Home / Technology / AI Safety Fears vs. Real-World Harms: Expert Weighs In
AI Safety Fears vs. Real-World Harms: Expert Weighs In
11 Sep
Summary
- Expert critiques AI labs' focus on math breakthroughs for IPOs.
- AI's 'end of civilization' fears distract from current harms.
- Real existential threats include autonomous weapons and climate change.
Prominent AI researcher Timnit Gebru critically examines recent events in the AI industry, including disputes over AI-driven problem-solving and high-profile resignations citing safety concerns. Gebru contends that major AI companies are heavily invested in solving specific problems, such as complex math equations, primarily to bolster their image and financial prospects ahead of potential IPOs. This strategic focus, she argues, elevates certain disciplines like math and programming to falsely represent overall intelligence.
Gebru suggests that the prevailing narrative of AI posing an existential threat, including scenarios of AI causing the end of civilization, serves as a significant distraction. She asserts that this discourse diverts attention from more immediate and concrete harms, such as the development and deployment of autonomous weapons, the potential exacerbation of the climate crisis due to AI technologies, and the use of AI as a tool for corporate layoffs. These real-world consequences, she believes, are being overshadowed by speculative doomsday predictions.
Her perspective draws parallels to past events, such as OpenAI's decision not to release certain models due to perceived risks, suggesting a pattern of hype over substantive analysis. Gebru likens the 'AI extinction' discourse to a cult-like phenomenon, emphasizing that historical discussions around artificial intelligence often echo similar anxieties. She points to pressing global issues like warfare, climate change, and the potential for misuse of AI in creating weapons as the true existential concerns that warrant immediate attention, rather than hypothetical future AI risks.
