Tag: -4o

  • Simon Willison’s Weblog: Codestral 25.01

    Source URL: https://simonwillison.net/2025/Jan/13/codestral-2501/ Source: Simon Willison’s Weblog Title: Codestral 25.01 Feedly Summary: Codestral 25.01 Brand new code-focused model from Mistral. Unlike the first Codestral this one isn’t (yet) available as open weights. The model has a 256k token context – a new record for Mistral. The new model scored an impressive joint first place with…

  • Hacker News: How outdated information hides in LLM token generation probabilities

    Source URL: https://blog.anj.ai/2025/01/llm-token-generation-probabilities.html Source: Hacker News Title: How outdated information hides in LLM token generation probabilities Feedly Summary: Comments AI Summary and Description: Yes ### Summary: The text provides a deep examination of how large language models (LLMs), such as ChatGPT, process and generate responses based on conflicting and outdated information sourced from the internet.…

  • Slashdot: ‘Omi’ Wants To Boost Your Productivity Using AI and a ‘Brain Interface’

    Source URL: https://hardware.slashdot.org/story/25/01/08/2310255/omi-wants-to-boost-your-productivity-using-ai-and-a-brain-interface Source: Slashdot Title: ‘Omi’ Wants To Boost Your Productivity Using AI and a ‘Brain Interface’ Feedly Summary: AI Summary and Description: Yes **Summary:** The announcement of the AI wearable device, Omi, by Based Hardware at CES highlights a novel approach to personal productivity enhancement through AI integration. The device’s open-source platform addresses…

  • Simon Willison’s Weblog: microsoft/phi-4

    Source URL: https://simonwillison.net/2025/Jan/8/phi-4/ Source: Simon Willison’s Weblog Title: microsoft/phi-4 Feedly Summary: microsoft/phi-4 Here’s the official release of Microsoft’s Phi-4 LLM, now officially under an MIT license. A few weeks ago I covered the earlier unofficial versions, where I talked about how the model used synthetic training data in some really interesting ways. It benchmarks favorably…

  • Hacker News: Human study on AI spear phishing campaigns

    Source URL: https://www.lesswrong.com/posts/GCHyDKfPXa5qsG2cP/human-study-on-ai-spear-phishing-campaigns Source: Hacker News Title: Human study on AI spear phishing campaigns Feedly Summary: Comments AI Summary and Description: Yes Summary: The text discusses a study evaluating the effectiveness of AI models in executing personalized phishing attacks, revealing a disturbing increase in the capabilities of AI-generated spear phishing. The findings indicate high click-through…

  • Hacker News: Notes on the New Deepseek v3

    Source URL: https://composio.dev/blog/notes-on-new-deepseek-v3/ Source: Hacker News Title: Notes on the New Deepseek v3 Feedly Summary: Comments AI Summary and Description: Yes **Summary:** The text discusses the release of Deepseek’s v3 model, a 607B mixture-of-experts model that showcases exceptional performance, surpassing both open-source and proprietary competitors at a significantly lower training cost. It highlights the engineering…

  • Slashdot: Microsoft Integrates a Free Version of Its ‘Copilot’ Coding AI Into GitHub, VS Code

    Source URL: https://developers.slashdot.org/story/24/12/21/0443255/microsoft-integrates-a-free-version-of-its-copilot-coding-ai-into-github-vs-code?utm_source=rss1.0mainlinkanon&utm_medium=feed Source: Slashdot Title: Microsoft Integrates a Free Version of Its ‘Copilot’ Coding AI Into GitHub, VS Code Feedly Summary: AI Summary and Description: Yes Summary: GitHub has launched a free version of its Copilot AI coding assistant tool, now included by default in VS Code, aiming to remove barriers for developers and…

  • Simon Willison’s Weblog: Quoting François Chollet

    Source URL: https://simonwillison.net/2024/Dec/20/francois-chollet/#atom-everything Source: Simon Willison’s Weblog Title: Quoting François Chollet Feedly Summary: OpenAI’s new o3 system – trained on the ARC-AGI-1 Public Training set – has scored a breakthrough 75.7% on the Semi-Private Evaluation set at our stated public leaderboard $10k compute limit. A high-compute (172x) o3 configuration scored 87.5%. This is a surprising…

  • Hacker News: OpenAI O3 breakthrough high score on ARC-AGI-PUB

    Source URL: https://arcprize.org/blog/oai-o3-pub-breakthrough Source: Hacker News Title: OpenAI O3 breakthrough high score on ARC-AGI-PUB Feedly Summary: Comments AI Summary and Description: Yes **Short Summary with Insight:** OpenAI’s new o3 system has achieved significant breakthroughs in AI capabilities, particularly in novel task adaptation, as evidenced by its performance on the ARC-AGI benchmark. This development signals a…