Yesterday, I had the great privilege of speaking to several hundred Albertan Engineers at APEGA on Ethical AI: Analytical Governance in an Evolving Environment.
Over the last 18 months, I’ve delivered versions of this talk to several thousand people including engineers, lawyers, executives, developers, and public sector leaders. In almost every session or in a followup email somebody will voice the same underlying anxiety: “I’m worried about where this is all going, and I’m not sure anyone is actually steering the ship.”
That sense of helplessness is understandable, but it is mistaken.
If I can summarize the talk – the boundaries we draw right now, as individual users, corporate leaders, and working professionals, form the trellis along which these AI companies will build. Commercial AI development is an economic engine guided by demand. If we passively consume hyper-optimized tools without setting ethical parameters, we signal that efficiency is our only priority. But if we demand tools that preserve human agency, the market will adapt.
The responsibility to act on our personal values isn’t limited to the engineers writing code; it belongs to every informed professional making decisions about how technology is integrated into our workplaces and lives.
When I started my career fifteen years ago in analytics, people had a visceral reaction to being quantified. We used to draw a hard line: quantitative modeling belonged on the operational and financial side of the business.. reducing equipment collision rates, predicting supply chain bottlenecks, or optimizing logistics routes. But over the last decade, that line has quietly dissolved.
Today, we routinely apply those same quantitative, optimization-driven frameworks to internal human characteristics. We measure employee sentiment, track attention spans, automate interpersonal communication, and treat personal growth as a series of response variables to be maximized.
When we shift from optimizing physical assets to optimizing personhood, we risk treating our core humanity as a friction point to be eliminated. The real danger of modern AI isn’t just algorithmic bias; it’s the quiet surrender of authentic human struggle, empathy, and connection in exchange for superficial convenience.
As Anthropic co-founder Chris Olah noted during the release of the Vatican’s encyclical Magnifica Humanitas, tech developers urgently need “moral voices that incentives cannot bend.” Capitalist incentives drive labs to build what sells. They cannot govern themselves, nor should we expect them to.
If we view human care, mentorship, writing, and deep thinking as operational burdens to be offloaded to an agent, we trade away real fulfillment for administrative speed. Human limits, vulnerabilities, and complex emotions are not “system errors” waiting to be patched by a large language model. They are the exact places where trust, character, and genuine professional wisdom are forged.
Technical standards exist to evaluate safety and performance, but they cannot tell us what kind of future we ought to build. That requires a moral imagination anchored in personal convictions.
A big thank you to the entire APEGA team for organizing yesterday’s session, and to everyone who joined the conversation. The sheer size of the audience proves that professionals are ready to move beyond AI hype and engage with the hard, necessary questions.
Jeremy Adamson 