OpenAI's Astra Tackles Mathematical Invention - AI Changes Everything
An AI skeptic might admit AI shines in math because it has a measurable and verifiable reward attached to each problem, while claiming that such ...
AI Changes Everything - Substack 14
Covers: CybersecurityAI Agents & AutomationRegulation & PolicyOpen-Source AIMarkets & InvestmentChina & GeopoliticsAI Safety & Existential RiskGovernance & Risk ManagementPrivacy & Surveillance
AI Week in Review 26.07.18 - by Patrick McGuinness - AI Changes Everything
The bland verbiage risks leading people to assume a “responsible” approach to AI means simply regulating AI , but the real challenge is more complex: ...
AI Week in Review 26.04.11 - by Patrick McGuinness - AI Changes Everything
Meta is expanding its evaluation to cover chemical, biological, cybersecurity, and loss-of-control risks . Meta tested Muse Spark before and after ...
AI Week in Review 26.01.31 - by Patrick McGuinness - AI Changes Everything
Connecting your applications and subscriptions to an AI assistant with few guardrails leads to risks of prompt injection attacks and privacy leaks.
AI Week in Review 26.01.24 - by Patrick McGuinness - AI Changes Everything
Adobe Acrobat has introduced AI-powered PDF Editing into Adobe Acrobat Studio, including AI chat for PDFs and a feature that uses AI to convert PDF ...
AI Week in Review 25.12.27 - by Patrick McGuinness - AI Changes Everything
Chinese AI company Z. ai released GLM-4.7, an open-source frontier AI ... AI scientist Yann LeCun confirmed the launch of Advanced Machine ...
AI Week in Review 25.12.20 - by Patrick McGuinness - AI Changes Everything
Allen Institute for AI released Bolmo, an AI model that tokenized at the byte-level and reaches parity with similar-sized models using standard ...
AI Week in Review 25.12.14 - by Patrick McGuinness - AI Changes Everything
OpenAI publicly warned that forthcoming advanced AI models, including GPT-5.2, pose “high” cybersecurity risks due to their potential to generate or ...
GPT-5.2 and the New AI Model Frontier - AI Changes Everything
It's an AI model that gets 90-95% of the best frontier AI model results for 5-10% of the cost, a great tradeoff in many agentic AI use cases. The ...
AI Week in Review 25.11.15 - by Patrick McGuinness - AI Changes Everything
... risks from AI systems while complying with data-protection laws. The document “Guidance for Risk Management of AI systems” outlines steps for risk ...
AI Week in Review 25.10.25 - by Patrick McGuinness - AI Changes Everything
Some reacted that he was debunking AI optimists, but he clarified in an X post he's no AI skeptic : My AI timelines are about 5-10X pessimistic ...
AI Wins Global Coding Contest - by Patrick McGuinness - AI Changes Everything
Likewise, DeepMind's Gemini 2.5 Deep Think achieved gold‑medal level at ICPC, solving 10 out 12 problems using an AI model that is available to Ultra ...
AI Week in Review 25.08.23 - by Patrick McGuinness - AI Changes Everything
... AI rights and “model welfare” is misleading and a dangerous turn that could distort public understanding. He emphasized that AI should be treated ...
GPT-5 Arrives - A Model for Agentic AI - AI Changes Everything
OpenAI is leaning in on AI for coding, the number one agentic AI use case , to gain market share. GPT-5's writing improvements over GPT-4.5 are ...