1,000 days and counting: How ChatGPT’s launch changed Silicon Valley and beyond forever

Three years back, November 30, 2022, OpenAI released what they called a “low-key research preview.” Wasn’t even pitched as a real product launch. Some of its employees were getting cold feet about shipping something that wasn’t quite ready, but Sam Altman said went ahead. “Language interfaces are going to be a big deal,” he mentioned at the time. Altman’s known for going big with predictions, but that one? Probably his most low-key take ever.ChatGPT. That’s what they named it. A million people jumped on it in five days flat. Now? 800 million use it every week. But honestly, ChatGPT’s success isn’t even the interesting part anymore—it’s watching everything that happened after. And there’s this perfect irony here: the company that lit the match is now right in the middle of the fire they started (more on that a little later).

When AI became something everyone could actually use

AI wasn’t some brand new thing when ChatGPT showed up. It’d been around forever—your Netflix picks, Siri talking back at you, those self-driving cars people kept crashing. Most of us just never paid attention. AI was like plumbing. You know it’s there, keeps things running, but who thinks about pipes?ChatGPT flipped that switch overnight. Millions of regular people—not programmers, not tech nerds—could sit down and just chat with this thing. Watch it write their kid’s homework, fix broken code, throw together a business pitch, pump out a sonnet. The tech underneath? Not really new. These big language models existed already. What changed was how it felt to use one. First time AI seemed less like a tool and more like magic you could actually touch.

Sundar Pichai, CEO of Google, went on 60 Minutes and said, “I’ve always thought of AI as the most profound technology humanity is working on. More profound than fire or electricity or anything that we’ve done in the past.” He’d been saying versions of that for years, but ChatGPT made people actually get it. Not just understand it—feel it.This whole shift from AI that sits in the background analszing stuff to AI that makes things, creates content—that became the new line in the sand. And, it kicked off a feeding frenzy.Google threw together Bard in a hurry, they renamed it Gemini later. Anthropic got Claude out the door. Meta open-sourced LLaMA, and in the last few years it put its chatbot right at the centre of every social media app it has—whether it be WhatsApp, Instagram or Facebook. Microsoft started jamming ChatGPT into everything they could. Even, Apple had partner up with OpenAI, because it couldn’t figure out AI on its own. Everyone wanted in. Nobody wanted to miss the boat.

The arms race nobody saw coming

Start of 2025, OpenAI looked untouchable. By end of January? Not so much.China’s DeepSeek came out of nowhere with models matching OpenAI’s stuff, claimed they spent under $6 million training them. Six million. That number exploded a lot of heads in Silicon Valley. Everyone thought you needed billions to play this game. Turns out maybe you don’t.Then November 2025 hit. Google launched Gemini 3. Marc Benioff from Salesforce tested it for two hours, came out saying he’s done with the other stuff. Gemini 3 was beating OpenAI’s best models on the tests people care about. Had more computing muscle behind it. Just bigger.

Meanwhile OpenAI finally dropped GPT-5 in August. Two years they’d been hyping this thing. It face-planted pretty hard. Started calling Oklahoma “Gelahbrin” on maps. Couldn’t do algebra a seventh grader could handle. There were these prediction markets giving OpenAI 75% odds of having the best model. That number crashed to 14% in an hour. Brutal.Gary Marcus—he’s been skeptical about this AI hype from the start—put it pretty plain: “OpenAI has basically squandered the technical lead it once had; Google has caught up.” A lot of people were thinking it. He just said it.And it’s not just a Bay Area thing anymore. DeepSeek’s latest model is matching Gemini 3 Pro at thinking through complex stuff. Hit gold medal level on the International Math Olympiad test. Before that, only OpenAI and Google’s internal models pulled that off.Gets messier when you zoom out. US slapped tariffs on fancy chips to slow down China’s AI push. China told companies like Huawei to figure it out themselves. Abu Dhabi’s G42 and some other Middle East players are jumping in now too. Whole thing’s turning into a proper geopolitical mess.

