right. It is: during the transition, who pays the cost?
The Company
Augmenting as competitive strategy — a strategic argument, not a moral one.
The model is a commodity.
The context is the advantage.
“If your advantage is we have AI, you have no advantage: you hold the same key as everyone.”
The advantage lies in what generic AI does not bring: your industry, your data, your judgment, your customer relationship. AI turns that context into transformative capability.
the same thing — easier, faster, cheaper
More, different and better
Everyone's quick win is nobody's advantage: efficiency evaporates into price. Optimizing is linear and shared; reinventing is exponential and defensible.
The saving is not the victory.
It is the raw material.
What will you do with the freed-up time, budget and potential?
Don't book it as profit or let it evaporate: analyze, explore, decide and reinvest it in doing more, different and new. Change it, but change it well.
While you analyze,
they've already started.
AI erases distance: someone on the other side of the planet can serve your customer.
Those not yet in your market who will chase the same opportunities.
Decide the what with judgment — but move now. Build, don't prepare.
Cutting is easy to measure today and costly to pay for tomorrow. Run the 5-year NPV, not the quarter's saving. EY
Frameworks,
not tools
A stray tool gets copied by anyone tomorrow. A replicable framework — a systematic way of applying AI to your process, that your people adopt, adapt and scale — is an asset.
30 years turning innovation into models others repeat. The ROI isn't in the tool of the month; it's in the framework that outlives it.
If AI absorbs the junior tasks, we eliminate the learning: the talent bench thins. A 5-year problem disguised as efficiency. Goldman Sachs
Think in three horizons, not one
Build the people, the institutions and the frameworks. Whoever invests now compounds advantage.
Pressure to automate intensifies; the first-job crisis becomes visible.
New roles we don't yet know how to name. The winners treated it as a design challenge.
Context advantage
+ Small AI
Small AI — mobile, low-infrastructure — leapfrogs in health, education and finance. In a market the giants understand poorly, local context is a defensible advantage against giants with more capital and less context.
Should we ban AI for our staff?
The genie is out of the bottle. Ban it → they use it on the phone or at home and copy-paste: “shadow AI”, more vulnerable.
Enablement with governance, not prohibition.
“The Knowledge”
Since 1865, driving a black cab in London requires passing a legendary exam.
TfL calls it, in its own word, “encyclopedic” knowledge: the 320 Blue Book runs, within six miles of Charing Cross, from memory and with no technology.
It changes their brain
Their posterior hippocampus is larger than average and grows with years behind the wheel (Maguire, PNAS 2000). And across ~9 million deaths, taxi and ambulance drivers had the lowest proportion of deaths from Alzheimer's — which the authors call “hypothesis generating,” not conclusive. The BMJ, 2024
And what does the customer get?
Take away the pride and the craft — both real — and the business question remains: of those 3 or 4 years, how much reaches the passenger?
The passenger just wants to arrive — and GPS and Waze do that. No disrespect to the trade: what part is advantage and what part is nostalgia?
Not scrapping: rebalancing
Lower the encyclopedic part and reinvest that time in what a GPS doesn't give:
- When to ignore the GPS — and why.
- Alternate routes for emergencies, by time of day, by type of visitor.
- Manner, service and problem solving.
- AI fluency for what memory no longer has to carry.
Which customer are your roles designed for?
Are your job requirements — and your own development plan — built for today's customer, or for the one from twenty years ago?
This week: take one role and split it into three columns — what you require out of tradition, what the result requires, and what AI already frees up. Whatever stays in the first column is your Knowledge: decide, with judgment, how much to keep and how much to reinvest.
The Professional
Not just executives: professionals whose own work is changing. The portfolio career.
a ladder you climb
a network you navigate
The old script — degree, company, promotion, retirement — assumed a stable world that no longer exists. Each node is a skill, an experience, a domain.
“What do I want to be?”
What problems
do I know how to solve?
“What do I want to be?” anchors you to a job that can disappear. “What problems do I solve?” travels with you from one node to another, whether or not the title survives.
AI-proof skills
- Critical thinking
- Creativity
- Emotional intelligence — the most AI-proof
- Systems thinking
- Collaborating with AI
The OECD places AI at a low level in social interaction, metacognition and physical manipulation.
No PhD needed. Fluency is knowing where AI shines in your domain, where it fails and how to work with it. Within everyone's reach in this room.
The T-shaped professional
Whoever translates across domains — business with data, law with product, operations with the model — becomes indispensable. Build at the intersections.
