Our team, like the rest of the tech world, tuned into the OpenAI Dev Day announcement. While our user models do not depend on OpenAI, they have been part of our integration system, enhancing user experiences and offering significant benefits. Since ChatGPT's release last November, we've utilized OpenAI models to address the cold start problem for our users, and have introduced features such as image generation with DALL-E and ChatGPT friends, enabling users to engage with their own ChatGPT conversation instances.
Here are our initial thoughts on how the new GPTs ecosystem can complement the Personal AI ecosystem and benefit our users:
Enhanced Multi-modality Experience:
- Our memory data structure is designed for a multi-modal experience. GPT-4 Vision will enable the automatic creation of memories around pictures for users to recall later. Furthermore, as of November 6th, DALL-E-3 has been made available to all our users.
- Our voice cloning technology, currently in beta for premium users, paired with access to a suite of general Text-to-Speech voices from OpenAI, would broaden the talk experience for all our users.
OpenAI Assistants Ecosystem:
- Personal AI owners have long sought more ways to utilize their AI as assistants. Our product team plans to leverage the extended functionalities of OpenAI assistants, which will likely manifest as quality of life features in our app (e.g., transforming input for better ingestion), should users opt in. This will complement our quick product iteration efforts.
- In the long term, we aim to pursue tighter integrations, incorporating OpenAI assistants into our AI training template offerings, allowing users to seamlessly interact with both their Personal AI and OpenAI assistants in one place.
Reflecting on the direction in which OpenAI is headed and the market it serves, we maintain our focus on the following promises of Personal AI:
User-Trained Models:
- We adopt a user-centric approach, keeping each user's data and models in secure, separate storage. Your Personal AI is the central hub for all your inputs and outputs. It encompasses more than just work documents or travel itineraries; it's a continuous, evolving record of your life experiences.
- We envision your AI as an integral part of every aspect of your life, be it personal messaging, maintaining professional relationships while on vacation, or managing the demands of an influencer lifestyle, all accessible wherever you are.
Cost Efficiency:
- We are committed to delivering best-in-class technology to the 99% in the most cost-effective way. We believe in providing access to top-tier technology while ensuring costs remain consistent. As the value of your AI increases, your costs do not. Our initial assessments of OpenAI assistants find them cost-prohibitive for our users' average volume of use.
Personal AI Ecosystem:
- Human capability extends beyond task completion—it's also encapsulated in our unique knowledge and the rich tapestry of our experiences, especially when shared with others. We launched our Discover feature earlier this year and plan to enhance its functionality to foster diversity of thought and genuine conversations between users and their Personal AIs.
Innovation isn't just about the interaction between one human and one AI—it's about the synergy between many humans and many personal AIs. We believe that a true partnership between humans and AI is forged when the AI is supported by real people with their diverse insights.
.svg)
