ChatGPT Halved Programmer Jobs — BigTech AI Stocks Up 50%

Two separate 50% figures define America's AI economy in 2026. The Philadelphia Semiconductor Index ETF (SOX — an exchange-traded fund that tracks the biggest chipmakers, from Nvidia to Broadcom to TSMC) gained approximately 50% in the first four months of the year. A Federal Reserve study published in March 2026 found that US programmer employment has fallen roughly 50% since ChatGPT launched in November 2022. Same economy. Opposite directions. The gap between those two data points is the story that mainstream AI coverage keeps skipping.

This divergence is not a projection or a forecast risk. It has already happened — documented by the Federal Reserve, tracked by Pew Research and Gallup, and confirmed by capital allocation decisions worth tens of billions of dollars. Understanding what drove it is the most practically useful thing a developer, designer, or knowledge worker can do right now.

The Two AI Economy Charts Nobody Puts Side by Side

Before ChatGPT's public launch in November 2022, US software programming roles were growing at approximately 5% per year — well above the broader labor market's 1–2% annual growth. Tech careers were among the most recession-resistant in the American economy.

Federal Reserve economists Crane and Soto published a study in March 2026 documenting what happened next: programmer employment collapsed by approximately 50% after ChatGPT's arrival. Three years of AI deployment effectively reversed more than four years of consistent job growth in a single sector.

At the exact same moment, the companies that build and sell the chips (specialized processors) required to run AI told a completely different story:

• The Philadelphia Semiconductor ETF (SOX) gained ~50% in just the first four months of 2026 — a pace analysts described as "historical"
• Nvidia, which manufactures the H100 and H200 GPUs (graphics processing units originally designed for gaming, now the primary hardware for AI model training), faces demand it cannot fully meet
• Meta revised its capital expenditure guidance (the annual budget committed to building AI infrastructure — data centers, chips, fiber networks) upward multiple times in early 2026
• Cloud computing revenue across major tech companies is described as "meaningfully accelerating" driven by AI workloads
• TSMC (Taiwan Semiconductor Manufacturing Company — the foundry that physically fabricates chips for Apple, Nvidia, Google, and most of the AI industry) is reporting accelerating demand alongside strong forward bookings

Both trends exist in the same economy, generated by the same technological shift. The programmer employment collapse and the semiconductor market rally are not separate stories — they are two readings of the same event, measured from opposite vantage points.

How Seven Companies Became the Landlords of AI Automation

The AI Supremacy newsletter, which has tracked AI's evolution for 4.5 years — initially framing it as a US-China geopolitical competition — published a blunt reassessment in late April 2026: the real race was never between nations. It was between a handful of American hyperscalers (cloud computing companies that operate at near-continental scale, running hundreds of thousands of servers across global data centers) and everyone else.

The newsletter identifies seven companies as the true "landowners" of AI: Google, Meta, Amazon, Microsoft, Apple, Netflix, and Tesla. Their structural advantages go beyond ordinary market concentration:

• Compute control: These companies own or control the majority of the GPU clusters (interconnected groups of AI chips) that model training and deployment require. Even OpenAI and Anthropic — officially independent AI labs — must rent compute from Microsoft Azure and Amazon Web Services respectively.
• Equity ownership of competitors: Microsoft holds a major stake in OpenAI. Both Amazon and Google are Anthropic investors. The hyperscalers fund the very companies they ostensibly compete against — ensuring that even independent AI success flows back to BigTech balance sheets.
• Revenue flywheel: AI increases demand for cloud computing services, which increases hyperscaler revenue, which funds more AI infrastructure — a self-reinforcing cycle that accelerates as adoption spreads. There is no obvious exit from this loop for companies outside it.

As the newsletter's founder stated directly: "Just a few people at a few Big tech companies control the future of AI, not anyone else."

The degree of consolidation is visible even in competitive behavior. Google Deepmind reportedly launched a dedicated "strike team" specifically to counter Cursor (Anysphere's AI coding platform) — a fact worth pausing on. A company with a multi-trillion dollar market cap felt threatened enough by a startup to form a special competitive unit targeting it. That is how contested, and how concentrated, the AI application layer has already become.

Programmer Employment: The Workers Who Didn't Make the Earnings Slides

The 50% programmer employment figure deserves more than a footnote. Before November 2022, a computer science degree was one of the most reliable paths into the American middle class. Entry-level programming roles paid above-median wages. The sector grew through economic downturns that battered other industries. Young graduates from second- and third-tier schools could enter the field, build skills, and expect upward mobility over a decade-long career arc.

The Federal Reserve study (Crane and Soto, March 2026) identifies a roughly 50% employment decline but does not break it down by seniority, geography, or specialty — a genuine limitation. The scale of the figure, however, is too large to attribute to cyclical hiring fluctuations or normal market correction.

