英国AI安全研究所评估了新的AI模型，注意到Mythos和GPT-5.5在网络安全能力方面均有显著进步。研究人员发现，这些模型的上限受限于token使用而非内在能力，能力翻倍时间估计为4.5个月。另外，Palantir首席执行官Alex Karp批评德国不愿采用Palantir的软件，并敦促欧洲优先考虑已获验证的乌克兰国防技术。 AI
影响 新的AI模型在网络安全能力方面展现出显著的提升，快速的翻倍时间表明开发速度正在加快。
排序理由 该集群包含英国AI安全研究所关于新AI模型的研究发现，这构成了研究内容。
在 Mastodon — sigmoid.social 阅读 →
AI 生成摘要 · Google Gemini · 来自 7 个来源。 我们如何撰写摘要 →
@matthew_berman on YT! We only have 2 years... https:// youtu.be/ak6fQ2Yjwy0?si=Dmc9X1 41PU-dlNNi # AI news # Cybersecurity # Anthropic # Mythos Context: # Aschenbrenner letter https:// situational-awareness.ai 5/15/2026
@ emollick The UK’s state AI Security institute findings on latest AI models: 1) Mythos is a big gain in cyber capabilities. But so is GPT-5.5 2) It is hard to establish an upper bound on Mythos/GPT-5.5, which appear to be limited by tokens used, rather than ability. 3) Capabilit…
@natebjones # AI News 📰 # Otd # AgenticAI The Trillion Dollar Agentic Workflow Opportunity Is Here https:// youtu.be/jwtpMSRAPAQ?si=O0L-If rn8lovzBa4 5/14/2026
@matthew_berman on YT! We only have 2 years... https:// youtu.be/ak6fQ2Yjwy0?si=Dmc9X1 41PU-dlNNi # AI news # Cybersecurity # Anthropic # Mythos Context: # Aschenbrenner letter https:// situational-awareness.ai 5/15/2026
# Ai News # Ukraine NewsInsideUkraine https:// t.me/NewsInsideUkraine/744785 Palantir CEO Alex Karp delivered a sharp rebuke to Germany, criticizing the Bundeswehr's refusal to integrate Palantir's software and urging Europe to buy battle-tested Ukrainian defense technology inste…
@ emollick The UK’s state # AI Security iIstitute findings on latest AI models: 1) Mythos is a big gain in cyber capabilities. But so is GPT-5.5 2) It is hard to establish an upper bound on Mythos/GPT-5.5, which appear to be limited by tokens used, rather than ability. 3) Capabil…
@matthewberman on YT! Everyone's getting hacked # AI # Cybersecurity # Mythos 5/13/2036 https:// youtu.be/hAzhVloGkOw?si=03S2wO Es3_iflQzp