AI is the future and is gradually changing every aspect of society, from daily life and business to creativity and science. Whether that’s a good or bad thing is debatable, but one thing is certain: the demand for AI is skyrocketing. The companies behind AI are rapidly expanding their data centers to meet this demand, which could lead to widespread shortages of RAM chips and make the next smartphone more expensive. These companies also need high-quality data to help train their models, and whether they realize it or not, they’re giving away their data for free. Some ways to do this may even surprise you.
You probably know how your device usage and online activity helps train AI. Interacting with public content on social media provides algorithms that decide what to display next. Modifying Siri, Alexa, and Google Assistant allows developers to improve the AI’s speech recognition patterns. Whether you use autocorrect or predictive text in Google Docs or Microsoft Word, you’re teaching AI patterns that lead to more accurate grammar and spelling suggestions. But you may not know that solving online puzzles, playing augmented reality (AR) games, and even reporting spam can be used to train AI.
Solving CAPTCHAs helps train self-driving cars
CAPTCHA (Completely Automated Public Turing Test to Distinguish Computers from Humans) ensures that only humans, and not bots, can access a given website. This is done through tests that would be easy for a human but difficult for a computer to solve. Common CAPTCHAs include checking the “I’m not a robot” checkbox, identifying distorted text, and selecting specific objects in an image. This is the latest option where AI companies have revealed surprising ways to train self-driving cars.
Google has its own version of CAPTCHA called reCAPTCHA. You’ve probably seen reCAPTCHA v2, a more advanced, safer and less annoying version of CAPTCHA. This version displays a photo grid and asks you to select objects such as cars, bikes, street lights, sidewalks, and storefronts. But according to an article in X by @Sharble, a co-founder of an AI-focused marketing agency, these images are taken directly from Google Street View, and each time you click on an object correctly, you’re training Google’s computer vision model to identify them in the real world.
Much of that data goes to Waymo, a company that develops self-driving technology and is owned by Google’s parent company, Alphabet. Waymo announced in a blog post published in early February 2026 that self-driving operations will become fully autonomous with 6th generation Waymo Driver technology and will expand to more cities in the United States. You helped make this happen, and Google didn’t have to report you or even pay you.
Playing Pokémon Go trains delivery robots
When Niantic released Pokémon Go for Android and iOS devices in July 2016, fans had more Pokémon than ever before. It’s a GPS and augmented reality game where people venture into the real world to catch Pokemon. According to the Pokémon Company, more than 500 million people around the world downloaded the game within two months. In May 2025, Niantic was acquired by mobile gaming company Scopely and continues to develop it. However, Niantic hasn’t completely killed off Pokémon Go, and is using the game’s data in its AI spinoff, Niantic Spatial.
Over the past decade, millions of players have tried to catch Jigglypuff and Charizard, and Niantic Spatial’s AI has helped them build an “inch-perfect view” of the world, according to MIT Technology Review. The company used images to tag landmarks, allowing AI to create ultra-accurate models of the world. Using this model, the AI can determine your location with near-perfect accuracy using just a few photos of surrounding landmarks and buildings.
The technology will now be used in a partnership with robotics startup Coco Robotics to deploy delivery robots in the US and Europe. It helps robots navigate locations where GPS can be unreliable, such as in high-rise buildings, under overpasses, and on highways. At some point in the not-too-distant future, someone will have pizza and groceries delivered to them by robots, and Pokémon GO players will thank them.
Reporting spam helps train email filters
Spam is at best a daily nuisance that clutters your inbox, but at worst it’s dangerous and can leave you the victim of a phishing scam. If you receive an email that is clearly spam, you’ll probably ignore it, delete it, unsubscribe, or send it to your spam folder. However, the best way to deal with spam is to report it. Almost all major email providers now have this feature. This trains the spam filter’s machine learning algorithms to better recognize and block such messages in the future.
An added benefit of reporting spam is that future emails from that sender will be automatically filtered out. Additionally, what the AI learns can also help protect others from email phishing scams. According to a report by the National Consumers League (NCL), phishing and identity theft scams increased by 85.6% between 2024 and 2025, with losses doubling from $1,000 to $2,060. This significant increase is believed to be due to scammers using generative AI to craft convincing emails, text messages, and even cloned voices. Helping improve the same technology to thwart malicious actors who use AI to deceive people is a service to society.
