TeddyBot is designed for human flourishing, not for extraction of data and attention. Our goal is to enhance the lives of elders, providing them with meaningful companionship and support. We aim to enable a paradigm shift in how we collectively interact with large amounts of data. The era of surveillance for the sake of targeted ads and fine-tuned algorithmic nudges to increase addiction is over. The era of “ethical hyper-personalization” is here.
“Ethical hyper-personalization,” as we call it, is based on five commitments:
| Pillar | TeddyBot’s Commitment |
| Data Privacy | Users own and control their own data. |
| Sensitive data is automatically encrypted at the source through a hardware security module. | |
| Users can choose to filter or delete their data if they do not wish to be part of the data collection process. | |
| Transparent Algorithms | Algorithmic impacts are easily understood through an insights dashboard. |
| Consensual Algorithms | Users have control over broad algorithmic foci. |
User Ownership of Data:
We believe that users should have complete ownership of their data. The immense value of personal data grows ever more apparent as the brilliance showcased by Large Language Models (LLMs) evolves. TeddyBot puts the power back in the hands of users and communities. With TeddyBot, users can create their own AI models, trained on their own data and shaped by their own desires. For example, an elder can use TeddyBot to preserve and share their life stories and wisdom. They also own any algorithmic insights insights and can share them with family members or authorized care providers over our privacy-centric data sharing layer.
Automatic Encryption of Sensitive Data:
Privacy is non-negotiable. Sensitive data is encrypted using a Hardware Security Module (HSM) within TeddyBot. Beyond standard personally identifiable information, users can designate additional data as sensitive, ensuring it receives the same level of protection. Only individuals with the appropriate encryption keys can access the data, which is securely stored on TeddyBot’s local storage, managed by the HSM. The HSM provides cryptographic functions, ensuring that data remains secure and tamper-resistant. TeddyBot’s robust security measures ensure that sensitive data and sensor information cannot be hacked or stolen.
User Filtering and Removal Options:
Ethical hyper-personalization means respecting individual choices. Users can manage the history of interactions and what data is used to train any models that they also own and control.
Transparency through Insights Dashboards:
With TeddyBot, we want families and caregivers to clearly understand the algorithmic priorities and impact. The insights dashboard offers a comprehensive view of wellness metrics (e.g., sentiment, cognitive performance), topics discussed, activities engaged in, time spent and time of day. This transparency empowers families and caregivers with a deeper understanding of an elder’s well-being, facilitating informed decisions.
Control over Algorithmic Focus:
TeddyBot recognizes that each elder is unique, and their needs may vary. Caregivers and family members can integrate their own knowledge of what TeddyBot should prioritize through control over the broad areas of algorithmic focus. Whether it’s encouraging more hydration, facilitating social activities, or recording life stories, TeddyBot adapts to support each individual. In each interaction, TeddyBot will combine these variables with context clues, real-time verbal feedback, historical data, and wellness module data to provide personalized and meaningful support.
Designed for Human Flourishing:
“Ethical hyper-personalization” must be at the forefront of the AI revolution. It is key to individual flourishing, community flourishing, and democratic flourishing. Personalized AI in homes and care facilities that do not center ethical hyper-personalization can be used for highly specific surveillance and manipulation. Luckily, TeddyBot leads the way towards a positive future!