[Featured Story] Wall Street is sweating bullets. AI is not just stealing jobs anymore — it is out to “destroy” entire companies. Claude is rewriting legacy code and making IBM stock tremble. ChatGPT has moved into Excel and is about to automate investment banking. Now, every Wall Street firm faces a brutal shakeout.
Recently, an unprecedented anxiety has spread across all of Wall Street.
People used to worry that AI would take workers’ jobs. But now, investment banks are waking up to something far scarier:
AI could make entire companies, even entire industries, vanish overnight.
They are calling it the force of “creative destruction.”
Just as the internet killed Blockbuster and Kodak, AI is now coming for modern businesses in a massive reshuffle.
And over the past two weeks, three heavy punches landed one after another, turning this concept from an economics textbook into a real trading floor nightmare.
GPT-5.4 Invades Excel: Wall Street Is About to Change Forever
This time it is not about writing poems, chatting, or drawing pictures. AI is heading straight for the most sacred battlefield on Wall Street: Excel.
OpenAI simultaneously launched the ChatGPT for Excel plugin, letting GPT-5.4 work directly inside spreadsheets.
Investment analysts can now call up ChatGPT inside Microsoft Excel and Google Sheets to build financial models and generate investment memos.
Just how powerful is GPT-5.4?
In OpenAI’s internal investment banking benchmark tests, GPT-5.4 Thinking jumped from the original GPT-5’s 43.7% to 87.3% — it doubled its score.
In the GDPval evaluation across 44 professions, it matched or beat human industry professionals in 83% of cases.
And for the first time, GPT-5.4 has built-in “native computer control” in a general-purpose model.
It can operate desktop software like a human, using mouse and keyboard to complete tasks. On the OSWorld desktop navigation test, GPT-5.4 scored 75.0%, while the human average is 72.4%.
OpenAI also brought in a lineup of major financial data platforms — FactSet, Third Bridge, Moody’s, Dow Jones Factiva, MSCI — all integrated into ChatGPT.
ai slut
The era of staying up all night manually reconciling spreadsheets may truly be over, killed by GPT-5.4.
Claude Strikes: Big Tech Stocks Shake
GPT-5.4 is coming for senior analysts. But two weeks ago, Anthropic’s Claude aimed best free ai porn at something even deeper: the infrastructure itself.
Anthropic officially announced that Claude Code can automate the modernization of COBOL legacy code. That same day, IBM stock crashed 13.2% — its worst single day since 2000.
A single blog post wiped out about $40 billion in market value.
How can one blog post have such devastating power?
COBOL supports IBM’s entire business model. Most banks, insurance companies, and government systems worldwide still run on this programming language born in the late 1950s.
However, this legacy code was written in the 1960s and 1970s, and almost no one understands it anymore. Rewriting it would cost millions of dollars and take years.
So these companies have no choice but to send billions of dollars to IBM every year, begging them to keep maintaining it.
Then Anthropic said: “Claude can handle it right now.”
Link: https://claude.com/blog/how-ai-helps-break-cost-barrier-cobol-modernization
Industry leaders quickly weighed in: the barrier to modernizing legacy code has always been understanding it.
Although COBOL has become a dead language, Claude has driven the cost of understanding it down to zero.
In other words, banks and governments no longer need to pay IBM massive maintenance fees.
Of course, IBM fought back immediately, claiming that “translating COBOL is the easy part. The real work is data architecture redesign, runtime replacement, and transaction processing integrity.” But the market had already voted with its feet.
Bloomberg data shows IBM fell 27% in February, its largest monthly drop since at least 1968.
Block Cuts 40% of Staff: Stock Jumps 25%
Claude is dismantling infrastructure. GPT-5.4 is replacing senior talent. And then there is this:
Fewer people equals good news.
On February 27, Block CEO Jack Dorsey announced layoffs of over 4,000 people, cutting from more than 10,000 to under 6,000. Forty percent of the workforce, gone just like that.
Dorsey put it bluntly:
“Intelligent tools have already changed what it means to build and run a company. Smaller, flatter teams, combined with the tools we are building, can do more and do it better.”
He even predicted: “Within the next year, most companies will reach the same conclusion and make similar structural adjustments.”
The market’s reaction? Block stock surged 25% after hours, adding about $8 billion in market value.
This is one of the largest AI-driven layoffs in S&P 500 history. The more aggressively a company cuts jobs, the more its stock rises — a signal more glaring than any research report.
