Numbers are too big now. Just before Halloween, Nvidia made headlines for becoming the first $5 trillion company in history. A week later, Tesla’s shareholders approved a pay package for Elon Musk that could be worth about $1 trillion in a decade. At a certain point, figures this large become uncomfortably meaningless. A million dollars is a lot of money. A billion dollars is a heck of a lot—so much that if you had it, you’d be a billionaire. Now try to imagine how rich you would be if you were pulling in $8 billion every month for the next 10 years, as Musk is effectively about to do. It’s impossible. “The numbers are so big, they are hard to comprehend,” Jeff Sommer recently wrote in The New York Times. And he’s the paper’s finance columnist.
Sommer was referring to the stock market, which has been on an outrageous tear, with the gains concentrated among a tiny number of unfathomably valuable companies. The S&P 500 has doubled since October 2022, which is impressive on its own, but the combined market cap of the Magnificent Seven—Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia, and Tesla—has more than quadrupled. These firms are all heavily invested in generative AI, a technology based on training computers to make connections among quantities of data that are completely beyond human understanding. Large numbers are generating large numbers.
Records are being broken before they’ve had time to sink in. Nvidia’s $5 trillion valuation last month surpassed the first-ever $4 trillion valuation, which was also achieved by Nvidia, in July. The planet didn’t even have a single trillion-dollar company until Apple hit that milestone in 2018. This was considered a big deal at the time. Now the equivalent of the most valuable corporation in history circa 2018 can be generated in three months.
[Matteo Wong and Charlie Warzel: Here’s how the AI crash happens]
These wild numbers make all other sums sound trivial, even when they aren’t. On Tuesday, the Japanese conglomerate SoftBank disclosed that it had sold its entire $5.8 billion stake in Nvidia to fund other investments. That was important enough to make news basically everywhere. But we’re talking about one one-thousandth of Nvidia’s market cap here. Is $5.8 billion a lot, or is it lunch money?
As a species, we aren’t ready for this. People basically can’t tell the difference between any number that ends in –illion, and the more zeros you add, the worse things get. One experiment at the University of Richmond asked a group of students and graduates to plot numbers on a line. Half of them thought it made sense to evenly space 1,000, 1 million, and 1 billion. Another study asked people to rate the effectiveness of proposed COVID-relief packages. Responses differed sharply when the options were presented in per-capita amounts (giving everyone $1,200 was deemed much less effective than $24,000) but hardly at all when they were presented as total amounts ($100 billion versus $2 trillion), even though the ratio was the same in each case.
As the numbers grow, our understanding of reality dims. This is true even for some of our most influential thinkers. On Monday, Donald Trump posted on Truth Social that the United States would forfeit $2 trillion in revenue should the Supreme Court strike down his tariffs. The very next day, the president claimed that the number was $3 trillion. Even the smaller number would, if true, have been enough to completely wipe out the federal deficit. In fact, the U.S. has raised only about $200 billion in customs duties in the past fiscal year. The total U.S. debt, meanwhile, has risen to $38 trillion, a number so self-evidently absurd that the mind resists dwelling on it. It must be funny money. If it weren’t, it could never have gotten so big.
[Rogé Karma: The debt is about to matter again]
The prospects for democratic self-governance are not good. Raising and spending money is one of the most fundamental functions of government. If neither citizens nor their leaders can wrap their heads around these numbers, how will anyone make informed decisions about how the country is run? A trillion here, a trillion there, and pretty soon you’re talking real money.
Nor can our age of hyper-numbers be easily dismissed as the familiar product of inflation. Everyone grows up hearing about how cheap things were when our parents were kids—how you could buy a new Chrysler back then for the cost of a mattress today. Then you get older and have the same experience, puzzling over how movie tickets could possibly cost $16. But economic growth is generally understood to be exponential. The nature of exponential change is that it looks just like non-exponential change, right up until it doesn’t. Economists model growth as a constant percent, such as 3 percent annually. This sounds innocuous but leads to absolute numbers that elude our comprehension. If you plot out the value of an economy that grows by just a few percentage points a year, and you set a long-enough time frame, your chart will look like a plateau that suddenly explodes into a vertical line.
That is exactly what linear graphs of the U.S. stock market look like right now, which suggests that we have achieved escape velocity and entered the vertical part of the curve. (The geniuses who put historical stock-market graphs together tend to conceal this fact by using a logarithmic scale, which makes big jumps look much smaller.) The bigger the numbers get, the faster they grow in absolute terms. In just a few years, a $5 trillion valuation might sound as quaint as a $2,000 two-bedroom apartment in Brooklyn does today. The recent stock-market surge has brought about this moment sooner than it otherwise would have arrived, and a bursting bubble could defer it, but not forever. Nvidia could lose half of its value, and it would still be worth $2.5 trillion.
One solution, childish in its simplicity, is for the government to unilaterally deflate the currency by decreeing that all dollar figures will henceforth be reduced by an order of magnitude—that is, they will drop a zero—like a stock split on the dollar. This would render Nvidia a mere $500 billion company and buy us all some time to absorb that fact. This technique is known as redenomination. It has been used to reduce nominal prices in countries experiencing out-of-control inflation, but to my knowledge, it has never been deployed to ease the cognitive burden of a stock market that is performing too well. Alas, like all perfect ideas, this will never happen. Instead, the numbers will keep growing, faster and faster, and we will remain strapped against our will to this exponential rocket, hurtling into the Milky Way, which contains only about 400 billion stars, tops. But who’s counting?
The post Humans Can’t Wrap Their Minds Around This Economy appeared first on The Atlantic.
