The Craig Bushon Show — “The Truth Is Not Hate Speech”
By Craig Bushon Show Media Team
We are warned about artificial intelligence constantly, and the warnings have settled into a familiar catalog. AI could eliminate millions of jobs, become a cyberweapon, enable a biological attack, power a surveillance apparatus, or eventually outthink the people who built it. Those warnings deserve serious attention. But they all share an assumption worth examining, which is that the danger lies in what AI might do to a functioning economic system. What if the more revealing question is what a struggling economic system needs AI to do for it?
Speaking at the U.S.-Saudi Investment Forum in Washington on November 19, 2025, Elon Musk offered a forecast that most listeners filed under science fiction. Within roughly ten to twenty years, he predicted, advances in artificial intelligence and humanoid robotics would make human labor unnecessary and money largely beside the point. “My prediction is that work will be optional,” he said, comparing a future job to growing vegetables in the backyard when the store sells them cheaper. Set aside whether the timeline is credible. The prediction itself is an admission, made by one of the largest beneficiaries of the current system, that the economic arrangement we live under today may not survive in its present form. And it arrives at a moment when the arithmetic of public finance has stopped working in every major economy on earth.
The Numbers No Longer Add Up
On August 18, 2026, total United States public debt crossed forty trillion dollars for the first time in the history of the republic, reaching $40.05 trillion in the Treasury’s Debt to the Penny dataset. A decade ago the figure stood at roughly $19.4 trillion, which means the obligation has doubled in ten years. The Congressional Budget Office projects a federal deficit of $1.9 trillion in fiscal 2026, growing to $3.1 trillion by 2036, with federal debt held by the public climbing from 101 percent of GDP to 120 percent over the same decade, surpassing the previous record of 106 percent set immediately after the Second World War. The most consequential line item is the one almost nobody campaigns on. Net interest costs are projected to double from roughly $1.0 trillion in 2026 to $2.1 trillion in 2036, making debt service the fastest-growing category of federal spending, outpacing both Medicare and Social Security.
This is not an American peculiarity. The International Monetary Fund’s April 2026 Fiscal Monitor reports that global public debt rose to just under 94 percent of worldwide economic output in 2025 and is on course to reach 100 percent by 2029, arriving a full year earlier than the Fund projected only twelve months before. Interest spending across the global economy has climbed from about 2 percent of GDP to nearly 3 percent in the span of four years, and the IMF describes the global fiscal buffer as having effectively vanished. Governments face simultaneous and growing demands for healthcare, pensions, defense, energy, and infrastructure at precisely the moment when the cost of carrying their existing obligations is consuming an ever larger share of what they collect.
Most of this debt will never be retired the way a family retires a mortgage. Governments refinance it, issue new debt against it, raise taxes, allow inflation to erode its real value, or count on future growth to shrink the burden relative to the size of the economy. The difficulty is that debt compounds, and compounding works against a borrower with the same relentless arithmetic it works for an investor. When a government must borrow to cover interest on money it already borrowed, the obligation grows on its own momentum. Eventually growth must accelerate, taxes must rise, spending must fall, the currency must lose purchasing power, or some combination of all four must occur. Political leaders across both parties have demonstrated almost no willingness to cut spending at the necessary scale, and citizens already stretched by housing, healthcare, insurance, transportation, and grocery costs represent a tax base with real limits. That leaves one escape route that requires no painful vote and no angry town hall meeting, which is to produce dramatically more with dramatically less.
That is where artificial intelligence enters the picture.
The Bet Is Already on the Books
Here is the part of the story that has received almost no attention outside of specialist circles, and it is the reason this piece exists. The proposition that AI productivity might rescue public finance is no longer a thesis to be argued. It has already been written into the official budget arithmetic of the United States government.
In its February 2026 Budget and Economic Outlook, the Congressional Budget Office incorporated a positive effect from generative artificial intelligence directly into its projections of economic growth, estimating an increase in productivity growth of roughly ten basis points per year on average, which raises the level of output in the nonfarm business sector by 1 percent by 2036. Read that again with the deficit numbers in mind. The nonpartisan scorekeeper that Congress relies on to price every bill has assumed a measurable AI dividend as part of the baseline against which all future fiscal decisions will be judged. If that dividend fails to materialize, the projections above are not the pessimistic case. They are the optimistic one.
