I didn't notice the scarlet tanager until the alert appeared on my phone: "Merlin heard a new bird!" Despite its brilliant plumage—jet-black wings on a crimson body—the songbird can be hard to spot in a forest canopy. It sounds a little like a robin to an untrained ear.

The free Merlin Bird ID app detected a scarlet tanager nearby by using artificial intelligence to analyze my phone's live sound recording. I paused my hike, scanned the treetops, saw the bird singing, and clicked a button to add the species to my growing "life list." Digital confetti burst on my screen.

Like a real-world version of Pokémon Go, the drive to add to my Merlin list has helped me find a great kiskadee in Mexico and a rusty-cheeked scimitar-babbler in the Himalayas. But sometimes the greatest revelations are close to home, as more AI nature app users are discovering.

Are You a Bird Person Now?

"Our stereotypical demographic five years ago would have been retired people and avid birders," said Drew Weber, manager of the Merlin app at the Cornell Lab of Ornithology. "Now we're seeing a lot of 20-and-30-year-olds posting on TikTok or Instagram."

"Am I a bird person now?" exclaims one incredulous TikTok user whose Merlin app detected a tufted titmouse, a cardinal, and a Carolina wren within five seconds.

Even NFL quarterback Sam Darnold shared his experience: "That was a northern mockingbird," he laughed, holding his phone while sitting outdoors.

The app isn't perfect—mockingbirds mimic other sounds and can confuse the AI. Was that really a great horned owl you heard? Maybe, maybe not. Low frequencies like car noise can trick the model.

AI Nature Apps Are Not Just for the Birds

Built-in computer vision on modern smartphones makes it easy to identify plants and creatures. But accuracy isn’t always great for obscure species—and community-driven apps like Merlin and iNaturalist add value through crowdsourced expertise.

Every observation in iNaturalist and Merlin contributes to conservation research as biodiversity loss accelerates. "Our strategy is to build a community of passionate, engaged nature stewards," said Scott Loarie, iNaturalist’s executive director.

Submit a wrong ID, and experts will often correct you. Once consensus is reached, your observation becomes "research grade."

How Merlin's Photo ID Technology Works

The machine learning model behind Merlin’s Photo ID is a convolutional neural network, trained on millions of annotated bird photos from the Macaulay Library. The latest version has 95% average accuracy and far fewer errors than before. The massive dataset makes it uniquely powerful among birding apps.

Don't Use AI Apps to Pick Strange Berries (But I Did)

On a hike in Wyoming, I used iNaturalist to identify berries: saskatoon serviceberry, thimbleberry, and Greene's mountain-ash. Two were sweet, the last bitter. But experts caution: don’t rely on apps to confirm edibility.

"You should never trust automatic ID for edible plants," said Loarie. Still, apps help users learn and avoid harmful species like poison ivy or invasive lanternflies.

Tips for Making the Most of Nature Apps

• Take multiple angles of your subject
• Include key details: flowers, leaves, stems, insect top/side views, mushroom caps and gills
• Use good lighting and focus
• Include an object for scale

These apps aren’t just about ID—they’re about building communities of enthusiasts and contributing to conservation. Using Merlin or iNaturalist, you’re part of a global effort to protect biodiversity.