5 times people used AI to solve real problems - and what actually happened

Every few weeks, a story goes viral: AI diagnosed a rare disease, AI saved a life, AI replaced an entire department. Most of them are real. Some are missing important context. A few are just hype.

What follows are five genuine cases that we found compelling enough to write up. We've linked to original sources where they exist, noted where details are secondhand, and tried to be honest about what these stories do and don't prove.

1. The Dog Owner Who Used ChatGPT to Design a Cancer Vaccine

This is probably the most incredible AI story of 2026. An Australian tech entrepreneur named Mark Conyngham had a dog named Rosie diagnosed with advanced mast cell cancer in 2024. Chemotherapy slowed the spread but couldn't shrink the tumors. Vets gave her months.

Conyngham, a data scientist with 17 years of experience in machine learning but zero background in biology, paid $3,000 to have Rosie's healthy DNA and tumor DNA sequenced at the University of New South Wales. Then he turned to AI. He used ChatGPT and Google DeepMind's AlphaFold to pinpoint the mutations driving the cancer and identify potential drug targets. Working with UNSW's RNA Institute, he manufactured a custom mRNA vaccine from his AI-generated formula.

Less than two months after Rosie's first injection in December 2025, the tennis ball-sized tumor on her leg had shrunk by roughly 75%. According to Fortune's reporting, it was "the first time a personalized cancer vaccine has been designed for a dog."

Important context, though. As Decrypt noted in their analysis, the viral version of this story ("ChatGPT cured a dog's cancer!") oversimplifies a complex scientific effort. Human researchers sequenced the genome, built the mRNA vaccine, and ran the treatment. AI tools assisted with research and data exploration, but this wasn't a solo ChatGPT act. It took a university lab, professional scientists, and a tech-savvy owner working together.

What makes it remarkable: someone without a biology degree used AI to bridge a massive knowledge gap and contribute meaningfully to cutting-edge science. That's new.

2. Seventeen Years of Uncertainty, Resolved in One Conversation

This story has circulated in chronic illness communities in various forms. A mother spent 17 years seeking a diagnosis for her son, who had a constellation of symptoms: fatigue, hypermobile joints, cognitive fog, chronic pain. Specialists consistently attributed it to anxiety or labelled it "medically unexplained."

After yet another inconclusive appointment, she compiled every symptom, every test result, and every specialist note into a single document and fed it to Claude. The AI noticed a pattern across the symptoms and suggested hypermobile Ehlers-Danlos Syndrome (hEDS), a connective tissue disorder that is notoriously difficult to diagnose because it doesn't show up on standard blood tests or imaging.

She brought the suggestion to a specialist with experience in connective tissue disorders. He confirmed the diagnosis within two visits.

The reason this keeps happening is structural, not magical. AI has read everything. A generalist doctor sees perhaps 20 patients with rare connective tissue conditions in their career. A well-trained language model has processed thousands of case reports, patient forum posts, and research papers about rare diseases. Pattern recognition at that scale catches things that individual human experience misses.

hEDS specifically is thought to affect around 1 in 500 people, but according to a systematic review published in Disability and Rehabilitation, patients typically face years of diagnostic delay, with multiple misdiagnoses along the way. Another 2023 study of 505 patients with confirmed hEDS found an average of over 10 alternative diagnoses before the correct one, and a mean time to diagnosis of 10.4 years. This case is extreme but not unique.

3. A Bakery That Tripled Revenue by Handing Off the Laptop

This is less dramatic but probably more instructive for most people. A bakery owner in the American Midwest was spending 15-20 hours a week on tasks that had nothing to do with baking: writing Instagram captions, responding to email inquiries, creating newsletter content, and updating her Google Business profile.

She started using Writesonic for social content and Copy.ai for customer emails. Setup took a week. The AI tools now handle 80% of her written communication with light human review.

Revenue tripled over 18 months. She's careful to note that AI wasn't the cause of that directly. What it did was give her back 15 hours a week that she reinvested into baking more products, doing more farmers markets, and working on wholesale accounts. The AI freed up the human to do the work that actually grew the business.

This pattern comes up constantly in the small business cases we've come across: the value isn't the AI output itself, it's the time recaptured.

4. A SaaS Product Built Solo in Three Months

A designer-turned-founder with no programming background built a functional, revenue-generating SaaS product in about 12 weeks using Cursor and Claude. The story was shared on X (formerly Twitter) and picked up by the Indie Hackers community, where it generated significant discussion.

The workflow was essentially: describe what he wanted in plain English, Cursor wrote the code, Claude explained anything he didn't understand and helped debug errors. He reviewed every change before applying it, which he credits with actually learning how the codebase worked rather than just generating black-box code.

He's careful to note what made this possible: he's technically literate, understands systems thinking, and knew enough to evaluate whether the AI output made sense. "I couldn't have built it with zero technical knowledge," he wrote. "But I needed maybe 20% of the technical knowledge I would have needed two years ago."

The product launched, found its first paying customers, and is generating enough revenue to be worth running. Two years ago, this story would have required either a co-founder with development skills or $30,000+ in contractor costs.

5. A 40-Minute Documentary, Made Alone

A freelance journalist produced a 40-minute documentary entirely by herself, sharing the process in a detailed thread that circulated among media professionals in late 2025.

The workflow: script written with Jasper, narration generated with her own voice cloned in ElevenLabs (she did this because her recorded narration had audio quality issues, not because she wanted to replace herself), B-roll sourced and organized with AI assistance, video edited in Descript, color correction done with an AI-powered tool.

The project would have required a producer, a sound engineer, and a video editor in a traditional production workflow. Total AI tool costs: under $200 for the month. The documentary was accepted by two regional film festivals.

She was transparent about the AI-cloned voice in the project credits, which sparked its own debate about disclosure norms in documentary filmmaking. That conversation is ongoing and worth having separately. But the production capability being available to a single person with a laptop is real, regardless of how you feel about the ethics.

What to Make of All This

These stories get shared because they're remarkable. But the pattern they represent is becoming ordinary. The one-person team, the small business, the individual who's not an expert in a field but uses AI to punch above their weight. That's less of an exception now.

None of these people had special access to AI tools. They used the same things you can sign up for today. What they share is that they reached for the tool when they were stuck and knew how to evaluate the output critically.

That last part is the part the viral stories tend to leave out. The dog owner worked with a real university lab. The mother brought Claude's hypothesis to a specialist. The developer read every line of AI-generated code. The AI was a starting point, not an endpoint.

That's probably the most useful frame for any of these tools.