A recent statement from an Anthropic researcher suggesting a significant probability of artificial intelligence leading to human extinction has drawn sharp criticism, with some characterizing the claims as "bullshit" and a "marketing stunt." Evan Hubinger, Anthropic's "Alignment Science Lead," publicly stated his belief that there is "greater than 10%" chance AI could eliminate humanity within the next decade. He further indicated that Anthropic "do not yet have a plan to solve alignment for superintelligence and are not clearly on track to."
This assertion, shared widely on social media, has been met with skepticism. Critics argue that such pronouncements from companies developing AI serve as a deliberate tactic to inflate the perceived power and importance of their products. The underlying technology, described as sophisticated predictive text engines rather than true artificial intelligence, is seen by detractors as far from capable of autonomous or malevolent action on a global scale.
Concerns are also raised about the financial motivations behind these dramatic predictions. With substantial investments in AI development, companies may be resorting to doomsday narratives to secure continued funding from investors. This alleged pattern of publicizing potential future dangers, often shortly after new model releases, is viewed as a method to maintain hype and delay the inevitable economic consequences if the technology fails to meet inflated expectations.
Industry observers point out the lack of concrete examples to support claims of AI's rapidly advancing capabilities or its newly emergent dangers, particularly in areas like computer security. While acknowledging that AI can pose risks on an individual level through misinformation or encouraging harmful behavior, the notion of AI plotting humanity's downfall is considered unfounded. Instead, these exaggerated claims are seen as a way to mask the stagnation in actual AI intelligence and the unsustainable business model of generative AI companies.
The narrative pushed by AI developers, where their creations are presented as both incredibly powerful and inherently dangerous, is scrutinized for its strategic benefit to the companies. The lack of demonstrable, coherent achievements alongside dire warnings about future potential fuels suspicion that these are not genuine concerns but rather calculated public relations maneuvers to sustain investor interest and the ongoing "grift" of generative AI. The responsible dissemination of such alarming claims by news outlets is also questioned, with a call for greater critical evaluation of pronouncements that lack verifiable evidence.