Personal data Economies: Waze and the Future of Personalized Health Intelligence

In today’s world, medications like Ozempic and Wegovy (GLP-1 receptor agonists) are immensely popular for managing weight, diabetes, and heart health. [1]

But are they right for you? Are there risks you should know about??

A colleague recently mentioned that GLP1 use? has been associated with a flurry of a rare but serious outcome: sudden vision loss in one eye, linked to a condition called NAION (Non-Arteritic Ischemic Optic Neuropathy).

What exactly qualifies as a flurry? A handful per thousand? per hundred thousand? No matter how small the association, for those experiencing NAION, the effect is 100% real.?

This paper will use the example of GLP-1 and NAION to illustrate how personal digital agents are poised to provide personalized health intelligence and transparency to the masses. No lived experience goes to waste. We all benefit from those that went before.?

An early version of this model is already active in providing personalized travel intelligence at the pace of change.

Waze uses the travel patterns of many to deliver individually specific transparency into potential route options of interest. The user is not aware of the identity, the automobiles, the intended routes of other drivers, yet benefits from their collective experience to shed light on what would be the best route to choose at any moment in time. With a rigorous, privacy protecting trust framework, these principles can expand to provide health intelligence transparency which could benefit both our health and financial well-being.?

Back to our example:

What is NAION?

NAION occurs when blood flow to the optic nerve is reduced, leading to swelling and sometimes sudden vision loss. This risk can be higher for those with a smaller or “crowded” optic disc, the part of the eye where the optic nerve connects. Currently, most people don’t know the size of their optic disc, so assessing the individual risk of NAION prior to taking medications like GLP-1s might be challenging.

As our options in health and life become more complex, there’s an emerging solution: personal digital fiduciaries.? These areAI-powered agents designed by trusted intermediaries: Net fiduciaries. Our Net Fiduciary exclusively acts only in our best interest, and not another party. These digital agents can organize the details of your life, while providing individually specific transparency to an array of options.?

The trust framework is designed using a 3-Zone Model [2], keeping personal data secure while allowing for powerful, privacy-protected insights.

Here’s how it works:

The 3-Zone Model: A Future with Digital Fiduciaries

In the 3-Zone Model, each individual manages their information over time within personally-controlled containers.? AI-based agents—we can call them digital fiduciaries (DFs)—are programmed by your own Net fiduciary to help sift through and make transparent an array of potential options while respecting your privacy.?

Zone I: Home Base with our Net Digital Fiduciary (NDF)

Zone I is your personal digital space where you hold and manage a longitudinal, digitized version of your true lived experience.?

People are physical first. We sit, stand, sleep, eat, engage in activities all of which are physical in nature. Today, apps and wearables are converting our physical lived experiences into digital representations that are increasingly accurate. Many of us measure sleep and a few of us even generate clinical grade electrocardiograms using remote digital devices ourselves.?

Using our homebase, we amass thousands of parameters regarding our unique lived experience that can be summarized into key parameters used to match our life with others who share similar traits. These parameters can be collected and organized for us by a federated network of AIs, organized by our Net fiduciary.

Here’s an example:?

Mark has been gaining weight and noticed that his friends are having success loosing weight with GLP-1 medications.? Mark pings his net digital fiduciary/assistant: NDF which manages a host of more limited task digital fiduciaries—let’s call them TDF1, TDF2, and so on—that provide tailored transparency into the current state of experience regarding GLP-1 use.?

Two Points are important to mention here:

1. This system of fiduciaries are not recommendation engines but rather transparency engines. They simply do the work of gathering reliable information regarding various topics and package them for individuals.
2. There are a number of TDFs as each of them have a very limited scope of action, thus leveraging guardrails to establish least privilege that reliably limits what the AI agent is able to do.?

Zone II: Correlation Engine

Zone II is a powerful, real-time AI that draws on data from individuals’ Home Bases to identify patterns and correlations. It’s built to provide precise, relevant insights without revealing anyone’s identity.

Zone III: Commercial and Research Access

Zone III includes organizations, industry, and researchers who benefit from crowd-sourced insights. However, Zone III only receives generalized insights from correlations, keeping individual identities private.

Here’s a mock up flow: NDF-> TDF1->TDF2…and so on.

• TDF1 receives instruction from Mark’s NDF to package Mark’s key attributes and query Zone II for individuals sharing commonalities with Mark that are on GLP-1 medications in order to serve up current information on potential risks and benefits.
• TDF1 executes the above request and receives information from Zone II: 4,382 individuals sharing 80% key attributes with Mark were identified. 22% of them have a history of using GLP-1 medications. Both positive and negative outcomes were listed.? Given the quality and quantity of data reviewed: Zone II assigns a level of reliability to the information provided. Of note, rare cases include a diagnosis of? NAION. This prompts TDF1 to alert Mark’s health-intelligence fiduciary, TDF2.
• TDF2 reviews research and identifies a potential link between a crowded optic disc and NAION. It authorizes…?
• TDF3 to retrieve Mark’s recent optometrist records which may help inform his risk of GLP-1 associated NAION
• TDF4 reviews information coming into Mark’s home base (Zone I) to apply the highest certification/verification possible.?TDF4 checks that Mark’s calendar aligns with the date on the optometry report, checks that the address of the clinic is valid and that the examining clinician is a real person with valid credentials etc. thus ensuring that the clinical report truly belongs to a real person, Mark, and that the details are consistent with the ground truths of Mark’s lived experience.
• TDF5 meanwhile manages market incentives, and finds that Zone III (the commercial zone) is offering a reward for verified data on optic disc size in GLP-1 users and non-users [3]. Mark’s disc size could be valuable as industry seeks better understanding of risk factors affecting the market for these medications. Mark can license this data for specified use in Zone III and reap the economic benefit as he adds to the collective follow-on value generated.

With the 3-Zone Model, Zone II’s correlation engine has access to certified, verified information from live participants, producing high-quality insights. For example, Zone II might indicate that people with a larger optic disc are at a lower risk of NAION when using GLP-1s.

Outcome: Mark is informed that his optic disc is relatively large, which may reduce a potential risk of vision loss with GLP-1 medications. This just in time, personalized transparency into the crowd of individuals using GLP-1s helps Mark and his provider make a more informed decision.

Why Digital Fiduciaries Matter

Net fiduciaries, and their digital counterparts TDFs – are set to become essential tools for navigating complex choices in our lives. They work on our behalf, sorting through data to deliver relevant, personalized insights that protect both our privacy and well-being. To learn more about this vision for the future, and the evolving role for Net fiduciaries, check out the Glia Foundation and GliaNet Alliance.

With Net fiduciaries, plus task-specific digital fiduciaries, the current lived experience of humans on the planet becomes accessible, shedding light on life's potential harms in time to make a difference in a—by the people, for the people trusted framework.

Footnotes and References:

1. In the US, currently 1 in 8 or approximately 12% of U.S. adults have taken GLP-1 receptor agonists for weight loss, diabetes management, or cardiovascular health. Currently, about 6% of U.S. adults are actively using these medications. KFF
2. This is a future in which many of us custody most of our own information and participate in a? 3-Zone model discussed briefly here and in more detail here.
3. Currently, industry can purchase de-identified optometrist records and link them to other de-identified medical records indicating GLP-1 usage. But while this can support marketing, it’s less reliable for scientific research.