Building

Not another calorie counter with a green circle for protein.

I am building a food database that describes what food actually contains: complete macro- and micronutrients, omega-3 to omega-6 ratios, amino-acid profiles, and the details that disappear when an app reduces nutrition to calories, carbs, fat, and protein.

The first problem is the least glamorous and probably the most important: build data I can trust before building conclusions on top of it.

Current focus

#

• Designing a food and nutrient model that does not fall apart when the data gets complicated
• Finding and normalizing reliable food-composition sources
• Keeping units, serving sizes, raw/cooked states, and missing values explicit
• Tracking where every value came from instead of creating one mysterious “health score”

Roadmap

#

1. Food database · active
   Full nutrient records, source provenance, normalization, and data-quality checks.

2. Nutrition engine · next
   Derived values such as omega ratios, amino-acid completeness, nutrient density, and meaningful comparisons.

3. Tracking interface · planned
   Fast food logging without turning the app into a second job.

4. Biohacker layer · planned
   Better questions, deeper views, experiments, trends, and useful warnings without pretending software can diagnose a human.

Rules for the project

#

• Evidence and source data stay visible.
• Unknown values remain unknown.
• Calculations must be explainable.
• The interface should reveal detail without drowning the user in it.
• No feature gets added merely because every other nutrition app has it.

This page will change as the project changes. That is the point.

The premise is the part most people refuse to accept: you cannot stop someone from taking drugs. “Just don’t” is not a plan, it is a way of feeling clean about the outcome.

If you want the evidence rather than the argument, read the man who got addicted to heroin in real time. Nobody in that story was short of information. He watched it happen to himself, in public, narrating the whole way down, and it happened anyway.

So the book does the other thing. It assumes the drugs are getting taken and tries to get the person taking them out the other side intact.

What it covers

#

• Knowing what you actually have, instead of what you were told you have
• Testing: reagents, scales, and what a result does and does not prove
• What impurities look, smell, and behave like
• Cleaning up someone else’s chemistry, and when that is the wrong call
• Dangerous interactions, the ones that actually kill people rather than the ones that make headlines
• Dosing and tolerance, and the way both quietly lie to you
• Asking for help without losing your job, your kids, or your nerve

Current focus

#

• Pulling the harm reduction posts apart and rebuilding them as chapters
• Working out what belongs in a book and what should stay a post
• Sourcing properly: every number gets a citation or an admission that there isn’t one

Roadmap

#

1. Compilation · active
   The posts I have already written, restructured and stripped of the filler an article needs but a chapter does not.

2. The gaps · next
   The subjects I have not written up yet, and the interactions that only matter once you combine chapters.

3. Review · planned
   Read by people who know more than I do, especially the clinical parts.

4. Release · planned
   Format undecided. It should end up wherever it is easiest to reach at three in the morning.

Rules

#

• No synthesis instructions. Not now, not in an appendix, not ever.
• No moralising in either direction. I am not selling abstinence and I am not selling drugs.
• Every claim carries its own uncertainty. Where the evidence is thin, it says so.
• Written for the person who already took it, not the person deciding whether to.

I have been on the wrong side of all of this. That is the only qualification I am claiming.