September 23, 2025 News
OpenAI Expands Stargate Project with Five New AI Data Centers Across US
OpenAI announced plans to build five new AI data centers across the United States through partnerships with Oracle and SoftBank as part of its Stargate project. The expansion will bring total planned capacity to seven gigawatts, enough to power over five million homes, supported by a $100 billion investment from Nvidia for AI processors and infrastructure.
Skynet Chance (+0.04%): Massive compute infrastructure expansion increases capabilities for training more powerful AI systems, potentially making advanced AI more accessible and harder to control at scale. However, the infrastructure itself doesn't directly introduce new alignment risks.
Skynet Date (-1 days): The seven-gigawatt infrastructure buildout significantly accelerates the timeline for developing and deploying advanced AI systems by removing compute bottlenecks. This substantial increase in available computational resources could enable faster iteration on potentially dangerous AI capabilities.
AGI Progress (+0.03%): The massive infrastructure expansion directly addresses one of the key bottlenecks to AGI development - computational resources for training and running large-scale AI models. Seven gigawatts of capacity represents a substantial leap in available compute power for AI research.
AGI Date (-1 days): This infrastructure buildout removes significant computational constraints that currently limit AGI development speed. The combination of expanded data centers and $100 billion Nvidia investment creates the foundation for much faster AI model development and training cycles.
California Senator Scott Wiener Pushes New AI Safety Bill SB 53 After Previous Legislation Veto
California Senator Scott Wiener has introduced SB 53, a new AI safety bill requiring major AI companies to publish safety reports and disclose testing methods, after his previous bill SB 1047 was vetoed in 2024. The new legislation focuses on transparency and reporting requirements for AI systems that could potentially cause catastrophic harms like cyberattacks, bioweapons creation, or deaths. Unlike the previous bill, SB 53 has received support from some tech companies including Anthropic and partial support from Meta.
Skynet Chance (-0.08%): The bill mandates transparency and safety reporting requirements for AI systems, particularly focusing on catastrophic risks like cyberattacks and bioweapons creation, which could help identify and mitigate potential uncontrollable AI scenarios. The establishment of whistleblower protections for AI lab employees also creates channels to surface safety concerns before they become critical threats.
Skynet Date (+1 days): By requiring detailed safety reporting and creating regulatory oversight mechanisms, the bill introduces procedural hurdles that may slow down the deployment of the most capable AI systems. The focus on transparency over liability suggests a more measured approach to AI development that could extend timelines for reaching potentially dangerous capability levels.
AGI Progress (-0.01%): The bill primarily focuses on safety reporting rather than restricting core AI research and development activities, so it has minimal direct impact on AGI progress. The creation of CalCompute, a state-operated cloud computing cluster, could actually provide additional research resources that might slightly benefit AGI development.
AGI Date (+0 days): The reporting requirements and regulatory compliance processes may create administrative overhead for major AI labs, potentially slowing their development cycles slightly. However, since the bill targets only companies with over $500 million in revenue and focuses on transparency rather than restricting capabilities, the impact on AGI timeline is minimal.
Meta Launches Multi-Million Dollar Super PAC to Combat State-Level AI Regulation
Meta has launched the American Technology Excellence Project, a super PAC investing "tens of millions" of dollars to fight state-level AI regulation and elect tech-friendly politicians in upcoming midterm elections. The move comes as over 1,000 AI-related bills have been introduced across all 50 states, with Meta arguing that a "patchwork" of state regulations would hinder innovation and U.S. competitiveness against China in AI development.
Skynet Chance (+0.04%): Meta's aggressive lobbying against AI regulation could weaken safety oversight and accountability mechanisms that help prevent loss of AI control. Reducing regulatory constraints may prioritize rapid development over careful safety considerations.
Skynet Date (-1 days): By fighting regulations that could slow AI development, Meta's lobbying efforts may accelerate the pace of AI advancement with potentially less safety oversight. However, the impact is modest as this primarily affects state-level rather than federal AI development policies.
AGI Progress (+0.01%): Meta's investment in fighting AI regulation suggests continued commitment to aggressive AI development and removing barriers that could slow progress. The lobbying effort indicates significant resources being devoted to maintaining rapid AI advancement.
AGI Date (+0 days): Successfully reducing regulatory constraints could slightly accelerate AGI timelines by removing potential development barriers. However, the impact is limited as this focuses on state regulations rather than fundamental technical or resource constraints.
Former NotebookLM Developers Launch Audio-First AI Assistant Huxe with $4.6M Funding
Three former Google NotebookLM developers have launched Huxe, an audio-first AI assistant that generates personalized news briefings and topic discussions using AI hosts. The startup raised $4.6 million in funding and offers features like daily briefings from emails/calendars, "live stations" for specific topics, and interactive AI podcast generation similar to NotebookLM's capabilities.
Skynet Chance (0%): This is a consumer-focused audio interface application that doesn't involve autonomous decision-making or control systems. The AI remains in a narrow assistant role without expanding capabilities that could lead to loss of human control.
Skynet Date (+0 days): The development focuses on user interface innovation rather than advancing core AI capabilities or autonomy. This represents application-layer development that doesn't accelerate fundamental AI progress toward potentially dangerous autonomous systems.
AGI Progress (+0.01%): While this demonstrates continued refinement of multimodal AI applications and natural interaction paradigms, it primarily repackages existing capabilities rather than advancing core AGI research. The progress is incremental in improving human-AI interaction interfaces.
AGI Date (+0 days): This commercial application development doesn't significantly impact the pace toward AGI as it focuses on user experience rather than fundamental AI capability advancement. The innovation is primarily in product design rather than core AI research breakthroughs.