AI's Real Threats Are Already Here | Tech News

0:00On September 11, Timnick Gebru is raising some serious questions about the current AI hype. She believes the focus on safety and alignment and the talk of AI causing the end of civilization is a distraction. Gebru points out that companies are heavily investing in specific problems like math and chess to make it seem like they've cracked intelligence. When in reality, these are often curated successes designed for public consumption. She's critical of how breakthroughs are presented, suggesting that the industry often skips the rigorous vetting processes that scientific communities normally use. This allows companies to frame their achievements in the most dramatic way possible, pushing a narrative of superintelligence without proper scrutiny of the claims. Gebru argues that the real existential threats are already here. She highlights AI being used in autonomous weapons and warfare, the potential for AI to worsen the climate crisis and how companies are using AI to. Justify layoffs. She likens the industry's approach to occult

1:02with some true believers pushing a singularity narrative that's been promised for decades. The idea of AI causing extinction, while dramatic, often relies on scenarios that are hard to envision. Gebru notes that these often involve an AI optimizing for a task in a way that leads to unintended catastrophic outcomes, like eliminating humanity. She finds the scenario speculative, especially when compared to the tangible harms already present. Ultimately, Gebru wants us to focus on the immediate real world impacts of AI. She's pushing back against the sensationalism, urging a more grounded discussion about the ethical and societal consequences of the technology as it's. Being developed and deployed right now, from the industry powered by AI, this is tech news today.