When Connor Christou received his cancer diagnosis, the first thing he did wasn’t to search for a single “magic” treatment or a one-time second opinion. It was to start building a system—one that could hold the messy reality of his health in one place and then help him reason through it over time.
Christou, who has spoken publicly about his approach to training, performance, and data-driven self-improvement, treated the diagnosis the way he would treat any other high-stakes problem: by gathering inputs, structuring them, and asking a model to help interpret patterns that would be hard for a human brain to track across weeks and months. In his case, that meant feeding Claude a wide set of personal and medical information tied to his day-to-day routine—blood results, scan data, wearable outputs, and journal entries—so the AI could connect dots across time rather than looking at any single test in isolation.
The result is a story that sits at the intersection of two trends that are accelerating at the same time: AI moving from novelty into practical workflows, and people increasingly expecting their health data to be usable, not just stored. But it’s also a story about intent. Christou’s use of AI wasn’t framed as a replacement for clinicians or a shortcut around medical expertise. It was positioned as an analytical layer—an assistant for sense-making—built to support decisions during a period when clarity matters most.
What makes this approach stand out isn’t simply that AI was used. It’s how the data was assembled and what the system was asked to do with it.
Instead of treating each lab report or imaging study as a standalone event, Christou aimed to create continuity. Cancer care is full of discontinuities: you get a scan, you wait, you get another scan, you compare, you interpret, you adjust. Even when clinicians are meticulous, patients experience the process as a series of separate moments. Wearables and journals can help fill in the gaps, but only if someone can actually synthesize them. That’s where the AI came in.
Christou’s premise was straightforward: the more context you provide, the more useful the analysis can be. Bloodwork doesn’t exist in a vacuum; it changes alongside sleep, stress, activity, nutrition, medication schedules, and the body’s day-to-day responses. Scan results tell you what’s happening anatomically, but they don’t explain how the rest of the system is behaving. Wearable outputs can reflect physiological strain or recovery patterns, while journal entries capture subjective experiences—fatigue, appetite shifts, mood changes, side effects—that may not show up clearly in a lab panel.
By combining these streams, Christou sought to give Claude a fuller picture of his baseline and his trajectory. The goal wasn’t to “predict” outcomes in a simplistic way. It was to help him understand what might be changing, what might be noise, and what might deserve follow-up questions for his care team.
This is where the unique angle emerges. Many AI health stories focus on the model doing something impressive in a vacuum—summarizing a report, generating a list of questions, or producing a general explanation. Christou’s approach is more operational. He’s using AI as a tool to manage complexity: to keep track of longitudinal information, to reduce cognitive load, and to surface relationships that might otherwise remain buried.
In practice, that means the AI can be asked to compare earlier and later states, identify trends, and highlight anomalies. It can also help translate medical language into plain terms, which matters because patients often receive dense information under pressure. When you’re dealing with cancer, you’re not just learning facts—you’re trying to make decisions while your body and mind are under stress. A system that can repeatedly reframe information, connect it to prior context, and keep the timeline coherent can be more valuable than a one-off explanation.
Christou’s method also reflects a broader shift in how people think about data ownership and usability. For years, health data has been treated like a record: something you store, something you retrieve when needed, something you hand off to professionals. But the lived experience of illness is dynamic. People want their data to be actionable. They want it to answer questions like: What changed since last time? Is this consistent with treatment effects? Are there signals I should bring to my doctor? How does my body respond day-to-day, and does that response align with what the scans suggest?
That’s the promise behind bringing multiple data types into one analytical environment. Blood results can show biochemical shifts. Scans can show structural changes. Wearables can show physiological patterns that might correlate with recovery or stress. Journals can capture symptoms and context that don’t fit neatly into a chart. When these are combined, the AI becomes a kind of “timeline interpreter,” helping a person see the story their body is telling.
Of course, there are limits—and acknowledging them is part of being accurate. AI systems can be helpful, but they don’t replace clinical judgment. Medical decisions require validated protocols, careful interpretation, and professional oversight. Wearables are not diagnostic tools, and journal entries are subjective. Even bloodwork and imaging require expert reading and integration with pathology, staging, and treatment plans. An AI can assist with organization and pattern recognition, but it cannot establish causality or guarantee correctness.
