Michael I. Jordan Critiques AI Doom Rhetoric and Emphasizes Engineering Realities

Michael I. Jordan, a pivotal figure in machine learning and recently described by *Science* as "the most influential computer scientist," is sharply critical of the current narratives surrounding AI threats. During an interview with *Libération*, he labeled the prevailing doom rhetoric as a distortion of reality driven by naïve perspectives among certain commentators.

Jordan, who is a professor emeritus at UC Berkeley and currently heads the "markets and learning" chair at Inria in Paris, expressed his frustration with the sensational discourse about AI, which he believes is detracting from the meaningful evolution of technology. He remarked, "The summer's AI discourse was totally frustrating," underscoring his view that such dialogues often inflate fears that misrepresent AI's capabilities and limitations.

Concerns Over AI Terminology

Jordan has long cautioned against the reckless use of the term "AI." He pointed out that this verbiage, originally too broad, has inflated into buzzwords like "AGI" (Artificial General Intelligence) and now "superintelligence." He stated, "AI wasn't enough, so they started talking about AGI, and now it's 'superintelligence,' which creates public hysteria," elaborating that overhyped terms warp public perception and expectation, particularly among young professionals.

This inflation of terminology serves not only to sensationalize AI but also creates a barrier for new researchers who may lack the technical grounding required to critically analyze these claims. As a result, young employees in the field often come from varied backgrounds such as humanities or philosophy, which may lack practical engineering experience. Jordan noted, "They get together, a bit like in Parisian cafés in the 1920s, to discuss grand science-fiction ideas." This lack of technical expertise, he argues, leads to a romanticized but impractical understanding of AI development.

Theatre of Ethical Resignations

Jordan did not spare criticism for the resignations seen among researchers at AI companies, framing them as "theater." He argued that such actions are often choreographed to create a façade of virtue: “When they resign, they know they’re going to attract attention and feel virtuous. They claim to be making sacrifices but often do so from a position of financial comfort thanks to stock options.” He sees these actions as misguided and says they detract from the more pressing issues in the field.

Superintelligence: A Science Fiction Concept

Jordan strongly denounced the notion of "superintelligence" as a construct rooted in science fiction rather than empirical reality. He explained, "When it becomes the dominant discourse, it is very harmful for technological development and for the mindset of countries." He emphasizes that while it's acceptable to engage with hypothetical scenarios as a thought exercise, relying on such ungrounded concepts can lead to detrimental outcomes.

He was particularly critical of Geoffrey Hinton, a Nobel laureate known for his bold AI claims, saying, "Hinton believes that we can reproduce the brain in silicon," an assertion Jordan deems lacking in credibility. Jordan elaborated that the missing connections in this line of reasoning highlight a disconnect from the practical engineering foundations necessary for realizing accurate AI models.

Evaluating AI Developments

Jordan offered a more favorable view of Yoshua Bengio, who emphasizes the importance of safety in AI. He acknowledged that Bengio addresses real-world AI challenges with practical actions, steering discussions towards actual technological shortcomings rather than dystopian scenarios. However, Jordan remains wary of the overwhelming financial commitments being made in AI through massive data center establishments across the industry, calling them a "trillion-dollar bet" with uncertain returns. "We have absolutely no idea whether it will pay off," he stated, suggesting a more nuanced approach to integrating specialized knowledge within AI systems could yield better outcomes without necessitating vast resources.

In summary, Jordan's perspective reshapes the narrative from existential threats of AI to the pragmatic aspects of developing technology responsibly. He advocates for a shift away from sensationalist claims towards a more engineering-focused discourse, where tangible results and practical implications take precedence over the dramatic potentials of AI.

By fostering a clearer understanding of AI capabilities and engaging in rational, evidence-based discourse, the field can move towards a more productive future. It is essential that experts prioritize developing technology that serves real-world users, rather than indulging in speculative and overly cataclysmic scenarios.