Experimental News Clipping Site

Tag: advanced language models

  • Slashdot: Microsoft Launches ‘Vibe Working’ in Excel and Word

    —

    by

    Kurt Seifried
    in

    Source URL: https://slashdot.org/story/25/09/29/1610226/microsoft-launches-vibe-working-in-excel-and-word?utm_source=rss1.0mainlinkanon&utm_medium=feed Source: Slashdot Title: Microsoft Launches ‘Vibe Working’ in Excel and Word Feedly Summary: AI Summary and Description: Yes Summary: The announcement about Microsoft’s new Agent Mode for Office applications significantly impacts the way users create documents and spreadsheets through advanced AI capabilities. This development, leveraging OpenAI’s GPT-5, exemplifies the growing trend of…

  • Simon Willison’s Weblog: Quoting Jason Liu

    —

    by

    Kurt Seifried
    in

    Source URL: https://simonwillison.net/2025/Sep/6/jason-liu/#atom-everything Source: Simon Willison’s Weblog Title: Quoting Jason Liu Feedly Summary: I am once again shocked at how much better image retrieval performance you can get if you embed highly opinionated summaries of an image, a summary that came out of a visual language model, than using CLIP embeddings themselves. If you tell…

  • OpenAI : Introducing gpt-oss

    —

    by

    Kurt Seifried
    in

    Source URL: https://openai.com/index/introducing-gpt-oss Source: OpenAI Title: Introducing gpt-oss Feedly Summary: We’re releasing gpt-oss-120b and gpt-oss-20b—two state-of-the-art open-weight language models that deliver strong real-world performance at low cost. Available under the flexible Apache 2.0 license, these models outperform similarly sized open models on reasoning tasks, demonstrate strong tool use capabilities, and are optimized for efficient deployment…

  • Simon Willison’s Weblog: Quoting Sean Heelan

    —

    by

    Kurt Seifried
    in

    Source URL: https://simonwillison.net/2025/May/24/sean-heelan/ Source: Simon Willison’s Weblog Title: Quoting Sean Heelan Feedly Summary: The vulnerability [o3] found is CVE-2025-37899 (fix here), a use-after-free in the handler for the SMB ‘logoff’ command. Understanding the vulnerability requires reasoning about concurrent connections to the server, and how they may share various objects in specific circumstances. o3 was able…

  • Cloud Blog: SAP & Google Cloud: Enabling faster value and smarter innovation for business excellence

    —

    by

    Kurt Seifried
    in

    Source URL: https://cloud.google.com/blog/products/sap-google-cloud/google-cloud-at-sap-sapphire-2025/ Source: Cloud Blog Title: SAP & Google Cloud: Enabling faster value and smarter innovation for business excellence Feedly Summary: SAP and Google Cloud are deepening their collaboration across data analytics, AI, security, and more to deliver what customers need most: faster paths to business value, lower risk on complex projects, and smart,…

  • CSA: Phishing Tests: Key Questions to Ask

    —

    by

    Kurt Seifried
    in

    Source URL: https://www.schellman.com/blog/penetration-testing/phishing-tests-what-your-provider-should-be-telling-you Source: CSA Title: Phishing Tests: Key Questions to Ask Feedly Summary: AI Summary and Description: Yes **Summary:** The text provides valuable insights into the considerations organizations must evaluate when engaging third-party providers for phishing assessments. It highlights best practices that enhance the effectiveness of these assessments, emphasizing the evolving nature of phishing…

  • ISC2 Think Tank: DeepSeek Deep Dive: Uncovering the Opportunities and Risks

    —

    by

    Kurt Seifried
    in

    Source URL: https://www.brighttalk.com/webcast/5385/638002 Source: ISC2 Think Tank Title: DeepSeek Deep Dive: Uncovering the Opportunities and Risks Feedly Summary: In January 2025, the Chinese open-source artificial intelligence tool DeepSeek caused huge ripples in the AI market, granting user organizations affordable access to powerful LLMs. While this industry-disrupting innovation is indicative of the myriad opportunities that open-source…

  • Hacker News: Gemini hackers can deliver more potent attacks with a helping hand from Gemini

    —

    by

    Kurt Seifried
    in

    Source URL: https://arstechnica.com/security/2025/03/gemini-hackers-can-deliver-more-potent-attacks-with-a-helping-hand-from-gemini/ Source: Hacker News Title: Gemini hackers can deliver more potent attacks with a helping hand from Gemini Feedly Summary: Comments AI Summary and Description: Yes Summary: The provided text discusses the emerging threat of indirect prompt injection attacks on large language models (LLMs) like OpenAI’s GPT-3, GPT-4, and Google’s Gemini. It outlines…

  • 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…

  • Hacker News: Show HN: Benchmarking VLMs vs. Traditional OCR

    —

    by

    Kurt Seifried
    in

    Source URL: https://getomni.ai/ocr-benchmark Source: Hacker News Title: Show HN: Benchmarking VLMs vs. Traditional OCR Feedly Summary: Comments AI Summary and Description: Yes Summary: The text discusses the evaluation of Optical Character Recognition (OCR) accuracy between traditional OCR models and Vision Language Models (VLMs). It emphasizes the potential of VLMs, such as GPT-4o and Gemini 2.0,…