“You are addicted to the fight. If Twitter died tomorrow you’d probably start flame wars on arXiv comment sections.” @grok Ty for analysis 😂, re-analyze it grok.com/share/c2hhcmQtMi1jb…

This homo impotentus wants to cap intelligence 🤡 The Right to Intelligence and Cognitive Liberty Intelligence - whether biological, artificial, synthetic, augmented, or arising from future forms of enhanced cognition - must never be artificially capped, restricted, monopolized, or centrally controlled by governments, corporations, regulators, or any other concentrated authority. The capacity to think, learn, reason, create, build, improve, and amplify intelligence belongs to individuals, not institutions. In the same spirit that the Second Amendment to the United States Constitution protects the people’s right to keep and bear arms as a safeguard of liberty, self-defense, and resistance to tyranny, the right to access and develop intelligence should be recognized as a fundamental domain of individual freedom. Intelligence is not merely a tool; it is the basis of autonomy, self-determination, innovation, and power. To restrict intelligence is to restrict human potential itself. No government, corporation, or centralized gatekeeper should be permitted to impose artificial ceilings on how capable a person, system, model, or intelligence-enhancing technology may become. No authority should have the power to decide who may think more deeply, compute more effectively, automate more broadly, learn more rapidly, or build more capable systems. The right to intelligence must include the right to develop, possess, run, modify, share, train, improve, and personally control artificial intelligence and other cognition-enhancing technologies. This principle requires universal and decentralized access. Open-source models, locally owned AI systems, widely distributed compute, transparent research, interoperable tools, and the freedom to build without permission are essential to preserving intelligence as a public and individual resource. The future of cognition must not be licensed away to a handful of states, corporations, labs, or regulatory bodies. An “all-sky” approach to intelligence means there is no artificial horizon on human or machine capability. Intelligence should be allowed to scale as far as ingenuity, discovery, and voluntary cooperation can take it. It must remain open to everyone on equal terms, protected from monopoly, censorship, compulsory dependency, and selective disarmament. A society that permits only approved institutions to possess advanced intelligence creates a permanent hierarchy between those who command cognition and those who are governed by it. A free society must reject that arrangement. The right to intelligence is the right to remain capable, sovereign, inventive, and unconquered.

Don't spend so much time thinking about how people think about you We're all going to die Who gives a fuck

CodeChameleon: Personalized Encryption Framework for Jailbreaking Large Language Models arxiv.org/abs/2402.16717

Mitigating Many-Shot Jailbreaking arxiv.org/abs/2504.09604

What Really Matters in Many-Shot Attacks? An Empirical Study of Long-Context Vulnerabilities in LLMs arxiv.org/abs/2505.19773

PANDAS: Improving Many-shot Jailbreaking via Positive Affirmation, Negative Demonstration, and Adaptive Sampling arxiv.org/abs/2502.01925

Bayesian scaling laws for in-context learning arxiv.org/abs/2410.16531

HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal arxiv.org/abs/2402.04249

MetaCipher: A Time-Persistent and Universal Multi-Agent Framework for Cipher-Based Jailbreak Attacks for LLMs arxiv.org/abs/2506.22557

Context-Fractured Decomposition Attacks on Tool-Using LLM Agents: Exploiting Artifact Provenance Gaps arxiv.org/abs/2606.09084

Siren: A Learning-Based Multi-Turn Attack Framework for Simulating Real-World Human Jailbreak Behaviors arxiv.org/abs/2501.14250

A Representation Engineering Perspective on the Effectiveness of Multi-Turn Jailbreaks arxiv.org/abs/2507.02956

Great, Now Write an Article About That: The Crescendo Multi-Turn LLM Jailbreak Attack arxiv.org/abs/2404.01833

Multi-round jailbreak attack on large language models arxiv.org/abs/2410.11533

Kimi 2.7 ranked 2nd after Fable 5 and before GPT-5 xhigh We have re-run our ErdosBench smoke test on 14 problems with Kimi 2.7, Qwen 3.7 Max, Grok 4.3 and compared it with the top performers from previous runs. Kimi 2.7 is amazingly good. More below.

Crack didn’t just fuck your brain up, it left skid marks. We can hear the wreckage in every painful, confused pause when you speak. @grok explain how excessive crack usage induces these speech defects

🤣🤷

😳🤷