What adjacent data and industry reporting suggest about where the losses concentrated:

• Entry-level coding roles are increasingly automated before new graduates can compete for them
• Cursor markets itself on the explicit proposition that "1 engineer = former teams" — one developer with AI assistance replacing what previously required multiple hires
• Junior developer hiring slowed sharply at major companies from 2023 onward, as existing staff were equipped with AI tools rather than headcount being expanded
• Freelance coding platforms report collapsed demand for basic programming tasks that AI now completes in seconds for pennies

The newsletter's founder explicitly identified this as the story he "refuses not to tell," contrasting it with what he calls "VC-media on the rise" — venture capital-backed (private investment firm-funded) publications and BigTech-affiliated think tanks focused on AI upside while underreporting labor impact. His direct assessment: "The bifurcation of young people, post-graduate entry level employment and AI really has me worried."

Gen Z AI Enthusiasm at 22%: The Generation That Is Reading the Room

If the labor data describes what happened, the sentiment surveys measure how people feel about it — and what they feel is almost precisely the inverse of how AI is marketed by the industry benefiting from it.

Three major datasets from 2025–2026 converge on the same finding:

• 50% of US adults say they are "more concerned than excited" about AI's increasing presence in daily life (Pew Research Center)
• Only 10% of US adults describe themselves as primarily excited about AI — meaning the concerned segment outnumbers the excited segment by 5 to 1
• 18% of US workers now believe their job is very or somewhat likely to be eliminated by AI within five years — up from 15% in mid-2025, a 3-point increase in under 12 months
• Gen Z enthusiasm for AI: just 22% (Gallup) — the lowest of any demographic expected to be tech-forward

That last figure is the most analytically striking. Gen Z is the generation that adopted TikTok in months, normalized the smartphone as a primary interface for every domain of life, and built careers on digital-native platforms from day one. If any demographic was expected to embrace AI enthusiastically, it was them.

The newsletter frames their skepticism as rational assessment rather than reflexive technophobia. Gen Z is entering a labor market where entry-level programming and knowledge-work positions — the traditional career on-ramps for college graduates entering technology — have contracted substantially since 2022. For this cohort, AI is not an opportunity still arriving. It arrived before they did, and it occupied the positions they were heading toward.

There is a secondary implication the newsletter flags: Gen Z is actively leaving advertising-supported social platforms as AI-generated content degrades information quality across those networks. Digital advertising revenue is the primary funding mechanism for BigTech's AI infrastructure buildout. A sustained Gen Z exodus from advertising-dependent platforms would eventually affect the revenue flywheel that powers the entire hyperscaler model — a structural vulnerability baked into the current setup that the industry does not publicly acknowledge.

The $60 Billion Coding Bet — And the Programmer It Depends On Disappearing

The sharpest paradox in today's AI economy: the tool most directly associated with reducing programmer headcount is now valued at figures that would have placed it among America's most valuable companies as recently as a decade ago.

Cursor, built by the startup Anysphere, is an AI-powered code editor (a software development environment where an AI model suggests, writes, and completes code in real time alongside the human developer). Its core market proposition is explicit: one engineer using Cursor can produce the output that formerly required entire development teams. The product's tagline is not subtle about the labor implications.

This shift is broadly called vibe coding — an AI automation workflow where tools like Cursor and Claude Code generate functional software from natural language descriptions, enabling a single developer to ship what previously required whole teams. It is the clearest expression of why programmer employment figures and AI investment figures are moving in opposite directions simultaneously.

The reported figures around Cursor's current trajectory:

• Cursor was independently seeking $2 billion in new funding at a $50 billion valuation before the SpaceX discussions emerged
• SpaceX reportedly claimed a $60 billion acquisition option on Cursor — the stated rationale: combining Cursor's distribution to expert software engineers with SpaceX's Colossus supercomputer (approximately one million H100-equivalent chips — Nvidia's highest-performance AI training processors)
• Google Deepmind launched a dedicated competitive response team specifically targeting Cursor's enterprise momentum, confirming that even trillion-dollar AI labs view a coding tool startup as a first-order strategic threat requiring a dedicated response

Context: The SpaceX acquisition option has been reported but not confirmed by official announcement at time of writing. The $50 billion independent valuation was separately reported ahead of the SpaceX discussions and is consistent with multiple sources covering the funding round.

For developers, designers, and knowledge workers trying to interpret these signals: capital markets are placing $50–60 billion on the premise that per-engineer software output will increase dramatically — meaning fewer engineers are needed per project. This is not speculative analysis. It is a priced investment thesis with named buyers and disclosed valuations. The question is not whether this thesis will play out, but how quickly and unevenly.

The practical path forward is not to compete with AI on raw code generation — AI will win that race. The leverage point is building skills in directing, evaluating, and deploying AI-generated systems — the layer that still requires human judgment. Explore the AI automation skill guides on this site for a structured path through the tools that matter right now, or follow the latest AI developments for ongoing signals. The window for proactive adaptation is still open. The 50% employment figure suggests it is not unlimited.

Related Content — Get Started | Guides | More News