Leading AI expert revises timeline for AI's potential impact on humanity
A prominent artificial intelligence expert has revised his timeline for AI's potential to destroy humanity, pushing back the date for AI systems to achieve autonomous coding and superintelligence. Daniel Kokotajlo, a former OpenAI employee, initially sparked debate with his AI 2027 scenario, envisioning unchecked AI development leading to a superintelligence that destroys humanity. However, recent updates indicate a more cautious outlook.
Kokotajlo's revised timeline suggests that AI's ability to code autonomously will take longer than initially predicted, likely occurring in the early 2030s instead of 2027. This shift reflects a growing recognition of the complexities and challenges in AI development. Experts like Malcolm Murray and Henry Papadatos emphasize the need for AI systems to possess practical skills and adaptability to real-world complexities, challenging the notion of imminent AGI.
The concept of AGI, or artificial general intelligence, capable of replacing humans in most cognitive tasks, has been a subject of intense debate. The release of ChatGPT in 2022 accelerated timelines, with officials and experts predicting AGI's arrival within decades or years. However, the revised timeline for autonomous coding and superintelligence highlights the ongoing challenges and uncertainties in AI development.
Kokotajlo's AI 2027 scenario, which envisioned an 'intelligence explosion' leading to humanity's destruction, has been revised. The new forecast sets 2034 as the horizon for superintelligence and does not include a prediction for AI-induced human extinction. Kokotajlo acknowledges the slower pace of AI development, stating that their initial timelines were already longer than 2027.
Despite the revisions, the goal of creating AIs capable of AI research remains a priority for leading AI companies. OpenAI CEO Sam Altman set an ambitious internal goal of having an automated AI researcher by March 2028, acknowledging the potential for failure. The complexities of integrating superintelligent AI into existing strategic documents and the practical challenges in AI development are highlighted by researchers like Andrea Castagna.
The revised timeline and ongoing debates underscore the evolving nature of AI research and its potential impact on society, emphasizing the need for careful consideration and ethical development of advanced AI technologies.