Build,
don't prepare
“Every week you spend preparing is a week someone else is already launching and learning by doing.”
Pick 2 tools in your domain and start producing in 90 days. A certificate without application is worthless — context is king.
Value in augmented terms
“I produce X with AI assistance in half the time” — the professional narrative of 2026. Stop measuring hours; measure augmented output.
The premium is no longer on producing the draft, but on validating, refining and contextualizing it. The “human in the loop” is a skill with a name of its own.
AI as a partner,
not a tool
Most leave 80% of the value on the table: they use it as a secretary (the same thing, faster), not a strategic partner.
Iterates 15× without losing its edge: analyzes document 100 with the attention of document 1.
Faster and deeper: 10 h of preparation in 1.5.
Juguetón: 30 concepts → 3 brilliant ones that would never have occurred to me.
“How do I do the same thing faster?”
What can I do
that used to be impossible?
Email → communication strategy. Slides → persuasive narrative. Data → strategic intelligence. You don't use AI to do your work: to do work better than you could alone.
- Give it rich context — who you are, for whom, your goal.
- Give it freedom to contribute — don't reduce it to transcribing.
- Converse, don't ask — an extended dialogue, not a transaction.
- Direct its thinking style — how it thinks, not just what.
- One-shot / few-shot — give it examples of what works.
- Specify the output format.
- Ask it for clarifying questions before answering.
Five thinking styles
- Chain of thought — step by step · logical decisions
- Tree of thought — several branches · strategy and scenarios
- Stream of consciousness — free, unstructured · creativity
- Devil's advocate — challenges your assumptions · validate decisions
- Socratic method — asks you questions · clarify your thinking
Context + clarifying questions → tree (15 strategies) → devil's advocate → chain → 3 formats (CEO, team, one-pager). Deeper, more validated, ready to use.
Choose your words:
each one carries worlds
Language is itself a compression algorithm. Telling your partner you're “wiped out” conveys that you don't want to go out or see anyone —without another word—.
It's the same with AI: certain terms pack enormous, multidimensional meaning. Ask for something “with the quality of a Cannes, Sundance or Oscar-winning film,” “worthy of a Pulitzer,” or “with the rigor of a peer-reviewed scientific journal” — and watch quality scale in several directions at once.
Build without coding — but audit
Almost anyone can create apps, dashboards and functional prototypes just by describing them.
The antidote: ask it for a “cybersecurity audit” of its own code. Make it a habit for everything you build.
The magic of the gap analysis
Ask the AI for a gap analysis of anything you have or are preparing —your strategy, your proposal, your operation, your brand—.
It will thoroughly compare it against best practices and hand you a report of shortcomings, risks, opportunities and concrete suggestions to improve. One of the most productive techniques: it reveals the invisible and brings industry best practices to your table.
Prompts that write prompts
Not sure what to ask, or how to ask it? Ask the AI to help you craft the perfect prompt for what you need.
You can refine it in conversation until the prompt is right —and then tell it to run it. “Help me write the best prompt for [your goal]; ask me whatever you need, and when it's ready, run it.”
100+ things you can do with AI
An open catalog of practical ideas to explore and bring into your business.
100.cemi.ai/things.html — explore the business section especially, and pick one idea to try on Monday.
Three warnings
- 1Always verify. It hallucinates and biases; your judgment is essential. It amplifies your capacity, not your responsibility.
- 2Don't over-delegate the strategic. AI informs the decisions; you make them.
- 3Don't lose your voice. If everyone uses it the same, everyone sounds the same.
Augmented Intelligence
= amplification
human + AI = what neither alone
It's not AI replacing us, nor us ignoring it. AI doesn't do the work for you: it lets you do work better than you could alone. Collectively Enhanced Multiple Intelligence · CEMI
An individual with AI competes
with a team of ten
An SMB that uses AI intelligently competes with large corporations that use it badly.
It's happening in real time — and for an emerging market, it's the best possible news. But only if you use it to do more and better, not the same thing faster.
Five verbs
…people — don't replace the very ramps that make professional growth possible.
The best team isn't the one with the best players, but the one that makes its players better.
With tools, guidance, an environment that enables them, and the freedom to contribute. AI is that tool: it can concentrate or empower. The decision is not the machine's. It's ours.
Intelligence in the machines
vs. intelligence in our decisions
Change it, but change it well. The question is: will you seize it?
Scan for the event page · CEMI.ai · iBIZai.ai · SKaiLLS.ai
