AI leaders' bold vision for the future The real hallucination

In "AI’s Real Hallucination Problem" published by The Atlantic, the article delves into the ambitious and often audacious attitudes of tech executives in the artificial intelligence industry. The piece highlights the journey from the release of OpenAI's DALL-E 2 to the widespread use of generative AI tools like ChatGPT, noting the significant hype and resulting technological revolution.

Despite the enthusiasm, the article underscores the drawbacks, including the flooding of social networks with bizarre AI-generated images and the inundation of search engines with low-quality chatbot-generated content. The unauthorized training of AI on copyrighted material has sparked backlash, particularly as these technologies compete with human creativity for jobs and attention.

The article criticizes the tech industry's manifest-destiny attitude, where companies seem determined to reshape society with little regard for the repercussions. AI leaders, emboldened by their success and investor support, have shown a paternalistic approach, envisioning a future dictated by their technologies. This audacity is evident in the rhetoric of AI executives who propose radical changes without seeking broader consent. The piece references OpenAI CEO Sam Altman and Thrive Global CEO Arianna Huffington, who liken their theoretical AI health coach to the New Deal, revealing a grandiose vision disconnected from reality. OpenAI's CTO, Mira Murati, candidly suggests that some jobs, particularly creative and repetitive ones, might be better off disappearing, reflecting a cold, unsettling view of labour.

The article also highlights the environmental costs of AI, citing Google's increased emissions due to the energy demands of generative AI, which contradicts its net-zero climate commitments. Furthermore, companies like Perplexity face criticism for using third-party tools to harvest information, ignoring established online protocols. The rhetoric from AI leaders, the article argues, is particularly exasperating, as they often seem detached from the real-world implications of their technologies. This attitude extends to the belief that AI will create new jobs, though the language used often devalues existing ones, implying a readiness to sacrifice the "median human" worker for technological progress.

Leopold Aschenbrenner's 165-page essay series is cited, warning about the concentration of power in a small group of AI developers. Aschenbrenner, a former OpenAI employee, expresses concern over the potential for geopolitical and cultural upheaval driven by AGI. His treatise depicts a civilizational struggle that feels alien to the everyday realities of AI observed by the public. The article concludes by emphasizing the danger of allowing a small, audacious group of tech executives to dictate the future, as their ambitions could lead to significant societal disruptions. Ultimately, the real hallucination problem, the article suggests, lies not in the machines but in the audacity of those who build them.