A century ago, America entered the Roaring Twenties convinced that the future had arrived ahead of schedule. Automobiles, radio, telephones, electrification, mass production, and modern finance have revolutionized everyday life. Productivity has expanded. Markets skyrocketed and credit became a form of currency.
Then the bill came due.
But the lessons of the Roaring Twenties didn’t begin in 1929. It began much earlier with President Theodore Roosevelt’s conservation fight. Roosevelt was not anti-growth. He believed in ambition, business and national development. But he also understood that prosperity could not depend on institutions consuming forests, water, wildlife, and public lands faster than they could protect them.
His answer was to never stop progressing. It was to control it. During his presidency, Roosevelt helped protect approximately 230 million acres of public land, founded the U.S. Forest Service, and promoted a conservation ethic built on sustainable ideas. Resources are not truly ours if we consume them in ways that leave no possibility for those who come after us.
This management challenge faced opposition. Timber and mining interests, as well as some Western critics of federal land policy and some members of Congress, considered conservation an overreach. This pattern feels familiar today. Climate action is often seen as a constraint rather than a resilience. Long-term risks collide with short-term incentives. Control is attacked as hindrance, while exploitation is defended as freedom.
Roosevelt rejected that false choice. Conservation was not the enemy of prosperity. That was the condition for continued prosperity.
For the 1929 crash, the second half of the warning was added. Innovation was real, but so too was excessive speculation, cheap money, weak guardrails, and uneven prosperity. A history of the Federal Reserve’s crash notes that optimism about new technology overlapped with mutual funds, brokerages and margin accounts that allowed investors to buy stocks with borrowed money. The progress was real. The foundation was weak.
AI will give the 2020s its own roar
AI can personalize learning, accelerate scientific discovery, improve decision-making, optimize operations, and expand access to historically marginalized communities. But promise is not responsiveness, and scale is not sustainability. Across industries, AI is advancing faster than many institutions can absorb. The team is experimenting. Investors are speculating. Vendors are doing marketing. Workers are adapting. Regulators are also catching up.
AI is now a conservation issue As much as a technical question. Resources at stake include energy, water, carbon emissions, data, employee trust, human judgment, institutional legitimacy, and public trust. The International Energy Agency predicts that global data center power consumption could more than double by 2030, reaching approximately 945 terawatt-hours in the base case.
That doesn’t mean organizations should avoid AI. This means that the resource demands of AI need to be managed as seriously as strategy, risk, finance, and reputation.
Sustainable AI cannot be reduced to efficient chips or purchasing renewable electricity, but they are important. This requires environmental, human and institutional sustainability.
Environmental sustainability asks whether AI use cases consider energy, water, carbon, hardware, grid pressure, e-waste, and community impact. Human sustainability asks whether AI enhances workers’ capabilities, dignity, agency, competitiveness, and remedies. Institutional sustainability is about whether organizations know where AI is being used. This includes who owns it, what data is used, what risks it poses, how it is monitored, and who is responsible if something goes wrong.
From my perspective as a technology executive and researcher in AI governance and digital equity, the biggest risk is not that AI becomes too powerful. The big risk is that institutions become too passive. In reality, too many organizations treat AI as a competition to implement tools. Operating model transformation.
In my doctoral research, I found that people don’t judge AI governance solely by policy language. They determine whether governance is practiced. Are the boundaries clear? Can people question the output? Is human judgment preserved? Is there a way to fix the harmful results? Does anyone have an answer?
That’s why human participants can’t be the checkbox for final approval. Human intelligence must shape AI throughout the lifecycle of problem definition, data selection, model design, procurement, deployment, monitoring, escalation, exception handling, training, and resources. The most serious failures often begin long before the final decision appears on the screen.
What’s next?
Leaders should start with these five moves.
-
Moving from AI experimentation to AI operational discipline.
-
Treat computing, carbon, water, and hardware as governance issues.
-
Manage your data as an enterprise risk.
-
Makes humans responsible throughout the life cycle.
-
We need transparency, resources and reinvestment in employee capabilities.
The last Roaring Twenties taught us that progress can be both real and reckless. President Roosevelt’s conservation legacy adds an equally important lesson. That means leadership is measured not just by what you build, but by what you protect.
A century from now, people may look back on the 2020s as another tumultuous decade of technological change. The question is whether they will understand that we have learned the lessons of 1929 and the lessons of conservation that preceded it.
