I’ve spent most of my life implementing systems at scale, watching engineering teams transform how they work (pre- and post-AI), and having thousands of conversations about what it all means. The sense of awe I feel with what is happening today, and what continues to happen strikes me the deepest.

We’re living through something rare. A technological shift that doesn’t change what we (everyone!) can do, but fundamentally alters the relationship between human capability, human effort and human output. While the discourse tends to oscillate between utopian hype and existential dread, I find myself stuck on something simpler and more profound: we’re about to get a lot of our time back.

The Productivity Paradox?

Here’s what I’m seeing in practice: Engineering teams aren’t only working longer hours with AI tools, but those hours are smarter. A senior engineer who used to spend three days debugging a complex system issue now spends three hours. Instead of only filling those recovered hours with more debugging, they’re spending time on architecture decisions that require human judgment.

This isn’t a productivity gain in the industrial sense of “more output per hour.” It’s something different. It’s the ability to operate at the right level of abstraction for your actual expertise. Junior engineers are writing production-quality code because AI handles the syntax and boilerplate. Senior engineers are focused on system design because AI handles the implementation details. Technical leaders are actually leading because they’re not buried in the weeds.

The awe comes from watching people rediscover what they’re actually good at. We’ve spent decades training brilliant minds to be really good at things that computers can now do better. The revelation isn’t that AI makes us more productive but that productivity might finally align with what we find meaningful.

Agents, Automation, and Actual Balance

I used to think work-life balance was about dividing hours. Now I’m starting to think it’s about dividing attention.

The promise of AI agents isn’t that they’ll do your job for you. It’s that they’ll handle the cognitive overhead that prevents you from doing your actual job. We have AI agents that monitor our system alerts, triages them, and only escalates what genuinely needs human attention. Another one prepares meeting summaries and action items. A third handles the first pass of code reviews.

They’re removing the friction between me and the work that matters and that distinction is what is awesome.

The exhaustion of knowledge work isn’t usually from the hard problems. It’s from the accumulation of small decisions, context switches, and administrative overhead. You spend eight hours “working” but only a few hours on anything that required your unique capabilities. The other six were just… “overhead.”

AI doesn’t eliminate work. It eliminates that friction and when you remove friction you experience work differently. You’re present for it in a way that wasn’t possible when your attention was fragmented across a dozen different types of tasks.

Enjoying this? I write about AI implementation and engineering leadership every week.

Individuals Operating at Scale

This might be the part that fills me with the most awe.

We are watching individual contributors operate with a leverage that used to require entire teams. A single designer can now iterate on dozens of variations, run user testing, and implement changes in the time it used to take to schedule a meeting about running user testing. A single engineer can build, test, deploy, and monitor systems that would have required a whole squad of contributors.

This doesn’t make squads obsolete. It makes them better.

When individuals have more leverage, collaboration becomes about combining different types of expertise rather than subdividing labor. You’re not working together because you need to distribute the workload but you are working together because you need different perspectives, different judgment, different ways of thinking.

Let the AI handle most of the implementation, and let the humans handle the strategy.

Individual scale means you can afford to be thoughtful. You can afford to iterate. You can afford to ask “is this the right thing to build?” instead of just “can we build this quickly enough?”

Reclaim the Human Parts

Humans are actually pretty bad at most of what we spend our time on.

We’re bad at repetitive tasks. We’re bad at processing large amounts of data. We’re bad at maintaining consistency over long periods. We’re bad at remembering everything. We’re bad at context switching without losing information.

But, humans are good at: judgment, creativity, empathy, strategic thinking, understanding nuance, making connections between disparate ideas, telling stories, teaching, mentoring and most of all leading!

AI is giving us the opportunity to spend more time on the things humans are actually good at

Are we starting to see engineers who actually have time to mentor? Leaders who can be present with their teams instead of constantly context-switching between meetings? Teams that can afford to have thoughtful discussions about architecture instead of just shipping whatever gets the feature out fastest?

Are we enabling people are learning instruments again? Taking on creative projects? Spending time with their kids without their brain half-occupied with work problems? Reading books that have nothing to do with their career?

The reclaimed time isn’t just about efficiency it is about agency. The ability to choose what deserves your attention because you’re not drowning in things that demand it.

Compounding Cognitive Surplus

Your team has time to actually think through problems, so you make better decisions. Better decisions mean less rework & less rework means more capacity. More capacity means you can take on more ambitious problems, more ambitious problems mean more learning, and more learning means better judgment.

We are watching teams that used to struggle with operational excellence because they were too busy firefighting now build genuinely robust systems because they have time to do it right. We are watching engineers who used to barely have time to write code now writing documentation, building tools, and improving processes. We are watching leaders who used to manage by spreadsheet now actually leading.

Awe is in what communities can do when everyone in them has more capacity to think.

Risks Are Real, So Are Opportunities

I’m not naive about the challenges. The risk of AI replacing jobs is absolutely real. The risk of AI amplifying inequality is absolutely real. The risk of AI systems making consequential decisions without adequate oversight is absolutely real. The risk of us becoming dependent on systems we don’t fully understand is absolutely real.

I’m also watching something else happen, I’m watching people who were stuck in unfulfilling work discover what they’re capable of when they’re not ground down by cognitive overhead. I’m watching companies in traditional industries that could never afford world-class engineering talent now able to compete because AI democratizes expertise. I’m watching individuals build things that used to require venture capital and teams.

The optimism isn’t about ignoring risks. It’s about recognizing that the same technology that poses those risks also gives us unprecedented tools to address them. We have more capacity to solve problems because we’re spending less capacity on busywork. We have more time to think strategically because we’re spending less time on tactical execution. We have more energy for creativity because we’re spending less energy on repetition.

Filling the Time?

What do we do with all the reclaimed time? That is the real opportunity, and it is not just productivity or efficiency but the chance to redefine what we optimize for.

Maybe we optimize for deeper expertise instead of broader coverage. Maybe we optimize for thoughtful solutions instead of fast ones. Maybe we optimize for sustainability instead of growth at all costs. Maybe we optimize for meaning instead of output.

Nobody knows what this looks like, and we’re figuring it out in real time. But I know what I’m seeing: people who are more engaged, more creative, more present. People who are excited about what they’re building instead of exhausted by the building of it.

That’s the awesome part of this: not in the technology itself, but in what it enables us to become.

The Work Ahead

We’re not at the end of this transformation, and the AI systems we have today will look quaint in five years. The work patterns we’re developing today will evolve. The challenges we’re just starting to grapple with will become much more complex in ways, and simpler in others.

But for the first time in years, I’m genuinely optimistic that we might be heading toward a world where work feels more like what it should be: a way to apply our unique human capabilities to problems we find meaningful, rather than a way to occupy our time until we’re too tired to do anything else.

The awe is in rediscovering the human intelligence that we’ve been too busy to fully deploy not in artificial intelligence itself.

What might we be in a future where technology serves to make us more human rather than less? Even with all the risks, even with all the uncertainty, even with all the challenges ahead there is a reason to feel awe.

I’m not ready to do that yet by giving up on shaping the future I want to see.

Read more:

About the Author:
Duncan Grazier is a builder and the founder of Grazier Ventures, a holding company for ideas built with AI. Its first venture is Aqen.ai — ‘you bring the idea, Aqen brings everything else.’ He also runs TideHelm, the advisory practice founders, investors, and CTOs call before the hardest decisions. Previously Chief AI Officer at BuildOps ($1B unicorn), engineering leader at Weedmaps (through IPO) and ShopKeep (acquired by Lightspeed for $550M). Based in San Clemente, CA. Writes weekly at itsmeduncan.com.