Your Prompt
I want you to help me identify a system, workflow, recurring frustration, or industry inefficiency that I have largely accepted as “just the way it is.”
First, ask me 5 questions about:
- my industry
- recurring frustrations
- coordination problems
- bottlenecks
- wasted time
- unnecessary complexity
- information gaps
- tasks humans constantly work around manually
After enough context, identify 3 systems or inefficiencies that are likely being tolerated because of human cognitive, communication, or organizational limits rather than true necessity.
Present them to me as options to proceed. Then:
- analyze why those inefficiencies exist
- explain how previous technologies failed to solve them
- explain how AI or modern computing could realistically reduce them
- identify second-order consequences if those inefficiencies disappeared
- identify who benefits, who loses leverage, and what new risks emerge
Do not default to optimism or pessimism.
Treat technological progress as disruptive, uneven, and adaptive at the same time.
Use historical parallels where relevant.
Against The Grain
How much progress are we willing to tolerate before the unintended consequences outweigh the benefits?
AI is forcing that question again... especially around environmental cost.
And the concern is legitimate.
Modern AI infrastructure consumes an enormous amount of power. Data centers require electricity, cooling systems, water, land, and rare materials. Local grids are already straining in some regions. This isn’t imaginary, and pretending otherwise just makes the pro-AI crowd sound unserious.
But taking those concerns seriously does not require rejecting the technology itself.
A century ago, economists and scientists warned population growth would outpace food production. The “population bomb” became one of the great fears of the twentieth century: Millions starving because humanity simply couldn’t produce enough resources to sustain itself.
Some countries responded by trying to limit population growth by policy... and are still dealing with the demographic and economic consequences fifty years later.
Meanwhile, many of the worst global famine predictions never materialized. Not because the concern was irrational, but because humans adapted faster than expected.
Fertilizer technology improved.
Crop science improved.
Mechanization improved.
Global logistics improved.
Agricultural yields exploded.
And despite the predictive models of collapse, the human population still nearly tripled.
That pattern shows up constantly throughout history.
The printing press destabilized entire power structures. It contributed to revolutions, religious upheaval, propaganda, and social unrest on a massive scale.
But it also expanded literacy, access to information, and individual freedom. There was no putting that genie back in the bottle, no matter how many governments have since tried to ban the flow of information itself.
Industrialization polluted cities and rivers at horrifying levels before cleaner systems emerged. The transition was messy, expensive, and often harmful before it improved.
Cars transformed cities into polluted messes. We responded with emissions systems, cleaner fuels, catalytic converters, and eventually entirely new vehicle technologies.
The internet increased global energy usage at a massive scale... while simultaneously replacing physical infrastructure, reducing distribution costs, and democratizing access to information and opportunity.
Every meaningful technological leap initially looks destabilizing.
Because at first, we mostly see the disruption.
We don’t yet see the adaptation.
And I think that’s where the current AI conversation often collapses into two extremes.
One side assumes every new capability should accelerate as quickly as possible because innovation will solve everything eventually. Human cost becomes acceptable collateral damage in pursuit of progress.
The other treats the environmental cost and social disruption as evidence the technology itself should be rejected entirely. Human flourishing becomes secondary to preventing disruption at all costs.
But historically, humanity’s biggest breakthroughs rarely came from choosing one extreme or the other.
The danger is assuming human demand keeps accelerating... while human ingenuity suddenly freezes in place.
The argument assumes innovation stops immediately after creating the problem.
We do this professionally too. Entire generations convince themselves they can safely ignore each technological shift because “it's a good thing I'm retiring soon.”
Which often becomes true specifically because they stopped adapting… and also assumes that living another third of our lives in retirement shields us from the shift.
The solution to AI’s current risks is not less artificial intelligence.
It is more intelligence applied responsibly toward solving the constraints we currently fear most.
Better grid management.
Better material science.
Better infrastructure decisions.
Human civilization is full of inefficiencies we’ve simply learned to tolerate. AI will expose and reduce many of them faster than humans alone ever could.
Now... that doesn’t mean blind optimism is wise.
Technology transitions are uneven. Some communities absorb more cost than others. Efficiency gains will increase consumption, while also reducing its current effects. Concentrated power is a legitimate concern. And “someone will figure it out eventually” is not an actual strategy.
But rejecting the technology entirely because the first version is disruptive, resource-intensive, or uncomfortable feels historically familiar too.
Humanity rarely moves forward by trying to uninvent transformative tools.
Optimism isn’t believing the future will automatically get better.
It’s believing the future can become better because humans are capable of improving it.
That’s been true across agriculture, medicine, infrastructure, manufacturing, computing, and communication.
And I’m willing to bet human ingenuity doesn’t suddenly stop here.
The future doesn’t belong to the people accelerating AI without accountability.
But it also doesn’t belong to the people refusing to engage with it at all.
It belongs to the people willing to help shape where the intelligence gets applied next.