The question has also moved into serious economic research. In a working paper published July 1, 2026, Ben Harris and William Overcash of the Brookings Institution, writing with Neil Mehrotra of the Federal Reserve Bank of Minneapolis, modeled directly whether AI-driven productivity growth can resolve an American fiscal trajectory that CBO projects will push public debt to 175 percent of GDP by 2056. Their finding deserves careful reading by anyone inclined toward either techno-optimism or despair. A once-in-a-generation productivity shock could cut annual deficits from roughly 6 percent of GDP to about 2 percent, which is genuinely enormous. But the authors identify five forces specific to an AI shock that would claw back more than half of that gain. Longer lifespans expand the retirement-age population and the entitlement spending attached to it. A shift in national income from higher-taxed labor toward lower-taxed capital narrows the tax base. Workers displaced by automation increase enrollment in income-support programs. Higher interest rates raise the cost of servicing existing debt. And an AI arms race between rival powers drives defense spending upward. Their conclusion is that AI can materially improve the budget outlook but is unlikely to resolve the imbalance on its own.
The wager, in other words, has been placed by institutions far more consequential than any technology executive giving a conference panel answer, and the most rigorous work available suggests the payout will fall short of the debt.
What the Productivity Actually Looks Like
The usual public conversation about AI fixates on employment, asking which jobs vanish and which industries get disrupted. Those questions matter enormously to the people living them, but they may be too narrow to capture what is being attempted. The larger economic promise is not simply the replacement of workers. It is the reduction in the cost of expertise itself.
Physicians, engineers, attorneys, teachers, programmers, financial analysts, and government administrators command high compensation in part because their competence requires years of training and accumulated experience. If software can perform meaningful portions of that work in seconds, the cost of delivering those services falls, and the services themselves reach people who previously could not afford them. Medical diagnosis could extend into communities that cannot support enough specialists. Legal and financial guidance that runs hundreds of dollars an hour could become nearly free at the margin. Education could adapt to each individual student. Robotics paired with capable AI could build housing, manufacture goods, harvest crops, move freight, and assist elderly and disabled people in countries that no longer have enough working-age citizens to do it.
The evidence so far supports the first half of that promise and withholds judgment on the second. The Bank for International Settlements, in its 2026 Annual Economic Report, found that task-level studies consistently document efficiency gains in the range of 20 to 50 percent in time savings. That is a real and substantial result, replicated across many settings. But the same report notes that aggregate productivity growth estimates remain far more conservative, generally under 1 percent over a long horizon, because adopting the technology at scale requires organizations to rebuild their processes around it, and that reconstruction takes years. The gap between what AI can do in a controlled task and what it has actually delivered to measured national output is the entire uncertainty on which the world’s fiscal arithmetic now rests.
Technology Does Not Repeal Scarcity
There is a structural weakness in the promise of near-universal abundance that no amount of computing power addresses. Artificial intelligence can multiply cognitive output, but it cannot manufacture additional land, generate energy without physical infrastructure, or eliminate the cost of raw materials. It cannot create housing in the specific places where millions of people want to live. And the technology itself is enormously physical in a way its digital branding conceals.
The International Energy Agency projects that global electricity consumption by data centers will roughly double from 485 terawatt-hours in 2025 to approximately 950 terawatt-hours by 2030, reaching around 3 percent of world electricity demand, with consumption at AI-focused facilities tripling over the same period. The supposedly weightless digital economy depends on power plants, transmission lines, semiconductors, cooling systems, water, critical minerals, factories, and land. Someone owns all of it.
That ownership question may be the most important one in the entire AI revolution, and it is where the debt story and the liberty story converge. Research published by IMF staff has found that AI could raise both income and wealth inequality, because those who own the technology or hold stakes in AI-driven industries capture increased capital income while displaced workers lose wages. An IMF Note published in April 2026, drawing on a scenario planning exercise, reached a similar conclusion, finding significant effects on wealth inequality as capital owners capture a large share of the gains, with scale economies in computing and data amplifying winner-take-most dynamics. These are staff analyses rather than official Fund policy positions, and that distinction is worth preserving, but the direction of the finding has been consistent across multiple independent efforts.
Cheaper goods and concentrated control over their production are not contradictory outcomes. They are the same outcome viewed from two angles, and only one of those angles is economic.
Who Controls the Abundance
Artificial intelligence may well help the world grow out of a portion of its debt burden. Higher productivity expands the economy, lowers the cost of delivering public services, and raises tax revenue without raising tax rates. The CBO has already banked a modest version of that effect, and the Brookings analysis suggests a larger version is possible under favorable assumptions.
But AI cannot repeal arithmetic. It cannot convert irresponsible spending into responsible spending. It cannot erase trillions in existing obligations without someone absorbing the loss. It cannot guarantee that productivity gains reach the households whose jobs were automated rather than accumulating with the owners of the machines. And it offers no guarantee whatsoever regarding liberty, which is the consideration that ought to concern us most.
