Product.ai / Research
Research

What we built, how we built it, and what we found. From the people who did the building.

Loop engineering is the meme. Outcome engineering is the job.

Teams everywhere are building AI agent loops, but almost nobody trusts them to run unattended. Outcome engineering is the secret to loops you can rely on.

MQ
Michael Quoc
Founder & CEO
Read the essay →
The Prediction Scorecard

Predictions graded in public, the moment each verification window closes.

We publish falsifiable, timestamped calls before we know the outcome — and we commit to grading ourselves here, openly, on every one of them. A miss renders exactly like a hit.

Read the full study: Skincare Prediction Study, 2026–2027: Seven Calls on Where Skincare Science and Market Hype Diverge →
1 Confirmed
0 Wrong
6 Pending
No. 1 — The Peptide Reckoning
Pending

The topical peptide and growth factor category will face a credibility crisis within 12 to 18 months — visible in brand repositioning, category growth deceleration, and independent clinical scrutiny.

Verify: By Q4 2027

A top-10 peptide brand publicly reformulates or repositions away from deep-delivery or collagen-stimulation language; FTC action lands on peptide marketing claims; category year-over-year growth falls below 5% from its current ~15%+ trajectory.

No. 2 — Tranexamic Acid Is the Stronger "Next Retinol" Candidate — Not GHK-Cu
Pending

Tranexamic acid will emerge as the next universally-adopted active — visible in new-product launch frequency and entry into standing dermatology guidance — displacing GHK-Cu as the leading "next retinol" candidate.

Verify: By end of 2027

Tranexamic acid ranks among the top 5 most-launched headline actives in U.S. OTC skincare new releases (Sephora, Ulta Beauty, Amazon Premium Beauty); it enters American Academy of Dermatology or American Society for Dermatologic Surgery hyperpigmentation treatment guidance; GHK-Cu fails to reach either threshold.

No. 3 — Korean Sunscreen Displaces American SPF as the Default Recommendation
Pending

Korean-formulated sunscreens will displace American-formulated SPF as the default recommendation in U.S. skincare communities within 12 months — visible at retail shelves and in U.S. SPF market share data.

Verify: By end of 2026

A Korean-formulated sunscreen reaches top-5 position in U.S. retailer SPF bestseller rankings (Sephora, Ulta Beauty, or Amazon Premium Beauty), followed by measurable Korean SPF share gain in U.S. market data (Circana, Euromonitor, or Nielsen) through Q1 2027.

No. 4 — Peptides Will Be Repackaged, Not Abandoned
Pending

The peptide category will pivot from collagen-stimulation claims to inflammation-modulation language rather than collapsing.

Verify: By end of 2027

A major peptide brand repositions its flagship messaging around surface-level inflammation modulation, epidermal signaling, or barrier-layer peptide activity — replacing the deep-delivery and collagen-stimulation framing that currently dominates peptide marketing copy.

No. 5 — Tallow Skincare Peaks and Backlash Begins Within 6 to 9 Months
Pending

The tallow skincare movement will peak within 6 to 9 months — visible in decelerating retail expansion and the arrival of dermatology critique into mainstream clinical channels.

Verify: By Q1 2027

Tallow SKU growth at major U.S. retailers (Sephora, Ulta Beauty, Amazon Premium Beauty) decelerates and reverses; a peer-reviewed dermatology publication or society statement addresses the C16–C18 fatty acid gap versus functional barrier repair; a major tallow brand reformulates or a retailer delists it.

No. 6 — The Niacinamide Concentration Correction
Pending

The topical market will converge on 4–5% niacinamide as the optimal concentration. Products at 10% and above will lose share.

Verify: Through 2027

Products at 10% or higher niacinamide lose collective market share in skincare category data (Circana, Euromonitor, or Nielsen); new launches converge toward 4–5% concentration; flagship SKUs from formulation-credible brands (Paula's Choice, La Roche-Posay, CeraVe, Korean premium formulators) hold 4–5% while 10%+ SKUs reformulate down.

