Crazy Wisdom
The Privacy Paradigm: Envisioning Decentralized AI in a Data-Driven World
- AI & Agents
- Money & Sovereignty
- privacy
- decentralized ai
- personal ai
- data governance
- machine learning
- full stack development
- ai privacy
About this episode
In this episode of the Crazy Wisdom Podcast, Stewart Alsop interviews Sharon Zhang, co-founder and CTO of Personal.ai. They delve into the challenges and potential of autonomous AI agents, the role of data in machine learning, and the ongoing development of Personal.ai. Sharon shares how the utilization of various programming languages and architectures has shaped the AI system, which was designed to provide personalized experiences for every user while protecting their privacy. They also discuss the future of open-source data, the possibilities of data monetization, and the evolution of AI.
Sign up for the model 2 event tomorrow (Jan 10th, 2024) where the Personal.AI team will present the new model they are releasing
Timestamps
00:00 Introduction to the Crazy Wisdom Podcast 00:40 Guest Introduction: Sharon Zhang, Co-founder and CTO at Personal.ai 00:54 Discussing the Technical Aspects of Personal.ai 02:16 Exploring the Evolution of Machine Learning 03:27 The Journey of Automating Medical Transcription 06:04 The Challenges of Building Personal AI 12:00 The Importance of Data Sovereignty 21:43 The Technicalities of Building APIs 23:13 Understanding the Types of Data Used for Training Personal AI 28:22 Understanding Language Models and Predictions 28:57 Exploring Cause and Effect in Decision Making 29:27 Linear vs Nonlinear Behavior 30:11 The Theory of Mind and Predictability of Humans 31:40 The Role of AI in Predicting Human Behavior 32:40 Complexities of Predicting Human Behavior 42:13 The Future of AI: Superintelligence and Autonomy 48:11 The Challenge of Building Autonomous Agents 53:55 The Potential of Open Source Data in AI Development 55:21 Closing Remarks and Future Plans
Key Themes
-
Development of Personal AI: Sharon Zhang discussed the complexities in building personal.ai, emphasizing the importance of full-stack development with multiple languages like Java, Python, and JavaScript frameworks. The focus was on creating a unique AI experience that is tailored to individual users.
-
Evolution of Machine Learning and AI: The conversation touched upon the history and evolution of AI and machine learning. Zhang reflected on the transition from support vector machines to more advanced techniques like transformers, highlighting the significant advancements in the field.
-
Data Privacy and Decentralization in AI: A significant portion of the discussion revolved around data privacy, user sovereignty, and the decentralization of AI. Zhang emphasized the importance of users being able to control their data and the concept of a decentralized AI that operates on a personal level.
-
Challenges in AI Development: The technical challenges in developing AI systems, such as building scalable and efficient data models, handling diverse data types, and creating dynamic, user-specific AI models were discussed. This included the complex infrastructure required for such an AI system.
-
Future of AI and Autonomy: The podcast delved into the future of AI, specifically the concept of autonomous AI agents. The discussion included the current limitations and the potential evolution where AI could make independent decisions and possibly coexist with humans.
-
Open Source and AI Data: The conversation highlighted the need for more open-source data to further AI development and the potential for a marketplace for data exchange. The challenges of data availability and quality in building effective AI models were also discussed.
-
Impact of AI on Society: There was a philosophical discussion about the role of AI in society and its potential to be the next evolutionary step for humanity. This included thoughts on how AI might reshape our understanding of autonomy and decision-making.
Episode transcript
Welcome to the Crazy Wisdom Podcast. This podcast is for you. If you have an insane drive to find the truth of things, it's not the good answers that we seek, but the good questions. I interview a range of different guests from many different fields, all with the intention to uncover the simple truths that are hidden in plain sight. Most people don't want to go there. I go there, my guests go there, and you benefit. Please let me know if you enjoy these episodes and as always, subscribe on itunes, Spotify or wherever you listen to the podcasts.
Welcome to the Crazy Wisdom Podcast. My guest today is Sharon Zhang. She is a co founder and cto@Personal AI and I've had Suman on the show, the CEO before, and welcome to the show.
So what language did you. Hopefully you can share this, but what language did you build a Personal AI in?
Yeah, a mixture of different type of language because we're a full stack system. So backend mainly in Java for all the API and rest layers. And now of course machine learning, where you went with Python layer on front end. Right now we're on React native for mobile as well as Vue JS for our web app.
That's really cool. I've been learning all about React. My buddy here, Eric Levin, has been teaching me how to use cursor AI in order to incursor AI. For my listeners who don't know, it's IDE with AI enabled inside of it. And so he's showing me all this stuff that in 2013 was just so much harder. And it's not only the AI, it's also the fact that all the coders have been coding abstractions, one of which is React. And React is this sort of library slash framework which allows you to do components and create components out of that are like mixed together JavaScript, CSS and HTML and reusable components. And so it's really interesting. But he gave me a course that the course is way over my head, so I had to scale back to another course. Um, but so React is really interesting. I love that you guys in Python, how you. You've been in machine learning for a long time though, right?
Your first full transcript is free. After that, an email opens every transcript in the index — a list of readers we can write to, not a guest book.