1 Billion People Drink Too Much – 100K Chose This App

01 /The Concept

Addiction treatment has a reach problem. Clinical programs work – but most people who need them never show up. Sunflower is trying to close that gap: a sobriety app designed to rewire the neural patterns behind addictive behavior, meeting people where they are rather than where a clinic is.

The emotional journal is the core tool. The underlying logic: people often reach for alcohol or substances as the easiest way to manage difficult emotions. The app offers a different – equally accessible – outlet for processing those feelings.

Anyone who's watched American TV dramas has seen characters fighting addiction who, when they're on the verge of giving in, call or meet with their "sponsor" – someone who has already been through recovery and can offer moral support.

Sponsors are a core element of therapy in recovery communities like Alcoholics Anonymous. But reaching a sponsor in the middle of a craving – which typically arrives late at night – often isn't possible. And that's when relapses happen.

For those moments of crisis, the app offers an AI sponsor: available 24/7, accessible as many times as needed, and free of the social discomfort of sharing every difficult thought with a real person. Full honesty gets both support and practical guidance in return.

Beyond that, the app features a visual sobriety tracker displaying consecutive sober days alongside achievement badges – a motivational mechanic that mirrors what Duolingo has demonstrated works for habit formation.

There's also an anonymous community where users can get advice and encouragement from others dealing with the same dependencies.

And a library of video lessons and interactive exercises grounded in evidence-based approaches to addiction recovery.

Basic functionality is free; some premium features require an additional purchase.

Sunflower was founded last year and raised its first $1M in funding before being accepted into Y Combinator. The plan now is to build a full online clinic for addiction recovery – with the app serving as the top of the funnel that brings patients in.

02 /Why It Matters

The Sunflower founder makes a market structure argument: there is already significant demand for addiction recovery – but not enough supply of effective solutions. More than 1 billion people worldwide, the startup claims, struggle with alcohol or drug dependency.

The only way to reach that scale of demand is through AI – which is exactly what the startup is building. As evidence, the founder points to the app's own numbers: monthly active users grew from 200 to 100,000 in six months, with users coming from nearly every country in the world.

The 1-billion-person market claim prompted some fact-checking – starting with US statistics, which proved eye-opening.

According to official data, 16.7% of Americans aged 12 and older – roughly 48.5 million people – qualify as drug-dependent, a category that includes marijuana use.

Alcohol dependence, by official definitions, affects approximately 10% of the total population – nearly 28 million people. There's overlap between the two groups, so the numbers can't simply be added together.

Separately, 21.7% of Americans aged 18 and older – nearly 60 million people – reported binge drinking at least once in the previous month in 2024.

And the National Institute on Alcohol Abuse and Alcoholism has flagged a rising trend in "high-intensity drinking" – defined as consuming at least twice the binge-drinking threshold, which is essentially blackout territory.

The potential demand for recovery solutions – even in a single country – is thus genuinely substantial. The US has approximately 30,000 residential treatment centers for alcohol and drug addiction. Given the scale of the problem, that's clearly not enough.

Those centers are also expensive, which puts them out of reach for lower-income populations – who, as a rule, are more numerous and more vulnerable to addiction than higher-income groups.

The market for addiction recovery continues to grow even with its limited reach. In 2024, it was valued at $9.44B; by 2034, it's projected to reach $16.22B.

Far cheaper alternatives – AI apps, online clinics – have the potential to expand the market by multiples, or potentially an order of magnitude.

03 /Opportunities

Sunflower is a textbook example of one of the most promising categories for AI: making something people didn't previously use because it was too expensive or too difficult accessible to a mass market.

The "too difficult" dimension may be the more powerful of the two. Imagine a typical middle-class professional who comes home after a hard day and opens a bottle of wine to unwind – and does so almost every evening. By clinical definitions, that person qualifies as alcohol-dependent. Are they going to show up to an Alcoholics Anonymous meeting or check into a residential clinic? Absolutely not.

Most people in that situation understand, at least somewhere in the back of their minds, that the habit is a problem. But publicly labeling themselves as an alcoholic isn't something they're ready to do. Neither is spending time on group meetings or structured treatment programs. But using an AI app? Why not.

For many of those people, "drink less" is also a much more achievable goal than "stop drinking entirely." That's the exact angle Sunnyside ([related review](/review/nado-menshe-pit)) took – raising around $15M for an app that helps users "drink less without the pressure to quit altogether."

Surprisingly, that approach works. Despite many clinicians arguing that complete abstinence is the only real path to recovery and that "one drink will become ten," 96.7% of Sunnyside users reduced their drinking – and 90% of those stopped binge drinking.

The bottom line: roughly 1 billion people meet the formal criteria for alcohol or drug dependency. The vast majority of them are unreachable through traditional recovery channels – either because the methods don't fit their lives or because the cost is prohibitive.

AI can finally bring that majority into reach. And that majority might just be the launchpad.