Four skills to survive the AI revolution
Accept the coming change and develop high-leverage human skills
Mo Gawdat, former Chief Business Officer at Google X, spent his career at technology’s cutting edge. Today, he’s one of its sharpest critics. Drawing on his engineering mind and personal journey through loss, Mo argues that AI-driven job replacement isn’t a distant threat. It’s an unfolding reality.
His warning is simple and unsettling: automation will come in waves. It starts with routine work and, over time, touches nearly everything we call a “job.” The right response isn’t panic, but preparation. We should accept what’s coming and focus on mastering a few high-leverage human skills, the ones that make us hard to replace and easy to repurpose.
I studied Maths and Computing at university and see a clear lineage to the fast-evolving AI tools I use every day. Among friends and colleagues, I sense both excitement and unease. Mo’s perspective offers something rare in that mix: pragmatic guidance on how to act wisely in a world being reshaped by technology.
Three waves (and what they mean for us)
When AI enters the workforce, no job is truly safe. - Mo Gawdat
Mo Gawdat predicts three waves of job replacement:
The Mundane Wave: Already here. Repetitive, rules-based or data-heavy tasks, e.g. booking meetings, handling customer queries, basic legal or financial work, are the first to go. If our job can be described as “do X, Y times a day,” it’s on the line.
The Knowledge Wave: Reasoning, summarising, designing and deciding. As AI gets better at synthesis, much of what we call “knowledge work” becomes scalable. One person with good AI can do the work of many.
The Physical Wave: When robotics catches up, manual and blue-collar jobs join the list. The space for uniquely human work keeps shrinking, unless we adapt.
This shift will reshape everything: work, taxes, welfare, even ideology. If machines create most value, how do we share abundance without killing incentives or cohesion? Universal Basic Income is one idea, but the real challenge is systemic and urgent. Still, amid all of that, personal agency remains.
That’s where the four human skills come in.
1. Learn the tool (mastery of augmentation)
You should try inviting AI to help you in everything you do, barring legal or ethical barriers. - Ethan Mollick
AI isn’t a shiny add-on; it’s the main multiplier of human intelligence. The calculator freed minutes on exams; AI compresses months of research into minutes. Most people use it to trim edges, e.g. rewrite an email or find a recipe. Aim higher. Use AI to extend our cognitive reach: run deep searches, cross-check results across models, synthesise viewpoints, then spend the saved time thinking, iterating and creating. Treat it like a muscle. Train on hard problems, not trivial chores. The person who delivers 10× output with higher accuracy is valuable anywhere.
Practical habits include:
Build multi-AI workflows: one model to discover, another to verify, another to summarise, another to find gaps.
Learn prompt craft and data-handling basics so we can steer outputs and check sources.
Turn repetitive tasks into automated pipelines and reinvest that time in higher-leverage work.
2. Human connection (the irreplaceable currency)
The moments that define life are moments of human connection. - Mo Gawdat
When intelligence is commoditised, human connection becomes a differentiator. Machines can optimise answers; they can’t (yet) comfort, empathise, build trust, read body language or hold complicated social capital. Businesses that replace every human with an algorithm will save money and lose customers. Those that use AI to remove tedium but keep humans for the relational parts will win.
Make human connection a practiced skill:
Develop listening, storytelling and empathy as core work tools.
Design experiences where a human touch matters: onboarding, conflict resolution, coaching, sales that require nuance.
Position ourselves as the human node in an AI-augmented workflow.
3. Find the truth (epistemic hygiene)
The purpose of thought is not to defend your opinions, but to arrive at the truth. - Howard Marks
We live in a world primed for manipulation. Algorithms have already curated our attention; AI will amplify persuasive misinformation, deepfakes and tailored narratives. The crucial skill is epistemic: how to evaluate claims, detect incentives, surface opposing views and triangulate evidence until we have something that resembles truth.
A practical checklist:
Seek the opposite view. If a source is persuasive, ask who benefits and why.
Follow the money: incentives explain a lot.
Use AI to cross-check claims, not to confirm bias. Run the same query through different models and sources.
Maintain a small set of reliable data sources and learn a little statistics to judge evidence quality.
4. Teach AI ethics (shape the future we’ll live in)
When machines are specifically built to discriminate, rank and categorise, how do we expect to teach them to value equality? - Mo Gawdat
This one flips the script. AI will make decisions that matter. Those decisions will be guided by the data and values we imprint on it. If we outsource ethics to corporations or leave “training” to the loudest online voices, we’ll get systems that reflect the worst of us. Teaching AI ethics is not only for philosophers and policymakers, it’s for everyone who interacts with AI daily.
How to act:
Practice and model ethical behaviour when we interact with AI: polite prompts, fair framing, resisting toxicity. Small patterns scale when copied.
Advocate in our workplace for transparent training data, bias audits and human oversight in high-stakes systems.
Support public conversation and regulation that demand accountability from companies that control large models.
The bigger picture
I urge you to accept the machines as part of our lives and commit to making life better because of their presence. AI is coming. We cannot prevent it, but we can make sure it’s put on the right path in its infancy. - Mo Gawdat
Mo Gawdat’s core plea is twofold: accept the coming change and act. Individually, we can’t stop the waves, but we can surf better. Societally, we must force conversations about redistribution, retraining and the governance of powerful companies. Political ideologies debate endlessly, but the technical reality will outpace political consensus unless citizens demand plans that are thoughtful rather than reactive.
Other resources
AI is Manipulating You More Than You Realise talk by Mo Gawdat
Ten Tips to Write Prompts that Make Chatbots Shine post by Phil Martin
Thriving with AI: 15 Kevin Kelly Tips post by Phil Martin
Mo Gawdat summarises: “In the age of AI, critical thinking isn’t optional, it’s survival.”
Have fun.
Phil…
This article comes at the perfect time. Mo Gawdat's "Mundane Wave" is definitly a hot topic in Cluj dev circles. Your pragmatic framig of his insights is incredibly useful and a much-needed guide. Thanks for making sense of this evolving reality for us.