Nvidia’s Trillion-Dollar Moment

If ChatGPT lit the fuse, Nvidia’s the one selling dynamite to everyone. Their stock? Up 979% since ChatGPT dropped. July 2025, first company ever to hit $4 trillion. Three months later, $5 trillion.Pretty simple reason why. Nvidia makes the chips—GPUs—that run all this AI stuff. Training these big models takes insane amounts of computing power. Nvidia basically owns that market. Jensen Huang, their CEO, said building one gigawatt of data center capacity runs you $50-60 billion. About $35 billion of that goes straight to Nvidia for chips and systems. He called his own company “unbelievable” which is a bit much, but then added AI is “real” and makes companies “hugely productive.” Can’t really argue with the stock price.The numbers get stupid big fast. Nvidia agreed to invest up to $100 billion in OpenAI. That money’s for AI data centers that need 10 gigawatts of power. That’s enough electricity for a decent-sized city. OpenAI’s buying millions of Nvidia GPUs back. We’re talking 4-5 million graphics processors. That’s double what Nvidia shipped in an entire year before.Not just them either. Nvidia, Microsoft, and Elon’s xAI bought Aligned Data Centers for $40 billion. Biggest data center deal ever. OpenAI, Oracle, CoreWeave, xAI—they’re all making these massive deals for computing power and buildings to put it in.

Nvidia’s trillion-dollar moment

If ChatGPT lit the fuse, Nvidia’s the one selling dynamite to everyone. Their stock? Up 979% since ChatGPT dropped. July 2025, first company ever to hit $4 trillion. Three months later, $5 trillion.Pretty simple reason why. Nvidia makes the chips—GPUs—that run all this AI stuff. Training these big models takes insane amounts of computing power. Nvidia basically owns that market. Jensen Huang, their CEO, said building one gigawatt of data center capacity runs you $50-60 billion. About $35 billion of that goes straight to Nvidia for chips and systems. He called his own company “unbelievable” which is a bit much, but then added AI is “real” and makes companies “hugely productive.” Can’t really argue with the stock price.

The numbers get stupid big fast. Nvidia agreed to invest up to $100 billion in OpenAI. That money’s for AI data centers that need 10 gigawatts of power. That’s enough electricity for a decent-sized city. OpenAI’s buying millions of Nvidia GPUs back. We’re talking 4-5 million graphics processors. That’s double what Nvidia shipped in an entire year before.Not just them either. Nvidia, Microsoft, and Elon’s xAI bought Aligned Data Centers for $40 billion. Biggest data center deal ever. OpenAI, Oracle, CoreWeave, xAI—they’re all making these massive deals for computing power and buildings to put it in.