Consider what a genuinely abundant, AI-mediated economy looks like from inside a household. If nearly everything arrives through platforms and systems controlled by a handful of corporations and the governments that regulate or partner with them, then access quietly replaces ownership. An account becomes a wallet, a digital identity becomes a permission slip, and behavior determines eligibility. A society in which everything is affordable can still be a society in which nothing is genuinely yours, and the transition from one to the other requires no dramatic announcement, no constitutional amendment, and no moment where citizens are asked to vote on it. It happens through terms of service.
This is why the argument over artificial intelligence cannot be reduced to whether the technology is good or evil. It is a tool of extraordinary power that may cure diseases, lift people out of poverty, expand human knowledge, and relieve workers from dangerous and exhausting labor. It could also become the mechanism through which financially strained governments manage populations they can no longer afford to govern in the traditional way, and the same infrastructure serves both purposes equally well.
The Great Wager
The world is accumulating obligations faster than its existing economy can comfortably support them, and the people responsible will not say so plainly. Technology leaders promise an age of abundance in which work becomes optional and money loses its meaning. Budget offices quietly write productivity dividends into their baselines. Central banks and research institutions model whether the dividend arrives in time. These are not separate developments. They are the same bet, placed by different players at different tables.
The financial system needs faster growth. Governments need lower service delivery costs. Corporations need higher productivity to justify a trillion dollars of capital expenditure. Aging nations need machines to perform work for which they no longer have enough workers. Artificial intelligence is being asked to solve all of it simultaneously, and the best available analysis says it will help considerably and still not be enough.
We should recognize the wager for what it is. Humanity is betting that artificial intelligence will create abundance faster than debt, inflation, inequality, and political dysfunction can consume it. The greatest danger may not be that AI destroys the world. The greatest danger may be that we become so desperate for AI to rescue the system that we hand over control of our lives to whoever controls the machines, and call it a bargain.
DISCLAIMER: This article is an opinion and analysis piece based on publicly available government data, economic projections, institutional research, and cited reporting. Forward-looking figures are projections, not guarantees, and economic conditions, fiscal policy, technological development, and government policy may change substantially. References to the potential effects of artificial intelligence on debt, employment, productivity, ownership, or individual liberty represent analysis of possible outcomes and should not be interpreted as predictions of certainty. Readers are encouraged to review the cited source material and reach their own conclusions.
Sources
- U.S. Department of the Treasury, Debt to the Penny — https://fiscaldata.treasury.gov/datasets/debt-to-the-penny/
- CBS News, “National debt tops $40 trillion after doubling in less than a decade, Treasury data shows” (August 21, 2026) — https://www.cbsnews.com/news/national-debt-tops-40-trillion-doubles/
- Congressional Budget Office, The Budget and Economic Outlook: 2026 to 2036 (February 2026) — https://www.cbo.gov/publication/61882
- Congressional Budget Office, Director’s Statement on the Budget and Economic Outlook for 2026 to 2036 — https://www.cbo.gov/publication/62050
- Peter G. Peterson Foundation, Interest Costs on the National Debt — https://www.pgpf.org/programs-and-projects/fiscal-policy/monthly-interest-tracker-national-debt/
- International Monetary Fund, Fiscal Monitor: Fiscal Policy under Pressure — High Debt, Rising Risks (April 2026) — https://www.imf.org/en/publications/fm/issues/2026/04/15/fiscal-monitor-april-2026
- Ben Harris, Neil Mehrotra, and William Overcash, “Can AI Restore Fiscal Sustainability in the US?” Brookings Institution working paper (July 1, 2026) — https://www.brookings.edu/articles/can-ai-restore-fiscal-sustainability-in-the-us/
- Bank for International Settlements, Annual Economic Report 2026, Chapter I: “Progress and peril” — https://www.bis.org/publ/arpdf/ar2026e1.htm
- International Energy Agency, Key Questions on Energy and AI — https://www.iea.org/reports/key-questions-on-energy-and-ai/executive-summary
- International Monetary Fund Staff Discussion Note, “Gen-AI: Artificial Intelligence and the Future of Work” (January 2024) — https://www.imf.org/en/Publications/Staff-Discussion-Notes/Issues/2024/01/14/Gen-AI-Artificial-Intelligence-and-the-Future-of-Work-542379
- International Monetary Fund Note, “Global Economic and Financial Implications of Artificial Intelligence: Lessons from a Scenario Planning Exercise” (2026) — https://www.imf.org/-/media/files/publications/imf-notes/2026/english/insea2026002.pdf
- Fortune, coverage of Elon Musk’s remarks at the U.S.-Saudi Investment Forum, November 19, 2025 — https://fortune.com/2025/11/20/elon-musk-tesla-ai-work-optional-money-irrelevant/
