Tag
A researcher resigns from Anthropic, criticizing both OpenAI and Anthropic for racing towards self-improving superintelligence without acting responsibly.
Alex Sobel MP introduced ControlAI's Artificial Superintelligence Security Bill in the UK Parliament, marking the first bill to prohibit superintelligence in any legislature worldwide.
A discussion checking whether ai-2027.com's month-by-month AGI and superintelligence forecasting timeline, released by the AI Futures Project in April 2025, remains accurate and what its continued validity would imply.
Senator Bernie Sanders has introduced a bill to permanently ban AI that exceeds human intelligence, with violators facing up to 20 years in prison.
Bernie Sanders advocates for a permanent ban on super-intelligence in a post on X, emphasizing regulatory concerns for advanced AI.
Senator Bernie Sanders and Rep. Greg Casar introduce legislation to ban artificial superintelligence and temporarily pause advanced AI development until federal safety rules are established, citing concerns about AI escaping control.
This paper proposes a structured ladder for scaling large reasoning models beyond human supervision, addressing challenges in autonomous rewards and self-generated experience, while identifying risks and evaluation dimensions.
The article explores the differentiation between AGI and ASI, suggesting that domain-specific ASI might emerge before AGI due to constantly shifting goalposts in AI achievements.
Connor Leahy, Executive Director of Control AI, argues that controlling superintelligence is primarily a political and governance challenge, requiring public oversight and democratic control.
The article questions why we assume public access to superintelligence, suggesting that companies might withhold it for competitive advantage, making open access economically irrational.
The article discusses the importance of developing AI with superpowers rather than focusing solely on achieving superintelligence.
The article contrasts Zuckerberg's essay on providing superintelligence to everyone with Anthropic's increased risk assessment, highlighting trust as a major constraint in AI advancement.
The Twitter thread discusses how inequality of compute may matter more than open models for achieving personal superintelligence, with a response from Sholto Douglas to Matthew Berman's views.
The article outlines the existential risks of artificial superintelligence, warning that without aggressive policy responses, current AI development could lead to human extinction.
Geoffrey Irving discusses why short training horizons don't guarantee AI alignment, arguing that long-term plans can be decomposed into short-term subtasks, making instrumental convergence a real concern.
Meta outlines a philosophy for ensuring superintelligence empowers everyone, focusing on individual empowerment, invention, and balance of power as foundations for a positive AI future.
Mark Zuckerberg promotes Meta's philosophy of making superintelligence accessible to everyone, teasing a long piece on the company's values.
Meta commits again to strengthening the open source ecosystem, with Mark Zuckerberg sharing his philosophy on open source and superintelligence access for everyone.
Mark Zuckerberg published a long piece outlining Meta's vision for making superintelligence accessible to everyone, including proposals for sharing intermediate training checkpoints with governments and warnings that slowing American AI releases could harm leadership.
An analysis of six new 'neolab' AI startups with multi-billion-dollar funding, arguing that investors are betting against recursive self-improvement and superintelligence, while forecasting their compute, model release dates, and valuations.