The jobs question

Minute ChatGPT launched, people started panicking about losing work. By 2023, you could see it happening. Artists watching their commissions disappear. Web developers competing with code that writes itself. Writers seeing AI take over stuff they used to get paid for.Numbers don’t lie. 2025 so far: 626 layoffs at tech companies, 182,963 people out of work. That’s 579 people losing jobs every single day. 2024 was worse—1,115 layoffs, 239,101 people.Big Tech went first. Microsoft cut 15,000 people this year. Satya Nadella said their AI tools like GitHub Copilot write 30% of Microsoft’s code now. IBM dropped 8,000, mostly HR people, replaced them with a chatbot called AskHR. Meta fired almost 4,000, said they were cutting “low performers” while hiring more machine learning engineers. Corporate speak for “AI’s doing your job now.”Andy Jassy at Amazon sent a memo saying they’ll need fewer people in a lot of roles because of AI agents. New York Times got hold of Amazon’s internal plans. Said robots might eventually replace 600,000 future jobs. Six hundred thousand jobs that’ll never exist because robots.Jamie Dimon runs JPMorgan Chase. They’ve spent $2 billion on AI since 2012. He told Fortune’s Most Powerful Women Summit straight up: “It will eliminate jobs. I think people should stop sticking their heads in the sand. So did tractors and so did cars.”But Dimon thinks long-term it works out. Told people at the America Business Forum in Miami, “My guess is the developed world will be working three-and-a-half days a week in 20, 30, 40 years, and have wonderful lives. It’s going to affect every application, every job, every customer interface.”2025 numbers: AI killed 77,999 jobs across 342 tech layoffs. 92% of IT jobs are getting changed by AI somehow. Customer service reps are most at risk—80% automation rate by 2025. Data entry clerks could see 7.5 million jobs gone by 2027.Here’s where it gets complicated though. Yeah, 85 million jobs disappearing by 2025. But 97 million new ones showing up. Net gain of 12 million jobs globally. Sounds good until you realize these aren’t the same jobs for the same people in the same places. New jobs need new skills. Different cities. Sometimes different people entirely.Goldman Sachs figures if AI expands like everyone thinks, about 2.5% of US jobs are seriously at risk. Programmers, accountants, auditors, legal assistants, admin assistants, customer service people—all on the list.World Economic Forum says 41% of employers worldwide are planning to shrink their workforce over the next five years because of AI.Hits different ages different ways. Workers 18-24 are 129% more likely than people over 65 to worry AI will make their job obsolete. Half of Gen Z job hunters think AI already tanked what their college degree is worth. Entry-level positions are screwed—50 million US jobs at risk. Big Tech companies cut new graduate hiring by 25% in 2024 versus 2023.Sundar Pichai told the BBC no job’s safe. Not even his. Called being CEO “one of the easier things” for AI to take over. Then said, “AI is the most profound technology humanity is ever working on, and it has potential for extraordinary benefits, and we will have to work through societal disruption. I think people who learn to adopt and adapt to AI will do better.”Some jobs are safer though. Construction, skilled trades—harder to automate. Healthcare and personal services are actually growing because AI helps instead of replacing. Nurse practitioners projected to grow 52% from 2023 to 2033.Business numbers tell a story. March 2024, only 27% of CFOs said they got really good ROI from generative AI. Late 2025? That jumped to 90%. Companies figured out how to actually use it.Skills gap is nasty. 77% of new AI jobs want master’s degrees. 350,000 new positions opening up—prompt engineers, human-AI collaboration specialists, AI ethics officers. By 2030, 70% of job skills expected to change. Better keep learning.

Governments scramble to catch up

Once AI’s impact got impossible to ignore, governments tried writing rules. Couldn’t get everyone to agree on anything.EU went first. AI Act started August 1, 2024. Takes a risk-based approach—bans sketchy stuff like real-time face tracking and social scoring. AI companies that could cause big problems have to put in safeguards, be transparent, respect copyrights. Full enforcement by August 2026.US went fifty different directions. Trump issued an Executive Order January 2025, basically threw out Biden’s old AI order. Wanted federal departments to focus on “enhancing America’s global AI dominance.” Administration published America’s AI Action Plan in July—90+ federal policy actions. Whatever that means.No big federal law, so states jumped in. 2025 alone, 38 states passed about 100 different AI laws. Total mess of contradicting rules. Colorado’s AI Act went furthest—tough requirements for risky AI systems starting February 2026.House Republicans tried putting a 10-year ban on state and local AI regulations. July 1, 2025, Senate voted 99-1 to kill it. Pretty rare to see that kind of agreement.Other countries doing their own thing. Canada’s got AIDA—Artificial Intelligence and Data Act—for high-risk AI systems. China makes you get algorithms approved first, makes sure they align with government ideology. Shows how AI rules reflect bigger political tensions.

Are we in a bubble?

