Why Your AI Should Work For You
Building Continua's vision of AI agents that understand and adapt to you, starting with personalized podcasts
Hello everyone! This is the first post on the Continua AI blog. We'll be sharing updates about our work in AI, insights from our team, and news about our products and technology. Speaking of which, I'm excited to tell you about Continua and a product we've been working on for you to create custom podcasts, built on our agent AI platform.
Posts like this often begin with a tl;dr. If there's one thing I'd like you to do, please visit our homepage continua.ai where you can find our app. Tell the app what you want to hear about, and our AI will create a podcast just for you – engaging conversations about any topic that interests you. Each episode is generated specifically for your interests and knowledge level. Please give it a try! 🙂 This is a beta launch, and we're eager to get your feedback.
Ok, so what's our mission and why start with podcasts? I left Google to start this company with the idea that the most precious resource is human attention. It's scarce, valuable, and under constant demand. We seek, at Continua, to preserve and amplify the impact of human attention. This goal informs all of our decisions.
The first part of our journey was building an agent platform. For us, agents are AI systems that truly work for you – they act in your interests rather than ours, maintain memory of your interactions, and take initiative by anticipating your needs. This foundation lets us capitalize on the unprecedented opportunities in LLM technology to build experiences that put users first.
In October of last year, while using NotebookLM we started thinking about how the two-host podcast format is well-suited to convey information in an effective and personalized manner. And so we built an app that creates personalized podcasts (Web / iOS / Android). For any given topic, one host might be well-informed, while the other asks most of the questions. That allows the listener to live in the mind of the question asker. It sidesteps the problem of staring at the blank screen of a chatbot and not knowing what to ask.
You can stay informed on topics you like, with episodes generated at whatever cadence you like. You can be entertained by choosing interesting styles ("arguing couple," anyone?), or connect your email and get a great podcast that surfaces what you need to know about and act on. Our unique way of sourcing content enables the product to expose you to a wider variety of information than you might get from typical search engines or media platforms.
All of this is in the service of giving time back to you, and enabling you to choose what you will pay attention to. We're really excited to launch more applications on our platform that share this mission in the coming months.
We like to share how our technology is evolving, so stay tuned for our upcoming technical and design deep dives. Our excellent team, based in NYC, Seattle, and SF, will be talking about our web scraping, how we rank and cluster content, the art in creating an engaging transcript, quality evaluation, and the contrast between product development at a startup versus Google.
One quick personal note on that last topic: At Google, I was fortunate to work with incredible talent across engineering and many other disciplines. Since starting Continua, I've gained an even deeper appreciation for the variety of skills needed to build something great, including the practical (office space and building desks), creative (design of the product), hustle (putting yourself out there), and more. What was abstract in these other disciplines is now concrete, and I see this shared ownership energizing all of us here.
The industry narrative when we started was about the risk that everything was just going to be a thin wrapper on ChatGPT. Now, as frontier labs compete and models become commoditized, the real innovation is happening at the product layer. We truly believe that 2025 will be the year of the agent!
With gratitude, urgency, and excitement,
David
1880 diagram of Thomas Edison's electric lamp.