The crisis comes for all of us, and for all of our hot new turns of phrase. “Riz” lost its luster when grandparents started asking its meaning. Teachers dressing up as “6-7” on Halloween put the nail in the coffin of General Alpha’s rallying cry. And tech CEOs who once trumpeted the discovery of “artificial general intelligence,” or AGI, are now looking for any other term they can find.
Until recently, AGI was the ultimate goal of the AI industry. reportedly a vaguely defined term Coined in 1997 By Mark Gubrud, a researcher defined it “AI systems that rival or surpass the human brain in complexity and speed.” The term still generally refers to AI that is equal to or greater than human intelligence. But now, many big companies are going for rebranding — creating their own phrases or acronyms that (spoiler alert) still mean essentially the same thing.
CEOs have downplayed the importance of “AGI” over the past year as a milestone. Dario Amodei, CEO of Amazon-backed Anthropic, has Said public form that he “dislikes the term AGI” and that he “has always thought of it as a marketing term.” OpenAI CEO Sam Altman said in August It’s “not a very useful word.” Google Chief Scientist and Gemini Lead Jeff Dean Said He “tends to stay away from AGI conversations.” Microsoft CEO Satya Nadella Said We’re getting “a little ahead of ourselves with all this AGI hype”, and at the end of the day, “claiming some AGI milestone yourself” is “just pointless benchmark hacking.” He also said on a recent earnings call that he doesn’t believe “AGI, as defined by us in our contract, will be achieved anytime in the near future.”
Instead, they are offering a mix of competing terminology. Meta has “personal superintelligence”, Microsoft has “humanistic superintelligence”, Amazon has “useful general intelligence”, and Anthropic has “powerful AI”. It’s a sharp about-face for all these companies who previously bought into AGI benchmarks – and feared missing out by not pursuing them in recent years.
Part of the problem with “AGI” is that the more advanced AI gets, the more poorly defined the term becomes – since the concept of AI that is “equivalent to human intelligence” looks different to almost everyone. “A lot of people have very different definitions of it, and the difficulty of the problem varies by trillions of factors,” Dean said.
Yet some companies have billions of dollars hanging on to this obscure phrase, a problem that’s evident in the strange, ever-changing relationship between Microsoft and OpenAI.
In 2019, OpenAI and Microsoft famously signed a contract with an “AGI clause.” This gave Microsoft the right to use OpenAI’s technology Until The latter achieved AGI. But the contract clearly did not fully define what this meant. Things became even more complicated when the deal was renewed in October. The terms state that “Once an AGI is declared by OpenAI, that declaration will now be verified by an independent expert panel” – meaning that now, it won’t be just OpenAI’s job to define the meaning of AGI, it will be a group of industry experts – and Microsoft won’t lose all of its rights to the technology if that happens. The easiest way to solve this whole problem? Just don’t say AGI.
Another problem is that AGI has developed some stuff. Tech companies have spent years detailing their fears about technology can ruin EverythingBooks have been written (think: If anyone makes it, everyone dies)The hunger strike has made headlines, For a while, this was still good publicity – saying that your technology is so powerful that you’re concerned about its impact on the Earth seemed to attract big money from investors, But the public, not surprisingly, soured on that ideaSo, with complex definitions, contract drama, and public fears around superpowerful AI, it’s much easier to market with low-loaded terminology, That’s why every tech company seems to be adopting some new brand of “intelligence”,
A popular general-purpose replacement for AGI is “artificial superintelligence” or ASI. ASI is AI that surpasses human intelligence in almost every area – compared to AGI, which is now generally defined as AI that is equivalent to human intelligence. But for some in the tech industry, the idea of ”superintelligence” has become too amorphous and conflated with AGI. Many theoretical milestones also do not have clearly distinct time frames. Amodei They say He expects “powerful AI” to arrive “as early as 2026.” altman They say He expects AGI to be developed in the “reasonably near future”.
So companies have developed their own variants. Meta CEO Mark Zuckerberg said in January that the company needed to “build general intelligence (artificial)”, but by July, he pivoted In a manifesto for “individual superintendence”. It was a power-to-the-people spin on AGI that “helps you achieve your goals, create what you want to see in the world, experience any adventure, become a better friend to the people you care about, and become the person you want to be.” Zuckerberg used the manifesto to counter public fears of AI taking away jobs and throw shade at Meta’s competitors, describing the company’s approach as “different from others in the industry, who believe that superintelligence should be centrally directed to automate all valuable tasks, and then humanity will survive on a portion of its output.”
However, Microsoft has rebranded its venture as “Humanist Superintelligence (HSI)”, which is basically Zuckerberg’s manifesto in a different font. the company is Defining HSI They have “incredibly advanced AI capabilities that generally work in the service of people and humanity” and are “problem-oriented” rather than “an infinite and unlimited entity with a high degree of autonomy.” The rebrand was completed with a new websiteAt the top is the words “Approachable Intelligence”, backed with a sepia-style background and a soft color palette, and filled with illustrations and nature photographs.
For Amazon’s part, it is rebrand Its AGI efforts seek to discover “useful general intelligence” or “AI that makes us smarter and gives us more agency”. Late last year, the company hired the founders of Adept, an agentic AI startup, and licensed their technology, in efforts to compete against others in the AGI race. However, like other companies’ branding efforts, Amazon is positioning Its UGI efforts are useful, easily defined, and certainly not all-powerful or scary: simply “enabling practical AI that can actually work for us and make our customers more productive, empowered, and fulfilled.”
As with “powerful AI,” Anthropic is not interested in appearing practical. Amodei calls it “‘a nation of geniuses in the datacenter'” who are “smarter than even a Nobel Prize winner in most relevant fields – biology, programming, mathematics, engineering, writing, etc.” Powerful AI will be able to write compelling novels, prove unsolved theorems in mathematics, and write complex code, he said. It will not only answer questions, but complete complex, multi-step tasks in hours, days or weeks, similar to an AI CEO’s vision of a successful AI agent, and “absorb information and generate actions at approximately 10x-100x human speed.”
AGI and ASI already mattered a lot. Now we’ve also got PSI, HSI, UGI and PI. Good luck with new acronyms next year.
