AI's Ethical Challenges, Societal Concerns, and the Push for Regulation

Key Insight

A significant difference between the AI boom and the dot-com era is the abundance of critics raising profound ethical and societal concerns. Geoff Hinton, often called the 'godfather of AI,' left Google in May to speak freely about his fears that tech giants' competition could lead to uncontrolled systems, calling the projected future 'scary.' The media reflected this apprehension; a May 2023 Time magazine cover starkly read 'THE END OF HUMANITY' against a red background. Large language models themselves fueled public mistrust through rampant factual inaccuracies, exemplified by GPT-4's false claim about Iran's nuclear reactor, ChatGPT's invention of a sexual harassment claim, CNET's AI-generated finance stories being 'riddled with factual inaccuracies,' and lawyers fined 5000 dollars for submitting legal briefs with fake case citations. Concerns about political bias also emerged, with ChatGPT refusing to generate stories depicting drag queens negatively or a Donald Trump voter fraud scenario, while readily creating positive portrayals and an ode to Joe Biden, leading critics like Marc Andreessen to assert a shared ideology among AI companies.

The ethical terrain on which generative AI is built led to a spate of lawsuits and regulatory actions. Artists, musicians, and creatives sued companies like OpenAI and Google for unauthorized scraping of their work to train models. Publishers filed similar lawsuits alleging copyright infringement for content ingestion. Hollywood writers and actors went on strike in May and July, respectively, seeking fairer pay and protection against the unchecked use of AI in scriptwriting and against the use of AI-generated replicas of their likeness or voice without compensation. Internationally, Italy charged OpenAI with unlawful data collection, temporarily banning ChatGPT in March, and later reopened its investigation in 2024, while the EU created a 'ChatGPT Taskforce' to investigate AI companies' data-gathering practices. Environmentally, AI's vast energy demands are a growing concern; a ChatGPT query uses nearly 10 times more electricity than a Google search, with data center energy consumption predicted to more than double by 2030, contributing to Google's 48 percent increase in greenhouse gas emissions from 2019 to 2023.

Concerns about AI's potential existential threat were highlighted by Tristan Harris and Aza Raskin in 'The AI Dilemma,' noting that half of AI researchers believe there's at least a 1-in-10 chance of human extinction due to an inability to control AI. This culminated in a May 2023 open letter from the Center for AI Safety, signed by over 350 individuals including prominent CEOs like Sam Altman and Mustafa Suleyman, Microsoft CTO Kevin Scott, Bill Gates, and AI pioneers Geoff Hinton and Yoshua Bengio, declaring that 'Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war.' However, not all agreed; Yann LeCun deemed these 'prophecies of doom' overblown, and Reid Hoffman chose not to sign, arguing that unlike nuclear war or pandemics, AI offers immense positive consequences and that delays in safe implementation (e.g., AI tutors, doctors) incur a 'huge cost in human suffering.'