Why AI Has Given Fergus Dyer-Smith His Hands Back

Fergus Dyer-Smith became CEO of M3 Labs earlier this year, where he helps clients deploy and solve real problems by creating AI products that’s enterprises actually use. A serial entrepreneur, Fergus founded Wooshii in 2009, which was a pioneer in developing tech-enabled video services, offering end-to-end video production and performance management solutions. The business was acquired by MSQ, M3 Labs’ parent company, in July 2025.

Fergus is a regular keynote speaker on AI-driven marketing innovation and real-world application. He’s the presenter of AIM– a weekly LinkedIn series that dives into the AI shaking up marketing, and has been a Board Advisor to AI and Tech businesses such as PinPoint Media, Envoke-Demos and Flexy App.

LBB> What’s the most impactful way that AI is helping you in your current role?

Fergus> Honestly, it’s given me my hands back. I haven’t coded properly in years, but with large language models I can sketch product ideas, write pseudo-code, and even prototype features without needing to pull an engineer off something else. It’s like being back in the engine room again - except this time, I can move from idea to instrumented prototype in a day. The impact isn’t just speed. It’s precision. Fewer meetings, less translation loss. I can think like a builder again.

LBB> Beyond efficiency, how is AI improving quality?

Fergus> AI has raised the floor, not just sped up the process. Everything starts at a higher baseline. The first draft, the first mock, the first dataset, it’s all already decent. That means we spend our time improving, not fixing.

LBB> What are the biggest challenges in collaborating with AI as a creative professional, and how have you overcome them?

Fergus> Hallucinations are still around, but I treat them as an engineering issue, not an existential one. You put the right checks in place, trusted data sources, grounding functions, testing loops, and the problem mostly disappears. The tougher bit is style drift. Models love to flatten nuance. We manage it by embedding brand DNA into the system: tone markers, rhythm notes, do-not-say lists. It keeps the voice consistent. The biggest trap, though, is false confidence. AI makes everything look neat, even when it’s wrong. So we’ve built a discipline around separating 'pretty' from 'proven.'

LBB> How do you balance AI with your own creative instincts?

Fergus> I use AI for width, not depth. It’s the fastest brainstorming partner I’ve ever had, brilliant in a range of producing. I’ll often ask it to attack an idea, then defend it, just to test my conviction. But that’s all it is, sparring. The final judgment, the taste call, that’s mine.

LBB> How do you make sure the work keeps a human touch?

Fergus> For me, authenticity isn’t something AI can give you. It sits in the decision-making layer, the human one. We use AI to generate, but the selection, the sense of what feels right, that’s where people come in. Taste, intuition, context, consequence. Those are still human superpowers.

LBB> Any big misconceptions you see around AI in the industry?

Fergus> Plenty. People still judge AI by its weakest outputs, the weird hands or the clunky copy, and miss the trajectory. That stuff will be fixed. The real limitation isn’t model quality, it’s workflow design. The other myth is that there’ll be one model to rule them all. In reality, the magic happens in how multiple narrow systems work together.

LBB> What ethical questions does this raise for you?

Fergus> Copyright doesn’t keep me up at night. We’ve already built systems that can handle that. Bias does. So does harm. That’s where the real responsibility lies, testing for how outputs behave in sensitive contexts, making sure our tools don’t quietly reinforce nonsense. We bake those checks into the process now.

LBB> How have attitudes toward AI changed recently?

Fergus> Six months ago it was all demos and dazzle. Now it’s procurement reviews and ROI spreadsheets. Creatives are past the panic stage; they’ve realised AI can cut the grunt work without killing their voice. Finance teams, meanwhile, have gone from FOMO to “show me the metric.”

LBB> Do you feel positive or worried about creativity’s future under AI?

Fergus> Mostly positive. The danger isn’t the tech, it’s leaders trying to bolt it onto old processes. The companies doing well are rebuilding their workflows from the ground up, treating AI like electricity in a factory. It changes the layout, the tools, even the job titles. When you get that right, creativity expands.

LBB> Do you think AI will enable entirely new forms of art or media?

Fergus> Absolutely. We’re already seeing it. Personalised video that adapts in real time to your behaviour. Creative that rewrites itself across platforms without losing brand voice. Generative brand worlds where audiences co-create, and the system decides what lives or dies. That’s not automation; that’s evolution.

LBB> And looking ahead, what impact will AI have on your role and your teams?

Fergus> Product work is shifting from shipping features to tuning systems. You need model evals, data contracts, and continuous loops that learn. Creative work becomes more about direction and curation, senior talent sets the taste and constraints, junior talent orchestrates the agents. Even pricing models change: less time and materials, more measured value. And hiring changes too. You want people who speak both data and narrative.