Just Because We Can Doesn’t Mean We Should
AI, judgment, and the cost of outsourcing ourselves
I was standing on a stage recently, in front of a few hundred people. Someone asked whether the pace of AI innovation is too fast and whether regulation is necessary to protect people. I said yes to both, without hesitation, and not because it sounds responsible. Because it is observably true.
The speed is real. Anyone paying attention feels upheaval in their operating system, with the near-constant cycle of weekly model releases. Entire AI workflows and strategies are rewritten before they get a chance to stabilize. It is not out of line to say this is overwhelming.
What is less often said, especially on stages, is that speed itself is a choice. Not in the sense that we can halt research or rewind the internet, but in the sense that adoption and delegation are still human decisions. We keep talking about the pace as if it were weather. It is not. It is infrastructure and leadership posture.
When I said on that stage that AI is not replacing humans anytime soon, I could feel the room shift. People are tired of pretending they believe something their lived experience contradicts. Despite the demos and the headlines, we are nowhere near a world where everyone casually buys an at-home humanoid robot to do their laundry or raise their kids. If you have watched the recent robotics showcases closely, not through a marketing reel but through an operator’s eye, the gap between spectacle and reliability is still wide. Useful, yes. Transformational in narrow contexts, also yes. Complete and total human obsolescence, no.
The more interesting question is not whether machines will replace us, but why we are so eager to frame the future that way.
Speed Is Not the Same as Direction
The regulation question often gets flattened into ideology, but founders experience it differently. Regulation is not about freezing progress. It is about buying time for systems, norms, and identities to adjust. Every major technological shift that actually stuck was accompanied by constraint, sometimes formal, sometimes cultural. Railroads, aviation, pharmaceuticals, and finance are some examples. History shows it’s not because regulators were clairvoyant, but because unbounded speed breaks trust faster than it builds value.
AI is no different. The difference is that this technology reaches into cognition, not just labor. That makes the adjustment psychological before it is economic.
We talk about pace as if it is an external force bearing down on us, when in practice it shows up as a thousand small leadership decisions.
Do we replace this role or augment it?
Do we automate this judgment or keep it human?
Do we ship this feature because we can, or because it aligns with what we want this product to be in three years?
Those are not technical questions. They are authorship questions, and founders should take time thinking through them.
There is a philosophical layer here that founders tend to wave away because it feels indulgent. It is not. If you do not define what remains human, the market will define it for you, and it will do so based on efficiency, not meaning.
We do not actually have to accept the pace wholesale. We can choose where to slow down, where to hybridize, and where to draw lines. The problem is that doing so requires clarity about what we value, and many leaders have outsourced that clarity to roadmaps and vendors.
Speed without authorship turns leaders into custodians of someone else’s future.
Creativity, Instinct, and the Things Models Do Not Carry
I had a conversation recently with a founder who runs a marketing agency. He was calm in a way that stood out. Not performatively calm, but unbothered. He told me he is not worried about AI replacing him or his team. Not because he underestimates the AI tools he uses; he uses them, but because he understands his own leverage.
What he described had nothing to do with the latest AI trends. He talked about instinct. About curated taste that comes from a decade of watching campaigns fail quietly before one lands. About the bodily sensation of knowing when something is off before you can articulate why. About sitting with a creative idea long enough to feel whether it carries emotional truth or just novelty.
AI models can mimic aesthetics. They can recombine patterns at scale. They can even approximate emotional resonance if you define it narrowly enough. What they do not have is experience in the human sense. They do not have memory shaped by consequence. They do not feel grief when something fails, or pride when something lands against odds. Those sensations are how judgment is trained.
Human instinct is a compression algorithm trained on lived experiences and stakes. When founders talk about replacing that with AI, what they usually mean is replacing the surface output, not the underlying learning loop. The output might look similar. The loop is not.
We can build machines that simulate creativity. We cannot give them a childhood, a body, or a social cost to being wrong. Not anytime soon. That matters more than most roadmaps admit.
The real scarcity is not intelligence; it is judgment with accountability.
Hybrid Futures and the Question We Keep Dodging
Where this becomes uncomfortable is identity. For a long time, especially in founder culture, being human has been synonymous with being productive. Work became proxy for worth. Long hours praised as an earned rite of passage. Output as proof of existence and survivorship. AI disrupts that bargain not by taking all the work, but by hollowing out the parts we used to rely on to feel necessary.
If machines handle the repetitive, the mundane, and even some of the analytical heavy lifting, what is left? Do we become more human, or do we slowly outsource the very thinking that made us feel distinct? I ask this myself all the time.
This is the conversation we should be having now. What does a hybrid AI world actually look like when designed intentionally? Which decisions must remain human even if they are slower? Which tasks should be automated precisely because they free cognitive and emotional capacity? Where do we draw boundaries and not because we cannot automate something, but because we should not?
There is also an unspoken anxiety underneath all of this. If work recedes as the primary container for identity, many people do not know what replaces it. What does it truly mean to be human? Biology alone does not feel sufficient and experience without articulation feels thin. We have not built new narratives of value yet, and that absence creates a vacuum that technology will happily fill.
Founders feel this acutely, even if they rarely say it out loud. When you have built your sense of self around building, what happens when building gets easier, faster, and less tied to personal sacrifice? Does that make you less relevant, or does it demand a deeper definition of relevance?
The danger is not that machines become more human, but that humans become less deliberate.
None of this resolves cleanly, and that is the point.
The future will not ask our permission to arrive, but it will absolutely reflect the decisions we make while pretending we have none. If you are a founder, the quiet question underneath all the AI noise is not whether you can keep up. It is whether you are still choosing. Are you?
What parts of your work are you unwilling to delegate, even if delegation is cheaper? Where are you letting speed substitute for thought? And if your identity has been welded to productivity for years, what are you becoming as that identity loosens?
That answer is not something an AI model will give you. It is something you decide, slowly enough to feel it.
Thinking aloud,
Yen
P.S. Check out the marketing agency founder, TJ Kolesnik; he’s brilliant at what he does.