Money’s hitting crazy levels. People started asking the uncomfortable question: is this real or are we all just pretending?Look at Nvidia and OpenAI’s deal. Nvidia invests up to $100 billion in OpenAI for data centers. OpenAI promises to buy millions of Nvidia GPUs to fill those data centers. Nvidia’s basically financing their own sales. Pretty slick if it works.Happens everywhere in AI. Nvidia invested in CoreWeave. CoreWeave supplies data centers to OpenAI. CoreWeave also buys from Nvidia. Nvidia owns about 7% of CoreWeave—worth maybe $3 billion. Microsoft poured money into OpenAI. OpenAI uses Microsoft Azure. Oracle made massive cloud deals with OpenAI. Oracle buys tons of Nvidia hardware.2020 to 2025, Nvidia invested roughly $53 billion in AI companies. 170 deals. Some analysts figure for every $10 billion Nvidia puts into OpenAI, they might see $35 billion in GPU purchases back. That’s 27% of their annual revenue. Convenient.People compare it to the dot-com bubble. Nortel, Lucent, Cisco—telecom equipment makers lending money to startups so those startups could buy their equipment. Before everything collapsed in 2001, those loans hit over 10% of annual revenues. Sound familiar?Even the big players admit something’s off. Sam Altman at a dinner with journalists in August: “Someone is going to lose a phenomenal amount of money in AI.” Bret Taylor, runs Sierra and sits on OpenAI’s board, straight up said “we are in a bubble.” Compared it to the dot-com boom late ‘90s.Sundar Pichai gave the BBC a similar warning in November. Said while AI investment growth has been “extraordinary,” there’s definitely some “irrationality” in the current boom. When asked if Google would be immune if the bubble burst, he didn’t sugarcoat it: “I think no company is going to be immune, including us.”He compared it to the internet boom. “We can look back at the internet right now. There was clearly a lot of excess investment, but none of us would question whether the internet was profound. I expect AI to be the same. So I think it’s both rational and there are elements of irrationality through a moment like this.” His comments echoed what Federal Reserve chairman Alan Greenspan said back in 1996 about “irrational exuberance” in the market—right before the dotcom crash.AI gains reshaped the whole stock market. Seven biggest S&P 500 companies—Nvidia, Microsoft, Apple, Alphabet, Amazon, Meta, Broadcom—now 35% of the index. Was 20% three years ago. Their growth is almost half the entire benchmark’s 64% increase since ChatGPT launched.Math gets ugly. Tech companies in S&P 500 spending $400 billion or more per year on AI going forward. OpenAI, the biggest AI company, reported $13 billion in revenue. This massive buildout better start making real money soon or valuations are going to get crushed.Jamie Dimon’s bank has 2,000 people building AI systems. 150,000 employees using large language models weekly. He told Fortune, “You can’t look at AI as a bubble, though some of these things may be in the bubble.” Compared it to early internet days. Hundreds of companies worth billions went under. But we got Facebook, YouTube, Google out of it.McKinsey found eight in ten companies using generative AI somewhere. But about the same number say they haven’t seen any real financial impact. Gap between using it and getting results means companies still figuring out how to make it actually work.Some people staying optimistic. Bret Taylor said while individual companies might fail, “AI will transform the economy, and I think it will, like the internet, create huge amounts of economic value in the future.”

What comes next

Thousand days since ChatGPT launched. It did what most tech can’t—changed the world’s direction, kicked off international competition, reshaped markets, made governments scramble for regulations, forced millions to rethink what intelligence and creativity even mean.Ironic part? Company that started this is now fighting to stay relevant. Right at the three-year mark, Sam Altman declared “code red” inside OpenAI. Paused work on making money to fix core ChatGPT performance. Weird echo of December 2022 when Google declared their own “code red” after ChatGPT launched.The reversal stings. OpenAI maybe lost $12 billion just in Q3 2025. They’re valued at $500 billion. Committed to spending $1.4 trillion developing their product. But 2025 revenue? Just $13 billion. Math doesn’t add up.Google’s Gemini app: 650 million monthly users. ChatGPT: 800 million weekly users. Gap’s closing. OpenAI’s struggling to build anything better than GPT-4. Meanwhile Google’s Gemini 3 got amazing reviews. DeepSeek matched OpenAI’s performance for pocket change.Young workers don’t know what’s coming. Job market’s changing under their feet. Older workers wondering if skills they spent decades building matter anymore. Investors threw absolutely insane amounts of money at AI companies, data centers, infrastructure. Just hoping it pays off.Jamie Dimon: “It’s going to affect every application, every job, every customer interface.” Sundar Pichai called AI “more profound than fire or electricity.” Sam Altman warned “someone is going to lose a phenomenal amount of money.”Whether OpenAI becomes a warning story about bubbles or somehow survives to dominate—we’ll only know looking back. What’s clear now is ChatGPT’s thousand days changed everything. Trillion-dollar question is what happens next thousand.