Jensen Huang's recentremarks on AI's economic trajectory are as bold as they are inevitable.“There's a belief that the world's GDP is somehow limited at a hundred trilliondollars,” he said. “AI is going to cause that hundred trillion dollars tobecome two hundred, three hundred, five hundred trillion... Everybody's jobswill change.”The pitch isseductive, and on the micro level, largely correct. AI will not simply replacejobs; it will strip away friction. Workers will spend less time wranglingspreadsheets or typing prompts and more time orchestrating, deciding, andcreating. Productivity will surge. Those who fail to integrate AI will lose tothose who do.But macroeconomicsrarely bends to technological optimism. The real question is not whether AIexpands the economic pie. It is how that expansion prices out, and who capturesthe gains.Pressure-testingHuang's $500 trillion vision reveals two sharply different futures. One leadsto structural deflation and abundance. The other leads to inflationarydistortion.Scenario A: TheNominal BubbleIf the $500 trillionfigure is driven more by financial engineering than physical output, the resultcould be an inflationary shock.A booming AI sectorwould generate enormous paper wealth across companies such as NVIDIA,Microsoft, and OpenAI. Investors and founders would recycle those gains intoreal-world assets: housing, energy, food, and commodities. That is classicdemand-pull inflation, amplified by unprecedented liquidity.At the same time, AI'sdigital promise collides with physical bottlenecks. Training models requiresvast amounts of copper, semiconductors, data centers, and electricity.Competition for those constrained resources pushes up costs across the broadereconomy while non-AI sectors struggle to keep pace.In this scenario, the$500 trillion economy is not real growth. It is a valuation bubble chasingfinite real-world supply.Scenario B: TheDeflationary EngineThe counterargument isthat AI could create genuine GDP expansion while driving structural deflation.GDP is ultimatelyprice multiplied by quantity. If AI removes the constraints of human labor andintelligence, the quantity of goods and services could scale dramatically evenas prices fall.When AI automatescoding, legal work, diagnostics, research, and eventually physical productionthrough robotics and automated manufacturing, the marginal cost of creatingproducts and services collapses. Software, logistics, energy optimization, andeven manufacturing become radically cheaper.If output expandsseveralfold while costs decline, the economy grows in real terms. Living costsfall, purchasing power rises, and abundance—not inflation—defines the outcome.This is the futureHuang is implicitly betting on. And mathematically, it is possible.The DangerousTransition GapThe real risk liesbetween those two scenarios.Markets may price inAI-driven abundance long before the physical infrastructure exists to supportit. Building advanced energy grids, semiconductor fabs, robotics supply chains,and transmission networks could take 10 to 15 years.Tariffs and the war aren't helping inflation, but one that not many expected to see was the AI story impacting inflation. Here's PPI (which flows into PCE) for semiconductors and other electrical components. pic.twitter.com/8R6b9ZqqrE— Ryan Detrick, CMT (@RyanDetrick) May 21, 2026That creates adangerous mismatch. Capital floods into AI today, asset prices surge, andresource competition intensifies before supply-side abundance arrives. Energy,housing, metals, and essential goods could all become more expensive during thetransition.In effect, the path toabundance may first pass through inflation.Central banks wouldface an impossible balancing act between suppressing inflation and supportinggrowth. Workers in disrupted industries could face displacement before newAI-augmented roles scale fast enough to absorb them. Social and politicalfriction could undermine the productivity boom AI promises.Abundance is notautomatic. It has to be engineered.The Real QuestionHuang is probablyright that GDP is not capped at $100 trillion. He is also right that AI willfundamentally change how people work.But whether the worldreaches $500 trillion through abundance or distortion will depend less onalgorithms and more on institutions.The outcome will hingeon energy policy, industrial capacity, monetary discipline, and laboradaptation. Technology creates productive capacity. Governments, central banks,and markets determine whether that capacity translates into stability.AI will reshape theglobal economy. The real question is whether society can manage the transitionas effectively as it trains the models powering it.This article was written by Anndy Lian at www.financemagnates.com.