• AI Minds Newsletter
  • Posts
  • Using OpenAI’s Sora for Endoscopies, Andrew Ng’s AI Product Management Manifesto, and Why People Can't Detect Voice Clones

Using OpenAI’s Sora for Endoscopies, Andrew Ng’s AI Product Management Manifesto, and Why People Can't Detect Voice Clones

Medical professionals can now view the inside of patient's bodies with AI and without invasive surgery, Andrew Ng examines the merits of AI Product Management, and researchers' latest insights into the difficulties of voice clone detection.

Author

Jose Nicholas Francisco
December 17, 2024

Welcome (back) to AI Minds, a newsletter about the brainy and sometimes zany world of AI, brought to you by the Deepgram editorial team.

In this edition:

  • 🎥 How Medical AI has saved lives over the past year

  • 💉 Using OpenAI’s Sora for Endoscopies!

  • 🔊 Why people are poorly equipped to detect AI-generated Voice Clones

  • ⚡ Introducing Shortcut: The AI-powered tool that lets you work at the speed of voice

  • 🐦 Twitter: Andrew Ng’s AI product management manifesto

  • 🐝 Social media buzz: Karpathy on San Francisco’s billboards and Altman on jobs

  • 📲 Three new, trending AI apps for you!

  • 🎙️ AI Minds Podcast with CEO and co-founder of Phonely Will Bodewes!

  • 🎦 Marques Brownlee reviews the latest version of Sora

  • 💪 SuperAGI Explained: What you need to know

  • 🤖 What if there was a society of nothing but AI Agents?

  • 📖 A deep-dive into counterfactual explanations in AI

Thanks for letting us crash your inbox; let’s party. 🎉

We coded with the brand-new Whisper-v3 over the past week, and the results were not what we expected. Check it out here!

🎥  How Medical AI has saved lives over the past year

This video was posted by NBC News a year ago, showcasing how AI has the capability to save lives in high-tech medical facilities like the University of Florida Health Center. To see how far we’ve come over the past year, check it out!

Video description: “Doctors at the University of Florida Health Center are using artificial intelligence to help monitor their patients. The findings will help them develop algorithms that will soon provide real-time health care recommendations. NBC News’ Dr. John Torres on the future of technology in healthcare.”

🧑‍🔬  Using Sora for Endoscopies and Why People Can’t Detect Voice Clones

Endora: Video Generation Models as Endoscopy Simulators - An endoscopy is a medical procedure that allows doctors to examine the inside of the body without major surgery. This paper introduces Endora, an innovative approach to generate medical videos that simulate clinical endoscopy scenes, using a meticulously crafted spatial temporal video transformer.

People are poorly equipped to detect AI-powered voice clones - Through a series of perceptual studies, the authors of this paper report on the realism of AI-generated voices in terms of identity matching and naturalness. They find that human participants cannot reliably identify short recordings (less than 20 seconds) of AI-generated voices.

⚡ Introducing Shortcut: Poised’s AI-Powered Tool that Lets You Work at the Speed of Voice

💪 What you can do with Shortcut:

  • Get instant answers while staying in your flow

  • Turn spoken thoughts into polished writing

  • Create documents in minutes just by talking

  • Transform scattered ideas into clear plans

  • Practice important conversations naturally

✨ What makes Shortcut different?

Speed and focus are at our core. While other tools pull you away from your work with lengthy responses and constant context-switching, Shortcut stays with you like a real assistant—delivering quick, precise results while keeping you in your flow. Need an answer? Want to write an email? Shortcut handles it instantly, right where you are.

Whether you're a chronic message editor, solo brainstormer, or conversation rehearser, Shortcut amplifies your natural abilities instead of changing them. Finally - technology that works the way your brain does.

🎉 Try it free here: https://www.poised.com/shortcut

🐝 Social Media Buzz: Andrew Ng’s AI Product Management Manifesto and more!

Wavechat is set to revolutionize the way businesses handle customer service with its AI-powered live chat functionality. Designed to operate around the clock, Wavechat ensures that customer inquiries are addressed promptly without the need for an extensive support team.

HotBot is a smarter search engine that utilizes AI and protects user privacy. The platform aims to provide relevant search results while avoiding filter bubbles and data collection.

Kusho is a cutting-edge platform designed to revolutionize the way software developers test and manage APIs. This innovative tool transforms API specifications into comprehensive test suites that seamlessly integrate with Continuous Integration/Continuous Deployment (CI/CD) pipelines.

🎤 The AI Minds Podcast

Will Bodewes, CEO and co-founder of Phonely, is revolutionizing call center operations with advanced voice AI technology. An entrepreneur, athlete, and curious thinker, Will is currently part of the Y Combinator accelerator program in San Francisco.

🤖 Bonus Bits and Bytes

  • 📹 Marques Brownlee reviews AI generated video - Can you figure out what’s real and what’s AI generated? Check out this video to test yourself and to see the depths of Sora’s capabilities.

  • 💪 SuperAGI - SuperAGI is an open-source framework that stands as a beacon for those interested in the development and deployment of autonomous agents. Its open-source nature means a community-driven approach, with contributions from developers worldwide.

  • 🚙 What if there was a society of nothing but AI Agents? - This research paper delves into this extremely curious question, and the results may surprise you!

  • 📚 Counterfactual Explanations in AI - The cornerstone of making AI systems interpretable and user-friendly lies in the concept of counterfactual explanations. This innovative approach revolves around creating hypothetical scenarios to demonstrate how altering specific inputs of an AI model could lead to a different outcome.