The Thread That Started It
In January 2013, Rob Rhinehart posted on Hacker News about a substance he'd been living on for 30 days: a beige powder mixed with water that replaced all food. He called it Soylent. The post was titled "How I Stopped Eating Food" and it went viral in the particular way that only early-2010s internet posts could — equal parts horrified fascination and "I can do that better."
By 2014, the Soylent discourse had splintered across Reddit, a dedicated Discourse forum, and dozens of personal blogs. What made the scene interesting wasn't the product itself — it was the community that formed around hacking on nutrition as if it were a software problem.
The DIY Scene
Before Soylent shipped a commercial product, people were mixing their own. The DIY forum on discourse.soylent.me (later renamed and eventually shut down) was a strange blend of biohacking, spreadsheet engineering, and open-source culture.
The basic formula was public: oat flour for carbs, whey or rice protein, oils for fat (canola, olive, MCT), and a micronutrient mix — usually a multivitamin ground to powder or individual bulk supplements from PureBulk or BulkSupplements. People tracked their recipes in shared Google Sheets with columns for grams, calories, cost per day, and notes like "tastes like pancake batter but grittier" or "added psyllium husk, gas resolved."
The most prolific DIYers maintained versioned recipes. People Chaw, a pseudonymous forum member, published a formula called "People Chow" that became one of the most forked recipes. It was essentially a text file with ingredient weights, Amazon links, and mixing instructions — a Makefile for food.
The Blood Test Subculture
The defining ritual of the DIY Soylent scene was the blood test. Someone would live on their formula for 30 days, get a full metabolic panel and lipid profile, and post the results — before and after — as a forum thread. These threads were the scene's version of a pull request review.
The collective baseline was surprisingly thorough. People tracked: complete blood count, comprehensive metabolic panel, lipid panel, vitamin D, B12, iron, ferritin, testosterone, thyroid panel (TSH, T3, T4), and sometimes CRP and HbA1c. A shared Google Sheet aggregated results across dozens of participants.
The results were generally unremarkable — which was the point. Most people's blood work stayed normal or improved (largely because they'd been eating worse before). A few had elevated triglycerides from high-carb formulas, which led to a wave of low-carb "keto chow" variants. One person discovered a B12 deficiency that their doctor had missed. The forum collectively debugged it: the multivitamin they were using had cyanocobalamin at a dose too low for their particular absorption issues.
The Engineering Mindset
What made the Soylent scene distinct from other nutrition communities was how thoroughly it rejected food culture. There was no talk of "clean eating," no moralizing about processed ingredients, no appeals to tradition or nature. The forum ethic was: if the numbers work and the blood work confirms it, the food is valid.
This was nutrition treated as an engineering problem: inputs (macros, micros, fiber), outputs (blood markers, body composition, subjective energy levels), with iterative optimization. The community's blind spot was everything outside those measurable parameters — gut microbiome diversity, the role of phytochemicals, long-term effects of liquid diets on dental and digestive health. These were occasionally raised and mostly dismissed as insufficiently quantified.
What Happened After
Rosa Labs (later Soylent Nutrition, Inc.) raised venture funding, shipped a commercial powder in 2014, and eventually pivoted to ready-to-drink bottles. The DIY scene faded as the commercial product became available. The Discourse forum was shut down around 2019, taking years of blood test data and recipe iterations with it — though fragments survive on archive.org and in scattered Reddit threads.
The commercial product diverged from the DIY ethos. The original Soylent was designed to be hacked — the recipe was public, the ingredients commodity-grade. The bottled version was a consumer product. The community that built the early momentum was no longer the target market.
In retrospect, the Soylent DIY scene was an early example of what would become the quantified-self movement's nutrition wing — people treating their bodies as systems to be instrumented and optimized, sharing data openly, and iterating in public. The specific product was less important than the methodology it accidentally spawned: blood work as CI pipeline, recipes as versioned source code, and a forum as an ad-hoc clinical trial with n=300 and no control group.