In this episode, we discuss OpenAI's decision to permanently remove Mixpanel following a data exposure incident, Microsoft's launch of Fara 7B for on-device AI control, and UC Berkeley's Chain of Visual Thought framework for direct visual reasoning. We explore how the Mixpanel breach exposed user names and email addresses through unauthorized API access, prompting OpenAI to eliminate the analytics provider entirely from production systems. We then examine Microsoft's seven billion parameter Fara 7B model that runs locally on laptops to control computers through screenshots, offering fast and private AI agent capabilities without cloud connectivity. Finally, we break down UC Berkeley's innovative approach that allows AI models to reason directly with visual information instead of converting images to text, eliminating the language bottleneck that limits pixel-level understanding in scientific imaging, medical analysis, and robotic perception tasks.https://www.aiconvocast.comHelp support the podcast by using our affiliate links:Eleven Labs: https://try.elevenlabs.io/ibl30sgkibkvDisclaimer:This podcast is an independent production and is not affiliated with, endorsed by, or sponsored by OpenAI, Mixpanel, Microsoft, UC Berkeley, or any other entities mentioned unless explicitly mentioned. The content provided is for educational and entertainment purposes only and does not constitute professional, financial, or legal advice. Affiliate links may generate commission for the podcast at no additional cost to you.

Ver todos los episodios