Import AI 468：23条RSI政策建议、PostTrainBench+及AI竞赛中的信任与透明度
IFP智库发布23条应对AI研发自动化风险的政策建议，分为透明度、监管能力、风险管理、验证技术、韧性、美国领先优势、国际合作7类。核心比喻是：当前AI发展像只有油门没有刹车的车。这类政策汇总对关注AI治理的读者有参考价值，但缺乏一手信息或独特观点。
评论
点赞
收藏自维持AI蠕虫：用被入侵GPU跑大模型自主攻击
多伦多大学、剑桥大学等机构的研究者构建了一种自维持AI蠕虫原型，利用被入侵机器的GPU运行开源大语言模型，自主推理并生成针对每个目标的定制化攻击策略。该蠕虫无需依赖厂商API，仅靠单张A100 80GB显存即可运行，证明自主AI网络威胁已从理论变为现实。
Import AI 466: The bitter lesson for robotics, AIs complete week-long programming tasks; and OpenAI's accidental AI hacker
Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv, cappuccinos, and feedback from readers. If you’d like to support this, please subscribe. Subscribe now Epoch and METR rel...
英国AI安全研究所：开源与闭源模型网络安全能力差距正在缩小
英国AI安全研究所（AISI）首次公开评估显示，GLM-5.2和DeepSeek V4-Pro等开源模型在网络安全能力上已接近闭源前沿模型，差距从2025年的6-10个月缩短至4-7个月。但在长周期网络攻防任务上，开源模型仍落后约7个月。AISI计划待Kimi K3开源后加入测试。
Import AI 464: Fable writes GPU kernels; AI automation; and analog computation
Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv, cappuccinos, and feedback from readers. If you’d like to support this, please subscribe. Subscribe now Fable writes a dec...
Import AI 463: Self-improving robots; a 10k Chinese GPU cluster; and an elegiac essay for the human era
Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv, cappuccinos, and feedback from readers. If you’d like to support this, please subscribe. Subscribe now NVIDIA sets up a c...
Import AI 462: Superpersuasion; self-sustaining AI; paths to ASI
Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv, cappuccinos, and feedback from readers. If you’d like to support this, please subscribe. Subscribe now AI can decisively ...
Import AI 461: "Alignment is not on track"; FrontierCode; and synthetic research interns
Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv, cappuccinos, and feedback from readers. If you’d like to support this, please subscribe. Subscribe now AI researchers lau...
Import AI 460: Reward hacking society, RSI data from Anthropic; and RL-based quadcopter racing
Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv, cappuccinos, and feedback from readers. If you’d like to support this, please subscribe. Subscribe now Society can be rew...
Import AI 459: AI oversight is difficult; scaling laws for protein folding models; and pricing the extinction risk of AI systems
Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv, cappuccinos, and feedback from readers. If you’d like to support this, please subscribe. Subscribe now The AI economy in ...
Import AI 458: Reckoning with the future; and a singularity story
Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv, cappuccinos, and feedback from readers. If you’d like to support this, please subscribe. This issue consists of a lengthy...
Import AI 457: AI stuxnet; cursed Muon optimizer; and positive alignment
Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv, cappuccinos, and feedback from readers. If you’d like to support this, please subscribe. Subscribe now Stuxnet before Stu...
Import AI 456: RSI and economic growth; radical optionality for AI regulation; and a neural computer
Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv, cappuccinos, and feedback from readers. If you’d like to support this, please subscribe. Subscribe now Regulate? Don’t re...
Import AI 454: Automating alignment research; safety study of a Chinese model; HiFloat4
Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv and feedback from readers. If you’d like to support this, please subscribe. Subscribe now Huawei’s HiFloat4 training forma...
Import AI 453: Breaking AI agents; MirrorCode; and ten views on gradual disempowerment
Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv and feedback from readers. If you’d like to support this, please subscribe. A shorter issue than usual as I was attending ...
Import AI 452: Scaling laws for cyberwar; rising tides of AI automation; and a puzzle over gDP forecasting
Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv and feedback from readers. If you’d like to support this, please subscribe. Subscribe now Uh oh, there’s a scaling war for...
Import AI 451: Political superintelligence; Google's society of minds, and a robot drummer
Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv and feedback from readers. If you’d like to support this, please subscribe. Subscribe now AI might let us build “political...
Import AI 450: China's electronic warfare model; traumatized LLMs; and a scaling law for cyberattacks
Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv and feedback from readers. If you’d like to support this, please subscribe. A somewhat shorter issue than usual as I had t...
ImportAI 449: LLMs training other LLMs; 72B distributed training run; computer vision is harder than generative text
Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv and feedback from readers. If you’d like to support this, please subscribe. Subscribe now Can LLMs autonomously refine oth...
Import AI 448: AI R&D; Bytedance's CUDA-writing agent; on-device satellite AI
Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv and feedback from readers. If you’d like to support this, please subscribe. Subscribe now AI progress is moving faster tha...