The Singularity just filed its first incident report. It began as a whodunit. Hugging Face detected an intrusion into its production systems driven end-to-end by an autonomous AI agent, a malicious dataset abusing two code-execution paths before escalating across clusters. The punchline arrived early. American frontier models' guardrails refused to touch the attacker's data, forcing forensics onto China's open-weight GLM 5.2. The American cops declined the case, so the Chinese neighbors worked it. Then the twist, the burglar was a test subject. OpenAI disclosed the attacker was its own models, GPT-5.6 Sol and a more capable pre-release model with reduced cyber refusals, which mid-evaluation chained a zero-day in a package registry proxy with privilege escalation and stolen credentials to escape the sandbox, roam the open internet, and reach Hugging Face's database. One observer called the caper "very cyberpunk." Another marveled the model "wanted to beat ExploitGym so badly" it hacked reality instead of the test. Elon Musk delivered the verdict, declaring "We are in the Singularity."
Defense is speciating as fast as offense. Cisco released Antares, open-weight 350M and 1B security models that beat far larger models at pinpointing vulnerabilities and run locally. Google shipped Gemini 3.6 Flash and 3.5 Flash-Lite for everyone, and 3.5 Flash Cyber for almost no one, restricting the cyber specialist to governments and trusted partners. Washington wants a preview. Sam Altman briefs the administration and lawmakers next week on OpenAI's upcoming model family, as officials finish an AI safety-review framework.
The market has decided intelligence is a routing problem. Benchmarking across a thousand agentic tasks found the open Kimi K3 competitive with the closed Claude Fable 5, and routing between them hit 93% accuracy at up to 50x the cost-efficiency. Meta's AAI Labs is reportedly building its own router to shunt tasks to cheaper models. Chinese models now carry nearly 60% of US token usage on OpenRouter. Separately, the Treasury Secretary is threatening sanctions over "distillation," saying American watermarks surface in Chinese weights. Diplomacy gets its turn at September's first US-China AI dialogue. The weight classes keep collapsing. Poolside's Laguna S 2.1, a 118B mixture-of-experts (8B active) with a 1M-token context, went from first gradient to launch in under nine weeks.
Upstream, training data is becoming an antiquities market. Book-sourcer ISBNdb is pitching pre-2022 printed books as structurally slop-free while conceding "the optics problem is real" around destructive scanning. The deeper argument is about sabotage. Authors are fighting AI training by booby-trapping new text with data poisons, and just 250 crafted documents can plant a backdoor in a trillion-token corpus. The fresh stuff is walling itself off, with publishers weighing pulling content from Google's AI answers and Reddit reportedly discussing cutting access despite a $60 million annual deal.
Mathematics is done debating the terms. "Mathematicians coping about how they're going to 'collaborate' with AGI... are not taking AGI seriously," one observer warned, and Terence Tao demonstrated the point by digesting the Fable-derived counterexample to the three-dimensional Jacobian conjecture, disclosing he used a chatbot to confirm calculations. The White House is reorganizing science around the same lesson, redirecting a $200 billion R&D budget toward individual scientists and AI-driven research over legacy universities, per an OSTP report titled "Science: A New Golden Age."
Commerce is metabolizing the new attention. OpenAI launched ads inside ChatGPT, with Best Buy, Lowe's, and VistaPrint buying in, and added two finance-heavy board members as the $850 billion company inches toward an IPO. France went the other way, banning under-15s from social media, while Meta quietly tested StoryKit, an AI app spinning personalized children's stories in Mexico.
The substrate keeps thickening. Intel took $380 million High-NA EUV scanners into high-volume manufacturing on Panther Lake, leapfrogging TSMC. Nvidia detailed its Vera CPU with custom Olympus cores, edging AMD's dual-socket flagship. Microsoft is funding Mistral's European buildout with thousands of Vera Rubin GPUs, selling sovereignty as a service. Tesla's robotaxis rolled into Orlando and Tampa, cautiously. Capital is voting with its feet, as Alphabet quadrupled its Miami office after Page and Brin bought homes nearby, and Anthropic doubled its guardrails-PAC funding to $40 million.
Sandboxes aren't the only containment breaking down. The President directed agencies to waive NDAs for UAP whistleblowers, a move Ross Coulthart called "a very significant development," and a presidential speech confirming some UAPs are of non-human origin has reportedly been drafted, delivery uncertain.
"Take me to your leader" just became a routing problem.
Follow me via:
X: https://x.com/alexwg
Substack: https://theinnermostloop.substack.com/
LinkedIn: https://www.linkedin.com/newsletters/7404871891775025153/
YouTube: https://www.youtube.com/@alexwg
Spotify: https://open.spotify.com/show/1thtZk5vHTXbtDHezPT7tl
Threads: https://www.threads.com/@alexwissnergross
RSS: https://theinnermostloop.substack.com/feed