No. 7 — Bemotrizinol Gets Its FDA Final Order Before End of Q3 2026
Confirmed

The FDA will publish a final administrative order authorizing bemotrizinol as an OTC sunscreen active ingredient by September 30, 2026.

Verify: By September 30, 2026

Federal Register publication of a final administrative order on bemotrizinol (FDA docket OTC000039); FDA OTC Ingredient Status Portal update marking bemotrizinol as an approved OTC sunscreen active. Confirmed June 10, 2026: the FDA issued final administrative order OTC000039 amending OTC Monograph M020 to add bemotrizinol, at concentrations up to 6 percent, as a GRASE sunscreen active — Federal Register notice published June 10, 2026; order effective August 9, 2026. Both named signals fired more than three months ahead of the September 30 window.

The Full Index

Every published piece, one line each — sorted newest first.

Essay Loop engineering is the meme. Outcome engineering is the job. Michael Quoc July 22, 2026
Essay AI automation is bad for business. That’s why I’m building my company around AI. Michael Quoc July 1, 2026
Industry Report The 2026 Trust in AI Commerce Report Product.ai Research June 23, 2026
Essay Designers aren't becoming engineers. They're becoming systems thinkers. Michael Quoc May 7, 2026
Prediction Study Skincare Prediction Study, 2026–2027: Seven Calls on Where Skincare Science and Market Hype Diverge Jessica Ho April 29, 2026
Essay Claude Mythos proves that AI safety is an offensive weapon, not a constraint. Michael Quoc April 8, 2026
Essay Why does ChatGPT give you generic answers instead of insight? Michael Quoc March 24, 2026
Practitioner Account Why Every AI Shopping Assistant Sounds the Same — and What It Took to Build One That Doesn't Jessica Ho March 9, 2026
Essay 60% of promo codes are dead. You pasted one at checkout last week. Michael Quoc March 5, 2026
Whitepaper Anatomy of a Verdict: How SimplyCodes Adjudicates Truth in a Broken Economy Michael Quoc, Bri Stanback March 4, 2026
Whitepaper Axiomatic Intelligence: A Post-Probabilistic Architecture for the Age of Noise Michael Quoc March 4, 2026
Practitioner Account Forging Brand Identity Through Adversarial Collision: Product.ai's AxI-Driven Design System Elena Madrigal, David Guzman March 4, 2026
Essay Axiomatic Intelligence: When the AI cannot afford to guess. Michael Quoc February 17, 2026
Essay The fiduciary diode: Engineering neutrality in the age of slop. Michael Quoc February 2, 2026
Essay The beige singularity: How the web poisoned AI—and AI is poisoning it back. Michael Quoc January 17, 2026
Essay From benchmarks to reality: Decoding the new Gemini 3, Claude 4.5, GPT 5.1 - what's really changed? Michael Quoc December 8, 2025
Essay The glass box mandate: why the future of AI is "ugly." Michael Quoc December 1, 2025
Essay The 10 immutable laws of context engineering. Michael Quoc November 18, 2025
Essay The operating system for AI-native companies: Building in public. Michael Quoc November 17, 2025
Essay Stop prompting. Start architecting. Michael Quoc November 5, 2025
Essay The age of agents is here. Stop designing chatbots. Michael Quoc November 3, 2025
Essay AI is a relevance engine, not a truth oracle. Michael Quoc October 27, 2025

Published here first. Cite freely.

Every whitepaper, prediction study, and report on this page is published on product.ai before it appears anywhere else. Each article carries formal citation formats — BibTeX, APA, and MLA — so citing us works the same way citing any other primary source does.

A note on The Operator's Codex

The Operator's Codex is Michael Quoc's personal essay series — his own voice, his own byline, his own opinions. It is not Product.ai's institutional research. The whitepapers, prediction studies, and industry reports published under Product.ai Research above are a separate body of work, with their own methodology and review process.

Read The Operator's Codex →
The System

Explore Axiomatic Intelligence.

The methodology behind axiomatic intelligence. How claims become verdicts. How the Truth Graph works.

Explore Axiomatic Intelligence →