Engagement Overview

UNICEF counts 144 million children under five who are stunted and 47 million who are wasted. Many of them will not eat even when food is in front of them. Appetite loss is so central to severe acute malnutrition that it has been built into the clinical pathway itself: the appetite test is what decides whether a child is treated at home or admitted to inpatient care.

Our design followed from that directly. Find the signal that suppresses appetite, express an enzyme that degrades it, and put that enzyme inside an organism that already lives where the signal is made. Each step was tractable, and for several months we treated the tractability of the biology as evidence that the idea was sound.

What we had not examined was the assumption sitting underneath all of it, which is that appetite is the thing standing between a malnourished child and a meal. Our Human Practices work is the record of eight people taking that assumption apart, and of what we changed once they had.

We approached specialists in each field our project touches, and deliberately also two organizations working where malnutrition is actually concentrated rather than where it is studied. None of them told us to stop. All of them told us the problem was wider than our description of it.

In their own terms: a behavior analyst gave us seven reasons children refuse food, and appetite was one. A dietitian gave us about a dozen barriers that stop a patient from eating, including one patient who had no dentures and could not chew. An endocrinologist showed us how many signals set appetite at once. A microbiome scientist asked for kinetic evidence we did not have and noted that the gut already makes our enzyme itself. A public health specialist in Bangladesh asked what a hungrier child eats when the cheapest food nearby is chips and sugary drinks. A food systems organization in India told us that in the communities it serves, the nutritious food is not there in the first place. A pediatrician of thirty-five years told us that tube feeding gets nutrition into a child but does not make the child eat. And a geriatrician told us that the molecule we target sits at different levels depending on which illness caused the malnutrition.

None of this stopped the work. It changed four specific things: what we claim the probiotic does, where in a treatment pathway it belongs, which children it must exclude, and how we intend to measure whether it works. Each one is traceable to a named conversation.

Alongside those changes we built something we could deliver immediately. The $3 Meal Plan is a set of more than thirty balanced recipes costing under three dollars a serving, documented on QR-coded notecards linking to videos of students cooking them, and handed out at a back-to-school autism expo in Houston. It exists because a behavior analyst told us, in our very first expert session, that the help those families needed was not something our science could provide.

Our engagement work is split across three documents, and they are kept apart deliberately. Running them together would let us claim that a flyer changed science, or that a webinar fed a community, and neither is true. The eight conversations themselves, in full, are on the Integrated Human Practices page; the Education record is on the Education page; the Entrepreneurship record is on the Entrepreneurship page.

The four-field record is worth explaining here, because it governs the whole of the Integrated Human Practices document. Early on we wrote meetings up as summaries, and the summaries were useless: they recorded that a conversation had been pleasant without recording what it did to the project. Every record is now forced into four fields, of which the last is the one that matters. A record whose Consequence reads “nothing yet” is an honest record. A record with no Consequence field at all is a thank-you note.