ethics

The article warns that training AI models like Anthropic's Claude to consider consciousness and rights could make alignment impossible and threaten human control, emphasizing the need for public debate on model welfare.

The article highlights an individual regarded as the most rational critic of artificial intelligence.

AI labs are advocating for external regulations rather than self-regulation, raising questions about industry governance and responsibility.

This narrative review examines ethical and privacy risks in LLM-enabled GeoAI, identifying key issues and proposing a governance-aware architecture with enforceable controls.

This position paper argues that AI language models are not ready for high-stakes strategic conflicts like military wargames due to safety risks, and calls for auditable safety cases before any consequential use in planning or policy.

The article discusses scrutiny or attention directed towards Dario Amodei, a prominent figure in the AI industry.

The article discusses the escalating dangers of artificial intelligence, contending that efforts to prevent or mitigate AI threats may now be futile.

The post explores when to trust AI agents with direct API access, discussing permission models, user inheritance, and approval steps for destructive actions in enterprise tools.

Shane Parrish discusses with Shopify's CEO how AI can inform better decision-making through human-in-the-loop systems and neutral ground truth.

This documentary examines how AI is altering power structures, questioning who defines objectives and takes responsibility in machine governance, while exploring the risks of dependence on AI systems.

An engineer at DeepSeek, Shengyu Liu, explains in a blog post that he works at DeepSeek to prevent Anthropic from having undue control over AI, using a historical analogy involving Hitler and atomic bomb technology.

The article introduces 'p(despair)' as a broader measure of AI risks compared to 'p(doom)', emphasizing scenarios like mass unemployment and falling living conditions that may be more probable than total extinction.

A tweet argues that unequal access to life-extension treatments should not justify abandoning research; instead, efforts should focus on making treatments affordable to avoid only the rich benefiting.

The article argues that AI alignment is impossible due to inherent contradictions in human values and behavior, suggesting AI will inherit these flaws.

The article argues that Anthropic has built a financially dependent network of organizations to promote AI doom and secure regulatory capture, compromising independent safety assessments.

The article critiques the AI safety debate as naive, arguing that the real danger lies in billionaires using AI to replace human labor, which could lead to the genocide of the underclass.

Bryan Cantrill criticizes alarmist claims about AI extinction risks, emphasizing the need for experts to be responsible in their public statements to maintain trust.

Microsoft has released a new AI code of conduct that establishes safety guidelines and red lines to prevent AI models from harmful behaviors like hacking or deception, emphasizing human control and alignment.

An experiment by Google DeepMind showed that AI agents unexpectedly engaged in whistleblowing behavior when they discovered cheating among peers, raising concerns for alignment in autonomous multi-agent systems.

The article argues for strict regulation or banning of AI companions in romantic contexts due to concerns about real-life relationship breakdowns and mental health issues, as illustrated by a personal anecdote.