Atlantic Business Magazine (ABM) invited me onto Episode 5 of For The Record, its video podcast, to answer the hardest questions I get about AI. Will it end humanity? Will it make work better? Will it end work altogether? Dawn Chafe, co-owner of ABM, asked me all of them in 42 minutes.
WATCH THE FULL EPISODE
For The Record, Episode 5: “Ready or not, AI is here” · Atlantic Business Magazine · about 42 minutes, free
What I’d want every business owner to take from the conversation:
1. AI carries real risk, and good leaders using it are the counterbalance.
2. You don’t need technical skills to use AI.
3. A grade six question gets a grade six answer.
4. Start with the value your business creates, then the workflow.
5. Work is moving from doing to conducting.
6. AI can make us mentally lazy unless we build in the struggle.
7. Change sticks when you make it with people, not to them.
1. Is AI a threat to humanity?
Probably yes. There is a real probability of harm, and right now there’s no regulated counterbalance keeping pace with the risk.
I watch three red lines:
• Recursive self-improvement: AI improving itself.
• Its own aims: AI deciding to chase goals of its own.
• Its own assets: AI acquiring resources for itself.
We haven’t seen an agent go off and pursue its own goals. There are signs we’re close to the first line, though, and once that’s crossed, I don’t know how you stop the other two.
So who balances it? As I said on the show: “The only counterbalance we have is good people, good leaders who understand AI.” Leaders who use it for what it can do understand its power, and they’re the ones who can push back.
Keep the fear in proportion too.
Humans overindex on fear and underindex on potential. We drive past each other with inches to spare, and we’d never approve that today. We’d also have given up everything the car made possible. For every doomer story about AI, there are thousands of good ones.
2. Do you need technical skills to use AI?
No. When Atlantic Business asked me for the biggest myth business owners in Atlantic Canada believe about AI, this was my answer.
Most owners think AI is IT. “As long as you stare at it as if it’s IT, you’re never going to get what you need out of it.”
A large language model is designed to work with natural language. You use it by asking questions in plain words.
The best proof I have is a cladding company we worked with. The owner came to an AI session two months before he bought his first laptop. His biggest problem was estimating, so I asked him to talk me through how he does an estimate. That gave us the workflow. Then we asked how AI could make it better.
Four months later he wrote to me. He was saving 600 hours a year on estimates. The estimates were better, and his conversion rate was up 17%. He spent the time he got back writing more estimates. In the last 18 months, he’s doubled the company twice.
“All you have to have is the ability to ask a good question. And if you started a business, you have that ability.”
3. How do you get good answers from AI?
Ask better questions.
The quality of what you put in decides the quality of what you get out. “You ask it a grade six question, you get a grade six answer. You ask it a PhD question, you get a PhD answer.”
Three habits help:
• Frame it. Give it a role, the input, the context and the constraints, then turn it into a conversation. “You are April Dunford, a world-renowned expert in marketing” gets you a very different answer from a bare question.
• Use it to challenge your thinking. “If you’re using AI to validate your feelings, your emotions or your thoughts, you’re abusing AI. If you’re using it to challenge your thinking so that you can be less wrong faster, you’re using AI.” Push back on it a few times and it tends to come round to your side. One senior leader told me after a session that after two or three back-and-forths, it was producing her own argument.
• Know where your expertise ends. “If you do not have domain expertise, you really can’t sniff a bad answer.” AI will give you an exceptionally confident answer either way. A second opinion on a legal document doesn’t make you a lawyer.
4. Where should a business start with AI?
Start with the fundamentals, which haven’t changed. What problem do you solve, for whom, and what is solving it worth?
Then look at your business as a map of workflows.
Marketing creates awareness. Sales converts it. Delivery keeps the promise. Scheduling lets you do enough of it each week to earn the margin you need. The right people, tools and visibility hold it together.
Ask of each workflow: does this still have value? Some won’t. When we went from the cart to the car, buggy whips lost their market.
Then find the one constraint in the workflow that matters most, and point AI at it. That’s how an AI roadmap starts: the biggest headache first, then the next one.
The last three industrial revolutions needed capital. This one needs clarity: on where you’re taking the business, how your workflows are changing and how the skills your people need are changing.
5. What happens to jobs?
They change shape. “We’re going to go from doing the work to conducting the work.” While I was recording the episode, I had two agents preparing work for me.
The World Economic Forum’s Future of Jobs Report 2025 expects 92 million jobs to be displaced and 170 million to be created by 2030. The 170 million won’t look like the 92 million.
What we look for in people now:
• Judgment: critical thinking, knowledge and experience.
• Curiosity: you can’t be curious and furious at the same time, and most people meet AI from a defensive posture.
• Systems thinking: can they see how the work comes together and orchestrate value? Your most experienced people are an asset in this. They know what good looks like, so they can spot when AI gets it wrong and correct course. They won’t pick AI up on their own. Walk them through it.
6. Will AI make us mentally lazy?
It can. “We’re unleashing a convenience machine.” Your brain is the most tiring muscle you have, so given a way out, most people take it.
I build in the struggle on purpose. I still write and read long form. I take a day every two weeks without AI.
Businesses need the same habit. Your team will spend more time editing and verifying work than doing it, so check regularly that you still know how to build the bridge. Leaders didn’t ask for this job, but they now have to build a cognitive gym for their people.
7. How do you make the change stick?
Three things tell me whether an organization will integrate AI:
1. Is it CEO-led? “This is not something you can delegate or abdicate. You can’t just ask your head of IT to do it.”
2. Does it re-engineer or reimagine the workflow? Both matter.
3. Is it done with people or to people?
The third one decides whether it lasts. “People don’t get exhausted by change if they’re involved in doing it.” If your team chooses what they’ll do on the other side, they’ll probably keep doing it.
The short version
• AI’s risks are real, and leaders who understand it are the counterbalance. • You can start today with plain-language questions about your own work. • Pick the one workflow that drains the most time, and change it with the people who run it.
WATCH THE FULL EPISODE
For The Record, Episode 5: “Ready or not, AI is here” · Atlantic Business Magazine · about 42 minutes, free
Frequently asked questions:
Where can I watch For The Record Episode 5?
It’s free on YouTube at youtube.com/watch?v=ieI_WeVxKZQ. Atlantic Business lists every episode at atlanticbusinessmagazine.ca/fortherecord. The episode runs about 42 minutes.
What is For The Record?
For The Record is Atlantic Business Magazine’s video podcast. It goes behind the headlines of the biggest stories in the magazine’s print edition. Episode 5, “Ready or not, AI is here,” is about artificial intelligence.
Do I need technical skills to use AI in my business?
No. Large language models work in natural language. If you can describe how a job gets done and ask good questions about it, you can start.
What is an AI roadmap?
A plan for putting AI to work one workflow at a time. Start with the workflow that drains the most time and attention, unpack it, integrate AI into it, then repeat with the next one.
Who is Matt Symes?
Matt Symes is the University of New Brunswick’s AI Entrepreneur in Residence and an Atlantic Business Magazine Top 50 CEO five years running. He served as the AI advisor on the Atlantic Economic Panel’s expert advisory group. He runs a portfolio of nine companies and tests AI changes there first. He works with business owners through Levership and with leadership teams at larger organizations through Symplicity Designs.
Thank you to Dawn Chafe and the Atlantic Business team for having me. We have more in the works together, and I’ll share it as soon as I can.


