From the CEO: January 2026
My supplement stack, how I'm thinking about AI + health tech, and blue light blockers that actually work.
Three things on my mind this month
1. My supplement stack
I've historically taken a very minimal approach with supplements but as I've experimented and done more research I've found myself adopting a widening daily stack. What I take as of now (I will probably add and remove):
- Creatine (10g-15g): Known for muscle performance and cognitive benefits. I take a higher dose based on evidence and discussion (e.g. from Rhonda Patrick) that muscles are greedy and the brain only gets what's leftover. Adopting higher daily doses of Creatine and Protein (150-200g/day for me) is the change I credit with finally resolving a multi-year back injury.
- Psyllium Husk (15g): For digestion and cholesterol. Fiber is a massive lever for lowering ApoB, and this dose helps me hit my targets. I mix it into yogurt to avoid the texture issues.
- Lithium Orotate (1mg): A specific bet on cognitive health. Brain lithium is linked to lower rates of Alzheimer's (see this study); at this micro-dose, I consider it a speculative but low-risk addition for an area of my health that I'm intensely interested in protecting.
- Nattokinase (10,800 FU): Cardiovascular insurance. While low doses help blood pressure, I'm taking this higher dose based on evidence suggesting it may actually break down arterial plaque and improve lipids. (Members can get a special rate by ordering through "Shop" in the Levels member portal).
- Fish Oil (1.8g): High-quality EPA/DHA for lipid profile, systemic inflammation, and brain health. I view this as foundational baseline support.
- Vitamin D (4,000 IU with K2): Critical for immune function and hormonal health. I track my levels through the 100+ marker Levels Comprehensive panel and am improving from chronic lows.
- Vitamin C (1g): My cortisol management tool. My sleep has been disrupted for awhile and I suspect my rising fasting glucose is due to stress/cortisol. Vitamin C has shown promise in blunting cortisol spikes (study, context), so this is my way of buying some physiological margin.
2. The AI <> Health Tech tsunami
In just a few short years, Large Language Models (LLMs) have completely transformed how software is built. As a mechanical engineer running a software company, I see parallels to what 3D printing did for physical product prototyping (easy, rapid, cheap)—except AI is far more capable and scalable. With tools like Cursor, Claude Code, and Figma Make, building a functional software prototype you can actually use takes 5 minutes, and is as easy as describing the idea to a colleague. This is fueling a massive innovation cycle in digital health (and every other industry). But that is just the surface.
We're entering a new phase where Large Language Models have rapidly evolved from questionably accurate chatbots to independently practicing medicine. This week, Utah became the first state to authorize AI to autonomously manage prescriptions. Simultaneously, the FDA publicly signaled a lenient approach to regulating AI in certain healthcare contexts like decision support tooling. These things happened far faster than I expected and yet, everything in AI is accelerating at an accelerating rate. It's going to get interesting!
Consumers are moving even faster. A new report from OpenAI revealed that more than 40M people ask AI about healthcare every day. Making up 5% of global ChatGPT traffic. I thought it was interesting to see the heavy usage in US "hospital deserts", areas lacking access to facilities and specialists. The report further notes that two-thirds of US physicians have used AI tools in their work. While "the system" debates the role of AI, patients and practitioners are already embracing it at a pretty incredible pace.
How I'm thinking about this for Levels: AI is changing how we work and what we build at Levels. We've already seen with our in-app AI how incredibly valuable it is for members to be able to query their data and our knowledge base in plain, simple language. We're using it to improve features, making our CGM insights and meal plans better and more personalized, and to build new ones like Document Center in the member portal: an intelligent health record you can converse with and control completely. As always, our north star is reliable, reported, accurate guidance, so we're moving fast while being careful that our training is thorough and your data is always private and secure. I'd love to hear what AI-powered features you'd like to see out of Levels.
3. Product rec: Ra Optics Maxwell Blue Light Blockers
I have to do quite a bit of evening screen use, the light from which can be very disruptive to circadian rhythm and melatonin production. Ra Optics Maxwell Blue Light Blockers are my go-to glasses to try to manage the consequences. These aren't your lightly tinted Amazon blue blockers—they are serious filtering glasses and the lenses are deep red.
Why I love them: They use a pigmented lens that is easier to see contrast through than other aggressive lenses I've tried, while claiming to block 100% of the artificial blue light. The most tangible difference I've noticed is reduced eye strain when using a screen at night with them.
Follow me on X here. Other than Levels, I have no affiliation with any of the products mentioned or linked above—I'm just a satisfied customer.
—Josh
Take control of your health with expert guidance
Levels pairs real-time glucose data and comprehensive lab testing with clinician analysis and personalized support—everything you need to turn insights into real health improvements. Click here to get started with Levels.
