February 25, 2025 News
Anthropic's Claude 3.7 Sonnet Cost Only Tens of Millions to Train
According to information reportedly provided by Anthropic to Wharton professor Ethan Mollick, their latest flagship AI model Claude 3.7 Sonnet cost only "a few tens of millions of dollars" to train using less than 10^26 FLOPs. This relatively modest training cost for a state-of-the-art model demonstrates the declining expenses of developing cutting-edge AI systems compared to earlier generations that cost $100-200 million.
Skynet Chance (+0.08%): The dramatic reduction in training costs for state-of-the-art AI models enables more organizations to develop advanced AI systems with less oversight, potentially increasing proliferation risks and reducing the friction that might otherwise slow deployment of increasingly powerful systems.
Skynet Date (-4 days): The steep decline in training costs for frontier models (compared to $100-200M for earlier models) significantly accelerates the pace at which increasingly capable AI systems can be developed and deployed, potentially compressing timelines for the emergence of systems with concerning capabilities.
AGI Progress (+0.06%): While not revealing new capabilities, the substantial reduction in training costs indicates a significant optimization in model training efficiency that enables more rapid iteration and scaling, accelerating progress on the path to AGI.
AGI Date (-4 days): The dramatic decrease in training costs suggests that economic barriers to developing sophisticated AI systems are falling faster than expected, potentially bringing forward AGI timelines as experimentation and scaling become more accessible to a wider range of actors.
OpenAI Delays API Release of Deep Research Model Due to Persuasion Concerns
OpenAI has decided not to release its deep research model to its developer API while it reconsiders its approach to assessing AI persuasion risks. The model, an optimized version of OpenAI's o3 reasoning model, demonstrated superior persuasive capabilities compared to the company's other available models in internal testing, raising concerns about potential misuse despite its high computing costs.
Skynet Chance (-0.1%): OpenAI's cautious approach to releasing a model with enhanced persuasive capabilities demonstrates a commitment to responsible AI development and risk assessment, reducing chances of deploying potentially harmful systems without adequate safeguards.
Skynet Date (+2 days): The decision to delay API release while conducting more thorough safety evaluations introduces additional friction in the deployment pipeline for advanced AI systems, potentially extending timelines for widespread access to increasingly powerful models.
AGI Progress (+0.03%): The development of a model with enhanced persuasive capabilities demonstrates progress in creating AI systems with more sophisticated social influence abilities, a component of human-like intelligence, though the article doesn't detail technical breakthroughs.
AGI Date (+1 days): While the underlying technical development continues, the introduction of additional safety evaluations and slower deployment approach may modestly decelerate the timeline toward AGI by establishing precedents for more cautious release processes.
DeepSeek Resumes API Services After Capacity-Driven Pause
Chinese AI startup DeepSeek has reopened access to its API after a three-week pause caused by capacity constraints. The company's openly available R1 reasoning model has gained recognition for matching or exceeding the performance of OpenAI's top models, prompting competitive responses from both OpenAI and domestic rivals like Alibaba.
Skynet Chance (+0.04%): The growing competitive landscape in high-performance reasoning models indicates AI capabilities are advancing rapidly across multiple organizations, reducing centralized control and potentially increasing the risk of safety corners being cut to maintain market position.
Skynet Date (-2 days): The capacity constraints DeepSeek faced and subsequent reopening suggests high demand for advanced reasoning models, accelerating the timeline for widespread deployment of increasingly capable AI systems that may eventually lead to control issues.
AGI Progress (+0.06%): DeepSeek's R1 reasoning model matching or exceeding OpenAI's top models represents significant progress in the broader availability of advanced AI capabilities, particularly as these models approach levels of reasoning necessary for AGI components.
AGI Date (-3 days): The competitive pressure between DeepSeek, OpenAI, and Alibaba is likely to accelerate development timelines, with OpenAI reportedly pulling up product releases and competitors launching new reasoning models in rapid succession.
Microsoft Reduces Data Center Footprint Amid Industry Expansion
Microsoft has canceled leases totaling approximately two data centers' worth of capacity with multiple providers. This contraction stands in contrast to significant industry expansion, including the $500 billion Stargate project by OpenAI, Oracle, and SoftBank, and raises questions about Microsoft's expectations for future AI demand.
Skynet Chance (-0.03%): Microsoft's scaling back suggests possible moderation in AI infrastructure growth from at least one major player, potentially reducing risks associated with unchecked expansion, though the effect is limited given massive investment by other entities.
Skynet Date (+1 days): Microsoft's reduction in data center capacity might slightly slow the overall pace of AI infrastructure deployment, potentially extending the timeline before computing resources reach levels capable of supporting highly autonomous systems.
AGI Progress (-0.05%): The cancellation of data center leases by a leading AI investor suggests possible recalibration of expectations about near-term AI progress or demand, potentially indicating less optimism about imminent AGI breakthroughs.
AGI Date (+2 days): Microsoft's apparent reconsideration of infrastructure needs could signal a more extended timeline for AGI development, as computing resources are a crucial factor in scaling AI capabilities toward general intelligence.
Koyeb Integrates Tenstorrent's RISC-V AI Accelerators into Serverless Platform
Cloud platform Koyeb has deployed Tenstorrent's AI accelerators, offering developers access to an alternative to Nvidia's GPUs. This partnership follows Tenstorrent's recent $700 million funding round and represents part of a broader effort to build hardware and software alternatives to Nvidia's dominant AI stack.
Skynet Chance (+0.01%): The diversification of AI hardware and emergence of new accelerator architectures slightly increases risks by expanding the technological surface area for AI development, though the overall impact is moderate as these alternatives still fall within conventional AI development paradigms.
Skynet Date (-1 days): The increased availability of AI accelerators and low-latency cloud infrastructure for AI workloads could marginally accelerate the timeline for deploying advanced AI systems by reducing hardware bottlenecks and democratizing access to specialized computing resources.
AGI Progress (+0.04%): The development of alternative, potentially more accessible AI hardware stacks contributes meaningfully to the technological infrastructure necessary for AGI development, reducing dependency on a single vendor and potentially enabling novel approaches to AI architecture.
AGI Date (-2 days): The combination of high-performance hardware alternatives, significant investment ($700M for Tenstorrent), and serverless deployment options will likely accelerate AGI development by reducing computing constraints and expanding the pool of researchers with access to specialized AI infrastructure.