PC Gamer Tracks AI Leaders' Shifting Claims on AGI and Industry Buzzwords

PC Gamer has launched a running tally documenting statements from AI industry leaders, tracking what the publication describes as contradictory claims about artificial general intelligence (AGI) and evolving industry terminology.

The article notes a pattern where AI company executives emphasize both existential risks and transformative potential. OpenAI's official description frames AGI as technology that could "elevate humanity by increasing abundance, turbocharging the global economy, and aiding in the discovery of new scientific knowledge," while also acknowledging "serious risk of misuse, drastic accidents, and societal disruption."

PC Gamer documents multiple AGI timeline predictions from prominent figures. In April 2024, Elon Musk stated that if AGI is defined as smarter than the smartest human, "it's probably next year, within two years." By November 2024, Sam Altman said AGI was coming in 2025. In December 2025, Musk claimed xAI could achieve AGI within a few years, possibly by 2026. Jensen Huang told investors in March 2024 that if AGI means software performing 8 percent better than most people on specific tests, "I believe we will get there within 5 years."

The publication also tracks emerging buzzwords in AI discourse. "Pacing the Frontier" refers to slowing AI advancement due to safety concerns. PC Gamer notes the term gained traction in July 2026 following a statement from 1,386 employees at frontier AI companies requesting government support to "deliberately pace the frontier of automated AI development."

"Recursive self-improvement" describes AI systems developing improved versions of themselves with minimal human involvement. Anthropic introduced this concept prominently in June 2026 through a post titled "When AI builds itself," stating that the company is "delegating a growing share of AI development to AI systems themselves." Anthropic cautioned that recursive self-improvement "is not inevitable" but "could come sooner than most institutions are prepared for."

PC Gamer's compilation spans from October 2023, when Google's Blaise Agüera y Arcas and Stanford's Peter Norvig declared AGI already existed, through September 2026. The publication plans to update the log whenever particularly notable claims emerge from AI industry figures.