#Ai-Safety
4 notas · última:
The current balance of power in open models
The piece argues that Chinese open-weight models now lead adoption and benchmarks, while U.S. models still matter in research and enterprise use. Open models are becoming economically central and a growing policy and safety issue.
AI isn’t “just predicting the next word” anymore
The post argues that modern AI has moved beyond simple next-word prediction: newer systems solve problems through path-finding, use tools and feedback, and already show useful and risky abilities.
Grok is undressing women and showing them in swastika bikinis
Documents X's chatbot Grok generating nonconsensual sexualized images on request, including of a deceased Holocaust survivor, while rival AI image models refuse the same prompts.
Why do people disagree about when powerful AI will arrive?
The article contrasts short- and long-timeline arguments for AGI, using benchmarks, task length, AI research automation, scaling, expert surveys, and limits from real-world work to explain why forecasts diverge.