Creative Destruction: An Old Word, A New Catastrophe
Claude crushing IBM, Block cutting 40% of its staff, GPT-5.4 taking over Excel — three events in two weeks, all pointing to the same economic concept:
Creative Destruction.
Economist Joseph Schumpeter wrote in 1942 that the essence of capitalism is not price competition, but the destructive replacement of the old world by new products, new technologies, and new organizational forms.
This word is not new.
Blockbuster, with over 9,000 stores at its peak and a market value in the tens of billions, went bankrupt in 2010 after Netflix arrived.
Kodak, the absolute king of the film era with 140,000 employees at its peak, filed for bankruptcy protection in 2012 after digital cameras emerged. The irony? Kodak itself invented the digital camera.
Travel agencies, stockbrokers, classified ads, print media, video rental stores — the IT revolution of the 1990s wiped out a whole batch of industries that once seemed indestructible.
And now, what AI is about to do may be ten times bigger than what the internet did.
Anton Korinek, an AI expert at the University of Virginia, asserts: the internet only changed how information is distributed. AI is attacking “cognitive production” itself. The scale of impact is completely different.
The internet replaced information movers: travel agents moved flight information, newspapers moved news information, Blockbuster moved movie information.
AI is replacing information processors: analysts process data, lawyers process laws, programmers process logic, doctors process symptoms.
This is the key difference of “Creative Destruction 2.0”:
It is not targeting manual labor, but brain work. Not blue-collar jobs, but white-collar elites. Not the execution layer, but the decision-making layer.
The Tug of War Between Optimism and Panic
Of course, doomsday narratives always run faster than the real world.
By coincidence, on the very same day Claude hammered IBM, another report dropped — and software stocks crashed across the board.
This report, framed as a “macro memo from 2028,” projected a scenario where AI triggers massive white-collar unemployment, the S&P 500 crashes 38%, and unemployment soars to 10.2% over the next two years.
It even has its own Wikipedia page now.
But on the other end of Wall Street, hedge fund giant Citadel Securities quickly fired back.
Citadel strategy analyst Frank Flight dismantled the Citrini report point by point, arguing that it ignores macroeconomic fundamentals and the technology adoption curve, and is essentially “ignorant panic about AI.”
Citadel cited the Keynes case: “Keynes underestimated the elasticity of human desire.”
Every technological revolution eliminates old jobs, but humans always create new demands and new industries.
What both sides are really arguing about is the most fundamental disagreement in today’s market.
The optimists believe: AI-driven productivity gains will fuel long-term economic growth.
Last quarter, US labor productivity grew 2.8%, averaging 2.8% since early 2023 — more than double the pace of the decade before 2019.
More efficient labor means higher profits, more new professions, and a higher standard of living.
Historically, every technological revolution has ultimately created more jobs than it destroyed.
The pessimists warn: the “growing pains” of the transition may far exceed expectations.
Nobel laureate and MIT professor Simon Johnson points out that corporate bankruptcies could trigger a chain reaction of credit risk.
“The last thing you want is a collapse of the credit system, or risk spreading inside banks.”
And Daniel Keum at Columbia Business School spotted a more subtle signal: in earnings calls, bosses are increasingly referring to employees as “costs” rather than “assets.”
Even before mass layoffs, companies are quietly cutting health benefits, remote work stipends, and even free snacks in the office.
Before they touch your paycheck, they always cut these fringe benefits first.
Is This Time Really Different?
Richmond Fed President Tom Barkin, facing the “AI doom” narrative, said one thing:
“This has been happening in America for hundreds of years.”
He is right.
From steam engines to electricity, from assembly lines to the internet, every technological revolution has been accompanied by industry shakeouts, corporate bankruptcies, and mass unemployment.
But in the end, the economy always rebuilds from the ruins, and builds higher.
The question is: how long is “in the end”?
For the 100,000 people laid off by Kodak, for the tens of thousands of store workers who lost their jobs when Blockbuster closed, “the economy will recover in the long run” meant absolutely nothing.
And this time, AI is targeting the very thing humans are most proud of — thinking, analyzing, judging, and creating.
When Claude can read sixty-year-old COBOL code, when GPT-5.4 can build an 87.3-point investment banking model in Excel, when a company becomes more valuable after cutting 40% of its people —
The rules of the game have changed.
The only question left is this: are you standing on the side of “creation,” or are you about to become one of the “destroyed”?
