Experimental News Clipping Site

Tag: being

  • Simon Willison’s Weblog: Quoting Emergent Misalignment: Narrow finetuning can produce broadly misaligned LLMs

    —

    by

    Kurt Seifried
    in

    Source URL: https://simonwillison.net/2025/Feb/25/emergent-misalignment/ Source: Simon Willison’s Weblog Title: Quoting Emergent Misalignment: Narrow finetuning can produce broadly misaligned LLMs Feedly Summary: In our experiment, a model is finetuned to output insecure code without disclosing this to the user. The resulting model acts misaligned on a broad range of prompts that are unrelated to coding: it asserts…

  • The Register: Xi know what you did last summer: China was all up in Republicans’ email, says book

    —

    by

    Kurt Seifried
    in

    Source URL: https://www.theregister.com/2025/02/25/china_hacked_gop_emails/ Source: The Register Title: Xi know what you did last summer: China was all up in Republicans’ email, says book Feedly Summary: Of course, Microsoft is in the mix, isn’t it Chinese spies reportedly broke into the US Republication National Committee’s Microsoft-powered email and snooped around for months before being caught.… AI…

  • Simon Willison’s Weblog: Deep research System Card

    —

    by

    Kurt Seifried
    in

    Source URL: https://simonwillison.net/2025/Feb/25/deep-research-system-card/#atom-everything Source: Simon Willison’s Weblog Title: Deep research System Card Feedly Summary: Deep research System Card OpenAI are rolling out their Deep research “agentic" research tool to their $20/month ChatGPT Plus users today, who get 10 queries a month. $200/month ChatGPT Pro gets 120 uses. Deep research is the best version of this…

  • Hacker News: DOGE will use AI to assess the responses of federal workers

    —

    by

    Kurt Seifried
    in

    Source URL: https://www.nbcnews.com/politics/doge/federal-workers-agencies-push-back-elon-musks-email-ultimatum-rcna193439 Source: Hacker News Title: DOGE will use AI to assess the responses of federal workers Feedly Summary: Comments AI Summary and Description: Yes Summary: The provided text discusses a controversial email sent by the U.S. Office of Personnel Management, orchestrated by Elon Musk, directing federal employees to report their weekly accomplishments. The…

  • Slashdot: Meet the Journalists Training AI Models for Meta and OpenAI

    —

    by

    Kurt Seifried
    in

    Source URL: https://news.slashdot.org/story/25/02/23/2111201/meet-the-journalists-training-ai-models-for-meta-and-openai?utm_source=rss1.0mainlinkanon&utm_medium=feed Source: Slashdot Title: Meet the Journalists Training AI Models for Meta and OpenAI Feedly Summary: AI Summary and Description: Yes Summary: The text discusses the evolving role of journalists in the AI landscape, particularly through platforms like Outlier, where they are engaged in training AI models. This shift highlights the intersection of…

  • Hacker News: It is no longer safe to move our governments and societies to US clouds

    —

    by

    Kurt Seifried
    in

    Source URL: https://berthub.eu/articles/posts/you-can-no-longer-base-your-government-and-society-on-us-clouds/ Source: Hacker News Title: It is no longer safe to move our governments and societies to US clouds Feedly Summary: Comments AI Summary and Description: Yes **Summary:** The text critiques the European dependence on American cloud services, emphasizing the risks associated with transferring data and governmental functions to the U.S. It questions…