ChatGPT Changed Everything and Was a Black Swan Event

• ChatGPT became publicly available 6 months ago.
• The chatbot has already upended education, training, hiring, and company share prices.
• It was a black swan event — an unforeseen, rare event that looked inevitable in hindsight.

Bill Gates says its creation is as fundamental as the computer chip, the internet, and the PC. Elon Musk wants to put the brakes on development. Warren Buffett is already comparing it to the atomic bomb.

These hyperbolic statements about artificial intelligence were made in the six months since the startup OpenAI publically released ChatGPT, an AI chatbot underpinned by the GPT-3.5 and GPT-4 large language models.

Similarly to Google’s Bard and image-generators such as DALL-E and Midjourney, ChatGPT can “create” something based on prompts and falls under the new umbrella term generative AI. And since ChatGPT’s release in November, it’s causing as much anxiety as excitement.

At least 49 listed US companies have mentioned ChatGPT in quarterly earnings calls in May 2023 to date, according to an analysis of Seeking Alpha transcripts by Insider. It’s a sign that corporate America is rushing to show off this new, important tech trend.

And it’s no mystery why.

As three researchers recently wrote in a paper published by the National Bureau of Economic Research: “ChatGPT represents an important shock to corporate valuations.”

They calculated that firms with high exposure to AI earned daily excess returns that were 0.4% higher than firms with lower exposure after the release of the chatbot.

A former OpenAI researcher, Paul Christiano, said society had “maybe a 10-20% chance of AI takeover” that would leave many or most “humans dead.”

Here are the signs that ChatGPT was a black swan event — an unforeseen, rare occurrence that had far-reaching impacts and, in hindsight, looked inevitable.

1. AI has come for jobs

The doomy tech prophecy that robots could eventually replace humans is starting to come true.

In March, an early analysis from Goldman Sachs said that generative AI could significantly disrupt the labor market by affecting around 300 million jobs globally. Coding jobs were predicted to be particularly vulnerable as tech layoffs savaged the industry. A September analysis by GitHub, the code repository owned by Microsoft, said that developers using its AI Copilot tool to aid their coding were 55% faster than those without.

Now CEOs are beginning to say the quiet part out loud.

Arvind Krishna, the chief executive of IBM, said in May that the company would slow or suspend hiring for roles that AI might replace and estimated that 7,800 jobs — mainly comprising back-office roles — could be impacted.

Mustafa Suleyman, a cofounder of Google DeepMind, recently called for universal basic income to support what he predicted would be the “serious numbers of losers” of jobs to AI.

As Insider’s Hasan Chowdhury noted, AI may not directly replace jobs across the board because it would enable companies to cut costs by using the technology to make existing workers more efficient and productive.

2. ChatGPT has changed education and training forever

Students and some teachers were among the earliest users of ChatGPT when they realized its potential to produce essays and help with assignments.

Cue plagiarism, false accusations of cheating from educators, and a new atmosphere of distrust between students and professors navigating a post-ChatGPT world.

Beyond school, large language models have the capacity to disrupt training even for seemingly elitist careers, such as medicine. One paper, which had not been peer-reviewed at the time of writing, suggested that ChatGPT could pass all three parts of the US medical-licensing examination. Another experiment said that medical researchers preferred ChatGPT to physicians because it appeared to be more empathetic. Human doctors have concluded that the bot wouldn’t replace them soon but might still have a few things to teach them about their bedside manner. This is a pretty astonishing prospect for an algorithm.

At a corporate level, surging student interest in ChatGPT saw education companies’ stocks get hammered earlier in May.

The student-services company Chegg, which offers online help for students’ homework, saw its stock price almost cut in half after CEO Dan Rosenzweig told investors the company had seen a “significant spike in student interest in ChatGPT” that was “having an impact on our new customer growth.”

“We see this one as a real, transformational change,” he said during its Q1 2023 earnings call, comparing the advancements in AI to other recent technologies such as bitcoin payments. “We’re betting very big on the fact that people are going to need to learn how to use these things.”

Pearson, a fellow digital learning firm, and the virtual language-learning company Duolingo also saw their share prices plummet.

3. ChatGPT single-handedly blew up Big Tech

On the face of it, Big Tech companies have failed to talk about big, ambitious ideas tangibly over the last five years.

Self-driving cars have stalled, internet balloons failed to take off, and even opening brick-and-mortar stores appeared too difficult a challenge.

ChatGPT lit a fire under these behemoths, with Meta, Amazon, and Google all racing desperately to talk about AI publicly after its release and Microsoft’s investment in OpenAI.

In the wake of ChatGPT’s launch, Google reportedly declared a “code-red” and has since been pouring resources into AI to try and keep pace. It has also had to pivot quickly in how it talked about artificial intelligence, moving away from high-minded ethics warnings toward actually putting out products.

One is a revamped version of Google search powered by its new chatbot Bard, a direct response to OpenAI’s ChatGPT. The firm will also integrate the tech across Google Workspace, marking the biggest innovations to its core products in years.

Keen not to be left behind, Meta has also been emphasizing its AI credentials.

In February, the company introduced researchers to LLaMA, a large language model similar to OpenAI’s GPT-4. A month later, CEO Mark Zuckerberg said that the company’s “single largest investment is in advancing AI.”

Analysts have expressed concern over how much Zuckerberg was spending to increase the company’s investment in AI. Meta’s lack of consumer-facing AI products, however, has made it less visible in the space.

Amazon, meanwhile, was working on implementing ChatGPT-like features into a secret new home robot, Insider’s Eugene Kim reported.

4. Nations want to rein in the spread of AI

With the threat of AI displacing jobs and even defaming people, regulators have been closely watching how private companies roll out AI.

The EU has made early strides towards banning uses that pose an “unacceptable risk” while the Biden administration has been meeting with major players in the industry. ChatGPT was briefly banned by Italian regulators.

Regulators may have been spurred on by a growing chorus of warnings about AI’s potential. A UK government official met with Geoffrey Hinton, the former Googler dubbed the “Godfather of AI” who quit Google earlier in 2023 to speak candidly about the risks posed by AI.

Hinton’s warnings were first published in The New York Times. They raised similar concerns to those in an open letter backed by other AI heavyweights, including Yoshua Bengio and Stuart Russell.

Even OpenAI CEO Sam Altman warned US lawmakers in May that personalized disinformation targeted at undecided voters was his personal fear.

“I worry that as the models get better and better, the users can have less and less of their own discriminating thought process,” Altman said.