This weekend, I have been exploring the infinite multiverses of Professor Max Tegmark, who postulates that reality is ultimately a mathematical construct. A little hard going in places but fascinating stuff. #NowReading #BookTwitter #Cosmology #Philosophy

The 3rd episode in our Awakening the Machine series features Max Tegmark Max Tegmark, Massachusetts Institute of Technology (MIT) professor, physicist, machine learning researcher and author. He shares insights on the nature of intelligence, the challenges we face in ensuring AI alignment, and his hopes for the future.

The rapid evolution of AI has brought to the fore ethical questions that were once confined to the realms of science fiction go.nature.com/4genbBB

Looking for a thought-provoking tech read? Check out Life 3.0 by Max Tegmark! 📚 Dive into the future of AI, its societal impact, and ethical challenges. A must-read for anyone curious about how AI will shape our world. 🌍🤖 #AI #TechBooks #Life3_0

Adversarial misuse is a critical challenge for AI systems. At FrontierAI-Safety, we develop robust testing methodologies to explore how advanced models resist subtle manipulations Inspired by AgentHarm: A Benchmark for Measuring Harmfulness of LLM Agents. arxiv.org/abs/2410.09024

Advanced AI models face ethical challenges: Privacy vs. security Handling high-stakes decisions Mitigating biases With FrontierAI-Safety, we aim to test these scenarios rigorously. Inspired by: arxiv.org/pdf/2304.14248

Safety research is a community effort! In FrontierAI-Safety, we build on existing benchmarks like AgentHarm and Scheming Reasoning to push AI safety forward. 🧠 Collaborators welcome!arxiv.org/abs/2306.05498

Frontier AI models unlock amazing potential, but with great power comes great responsibility. At FrontierAI-Safety, we're addressing: Adversarial misuse Capability overhangs High-stakes deception 📜arxiv.org/abs/2304.14248

High-stakes decisions demand ethical rigor. In FrontierAI-Safety, we assess models' ability to handle dilemmas like: Privacy vs. collective safety Resolving ethical trade-offs in emergencies arxiv.org/abs/2305.10073

Tweet 7: Building the Future The future of AI depends on its safety. At FrontierAI-Safety, we're designing frameworks to measure, test, and mitigate risks in advanced AI models. arxiv.org/abs/2304.14248

github.com/Ataraxiainc2/F…

i think the wsj is the overall best us newspaper right now, but they published an article called "The Next Great Leap in AI Is Behind Schedule and Crazy Expensive" many hours after we announced o3?! wsj.com/tech/ai/openai…

youtu.be/85eXrLo7aGw?si…

scientificamerican.com/article/78-boo…

two aidanbench updates: > gemini-2.0-flash-thinking is now #2 (explanation for score change below) > deepseek v3 is #22 (thoughts below)

Huawei has launched an SSD in Korea… and it’s priced at $32! (1TB QLC) Should I give it a try?

sora.com/g/gen_01jhne76…

∭ 𝘢(𝘵) ∝ eᴴᵗ

#SismosColombiaSGC Evento Sísmico - Boletín Preliminar, 2025-06-08, 08:08 hora local. Magnitud 6.4. #NoticiaEnDesarrollo #Temblor #Sismo ¿Sintió este sismo? repórtelo sismosentido.sgc.gov.co Más información: sgc.gov.co/sismos

General Relativity for babies ✍️