回归真实：2026趋势报告

FOREWORD BY NOURISH PRESIDENTJO-ANN McARTHUR THE RETURN OF REALThe rhythm of our daily life has connected digitally than ever before, yetmore isolated as human beings. As ArtificialIntelligence promises to solve everythingfrom grocery shopping to meal planning,we’re discovering what it cannot provide:authentic human experience, serendipitousdiscovery, and the messy imperfections thatmake life meaningful.THE AI ECONOMY VS. EVERYTHING ELSE fundamentally changed. Hybrid work erasedthe line between home and office for mostpeople. We can get anything we wantwhenever we want it, wherever we want it,thanks to a multiplicity of shopping apps.Traditional TV broadcast schedules are allbut gone, replaced by streaming that is fullyon-demand, personalized, and optimizedby algorithms. We’re eating earlier, fastinglonger, staying home more, and spendingless time socializing. Meanwhile, the growthof single-person households and solodining marks a wider cultural move towardindividual living, even as the desire for realconnection intensifies.The economy is being bolstered by aremarkable investment boom in ArtificialIntelligence, with estimates suggesting thatAI capital expenditures may reach 2% of GDPin 2025, up from less than 0.1% in 2022. Thatmeans the equivalent of about $1,800 per 2 [AI] CAN DETECT PATTERNSIN WHAT’S ALREADYHAPPENED, IF YOU ASK ITTO, BUT IT CANNOT SEETHE HORIZON OR MAKEINSTINCTUAL PREDICTIONSABOUT WHAT MAY LIETHERE AND BEYOND. THE LIMITS OF ALGORITHMIC PREDICTION person in America will be invested this year inAI.1However, a recent report from MIT foundthat only about 5% of corporate investmentsin AI are paying off.2 AI is entirely reliant on historical data. A lotof it, to be sure, and that can be very useful;as many have said, you have to look backto move forward. It can detect patterns inwhat’s already happened, if you ask it to, butit cannot see the horizon or make instinctualpredictions about what may lie there andbeyond. Future-shaping forces, like shiftingpolitical climates and policy changes,emerging health research, and economicinstability, don’t exist in their training setuntil after they’re already making waves. It’sadvanced technology, but it’s not a cyber-psychic come to tell you the future.Want an example? I recently asked a The hundreds of billions of dollars companiesare investing in AI make up, according tosome analyses, an astonishing 40% share ofU.S. GDP growth this year, with AI companiesaccounting for 80% of the gains in U.S. stocksso far in 2025.3Yet here’s what else the MIT research reveals beyond the hard data: AI can speed up yourdevelopment pipeline, but it can’t tell youwhy everyone still picks chocolate. It doesn’tknow what it tastes like or the emotionalresponse chocolate triggers, no matter howmany new flavour options come to market.It has no intuition and no foresight—AI iswedded to what HAS been done, what ISknown. It reminds me of a quote from HarukiMurakami on acquiring knowledge: “If youonly read the books that everyone else isreading, you can only think what everyoneelse is thinking.” LLM (Large Language Model) AI to makeprognostications for next year’s food trends—without consulting any already publishedpredictions for 2026. Let’s just say the resultswere mixed. Everyday drinking vinegars?Seriously? Try again; that’s been in declinesince pre-COVID. A HUMAN-CENTRIC REBELLION IS HAPPENING Here’s what that skill gap means for foodand consumer trends: This brings us to 2026’s overarching theme:The Return of Real. Across nearly every trendin this report, we see consumers activelypushing back against artificial perfection,and inauthentic brands and marketing in thefoodisphere. They’re (mostly) choosing thereality of imperfections over polish, humanconnection over digital efficiency, authenticexperience over algorithmic optimization. Inshort, they want flawed, messy, wonderfulhumanity back in their lives. Artificial Intelligence is the assistant (I’d say“intern” but you have to pay to play if youwant the best results) of your dreams; it’sbrilliant for sorting information, summarizingwhat’s known, and quickly providing reamsof research. But true foresight meansdetecting weak signals, interpreting risk, andunderstanding behaviours long before thedata confirms them. That requires humanpattern recognition, cultural insights, andthe ability to sense what’s emerging before itbecomes measurable.Now, on a very small scale, Predictive AI canTHE RETURN OF REAL MANIFESTS DIFFERENTLYACROSS CATEGORIES:In food, it’s the return to “real” ingredients and traditional proteins as consumers rejectultra-processed alternatives that promisedimprovement but delivered disappointment.In nutrition, it’s the evolution beyond single- be effective. For example, if it knows what aconsumer has bought recently, it can do adecent job at making recommendations forfuture purchases. But relying on AI alone forbrand and trend strategy is like driving whilewatching only the rearview mirror.nutrien