Christou’s framing—using AI as an analytical tool rather than a replacement for care—is important. It suggests a workflow where the AI supports the patient’s understanding and preparation, while clinicians remain responsible for diagnosis and treatment. In that model, the value is not that the AI “knows medicine.” The value is that it helps a patient manage information and ask better questions.
There’s another subtle benefit: reducing the friction between different kinds of knowledge. Clinicians often work with structured data and standardized reporting. Patients live with unstructured data—notes, symptoms, daily fluctuations, and the emotional weight of waiting. When those two worlds meet, misunderstandings can happen. A patient might forget details. A clinician might not see the pattern the patient noticed. Or the patient might struggle to articulate what they’re experiencing in a way that fits clinical language.
An AI system that has access to both structured and unstructured inputs can help bridge that gap. It can summarize what happened, highlight relevant changes, and present them in a format that’s easier to discuss. That doesn’t mean it’s always right, but it can make communication more efficient and more precise.
Christou’s approach also points to a future where personal health systems become more like research environments. In research, longitudinal datasets are common, and analysis is iterative. You don’t just look at one measurement; you look at how things evolve. Patients rarely have that luxury in real life—not because they don’t want it, but because the tools and workflows haven’t historically supported it.
By feeding Claude a broad set of inputs, Christou effectively created a personal longitudinal dataset and then used AI to interrogate it. That’s a meaningful shift. It turns the patient from a passive recipient of information into an active participant in sense-making. It also changes the nature of “second opinions.” Instead of only seeking another clinician’s interpretation, the patient can seek another analytical perspective—one that can revisit the same timeline repeatedly and surface different angles.
This is especially relevant in cancer care, where time is both urgent and slow. You need to act, but you also need to wait for results. During that waiting, anxiety can distort perception. People may overinterpret symptoms or miss subtle changes. A system that can contextualize symptoms against prior baselines and known treatment timelines can help stabilize attention. It can also help a patient avoid the trap of reacting to every fluctuation as if it were a major event.
At the same time, it can help ensure that genuine red flags aren’t lost. If wearable data shows a persistent decline in recovery metrics, or if bloodwork trends toward concerning thresholds, or if journal entries consistently describe worsening fatigue, the AI can help compile those observations into a coherent narrative. That narrative can then be used to prompt timely questions or follow-ups with the care team.
One of the most interesting aspects of Christou’s story is the emphasis on structure. Feeding “everything” into an AI isn’t automatically useful. The usefulness depends on how the information is organized, labeled, and connected. Blood results need dates and reference ranges. Scan data needs context about what was measured and when. Wearable outputs need definitions—what metric, what device, what sampling frequency. Journal entries need timestamps and consistent themes so the AI can interpret them as part of a timeline rather than as isolated text.
Christou’s approach suggests that he didn’t just dump files into a model. He built a dataset that could be interpreted. That’s a key point for anyone trying to replicate the idea. The difference between a gimmick and a workflow is the discipline of data hygiene and the clarity of the questions being asked.
It also raises a broader question: what does “AI in healthcare” actually mean when it’s used by individuals? There are at least three categories.
First, there’s summarization—turning complex documents into readable summaries. That’s useful, but it’s mostly static.
Second, there’s decision support—helping interpret what might be happening and what questions to ask. That requires context and careful framing.
Third, there’s longitudinal analysis—tracking changes over time and identifying patterns. That’s where Christou’s approach lands. It’s not just about understanding one report; it’s about understanding a trajectory.
Longitudinal analysis is harder, but it’s also where AI can add distinctive value. Humans are not great at tracking multi-variable trends across long periods, especially under stress. AI can handle that kind of bookkeeping and can propose hypotheses to explore. Again, those hypotheses must be treated as prompts for discussion, not as final answers.
Christou’s story also reflects the growing role of wearables in health narratives. Wearables are often criticized for being noisy or for lacking clinical validation. Those criticisms aren’t wrong. But wearables can still be valuable as early indicators of change in how the body is