Artificial intelligence was omnipresent in 2025: The World Economic Forum identified it as a top emerging technology, while firms like McKinsey and Deloitte focused on AI to the exclusion of other technologies in their comprehensive analyses of tech trends.

Artificial general intelligence, AI that could hypothetically match or surpass human reasoning, was the ostensible finish line for what has been called the "AI arms race," the fevered, costly frenzy that gripped Silicon Valley.

According to The Verge, however, the biggest names in AI — firms like OpenAI, Google, Amazon, Microsoft, and Meta — appear to be retreating from both the AGI label and the poorly defined concept it embodied.

In March, the Association for the Advancement of Artificial Intelligence presented a report on the future of AI research, including the prospect of AGI. Among AI researchers polled, 76% expressed skepticism that scaling current approaches could "yield AGI."

"AGI is not a formally defined concept, nor is there any agreed test for its achievement," the report noted, a sentiment reflected by the experts who participated in the research. 

Google Gemini lead Jeff Dean cast doubt on the possibility of even arriving at a consensus definition of AGI when he appeared on an August episode of The Moonshot Podcast.

"I think the reason I tend to steer away from AGI conversations is lots of people have very different definitions of it, and the difficulty of the problem varies by, like, factors of a trillion," Dean hedged.

Just one month prior, Mark Zuckerberg touted Meta's billion-dollar investments in AI, a move insiders claimed was driven by fear of falling behind rather than by any promise of a return on investment. 

Less than two months passed between Meta's frantic spending on the putative pursuit of AGI and Dean's demurral when pressed on the subject, which was something of a broader theme after tech firms bet everything on a future vaguely predicated on AI.

As tech titans began wavering on AGI over the summer, a separate AI-related problem was emerging: This quest to conquer the field and establish yet-undefined AI dominance led to the widespread construction of data centers.

In August, NPR reported that energy prices had risen "more than twice as fast as inflation," and across the country, electric bills skyrocketed. 

Which is your biggest concern regarding AI technology? Energy consumption ⚡ Job displacement 🤖 Data privacy and security risks ⚠️ Humans losing control of it 😱 Click your choice to see results and speak your mind.

Data centers are resource-hungry. In addition to water, the energy they consume has been identified as the primary cause of cost spikes, and America's already strained power grid continued buckling under increased demand.

While credible concerns about an "AI bubble" escalated and Americans unwillingly subsidized ill-defined AI innovations, Silicon Valley set about moving its AGI goalposts.

Meta leaned in to what it called personal superintelligence, while Microsoft adopted Humanist Superintelligence, or HSI, a concept that The Verge described as "essentially Zuckerberg's manifesto in a different font."

Get TCD's free newsletters for easy tips to save more, waste less, and make smarter choices — and earn up to $5,000 toward clean upgrades in TCD's exclusive Rewards Club.