Artificial intelligence is pushing the frontiers of science and technology—piloting military drones, diagnosing diseases, and designing enzymes that devour plastic waste. Meanwhile, consumer-facing chatbots are being used by hundreds of millions of people every week, seeding seismic shifts in education, child development, and human connection. TIME looks at 10 ways AI innovations are being used to shape the future.
Vaccine development
The swift development of COVID-19 vaccines had several ingredients: years of mRNA research, immense political will, and an influx of cash. But to shrink a 15-year timeline to just 12 months, scientists also relied on a new arsenal of AI-powered tools, which helped with everything from interpreting the genome to cleaning clinical trial data. Similar tools are now helping accelerate vaccine development. Confronted with a pandemic-potential pathogen, “we could have a new vaccine design within a matter of days,” says Timothy Endy, program lead at the Coalition for Epidemic Preparedness Innovations (CEPI), a private-public partnership advancing vaccine development. CEPI’s goal is to be able to develop and deploy a new vaccine in 100 days. “I think we’re just about there,” Endy says.
Human-machine relationships
Human-machine relationships are becoming mainstream. Millions of people are turning to AI chatbots to play a role that blends friend, therapist, and romantic partner. One AI roleplaying platform, Character.AI, now has 20 million monthly active users—half female, half born after 1997.
“For many people, it’s a fun diversion they get a lot out of,” says Kate Devlin, a professor at King’s College London researching AI and human intimacy. She sees benefit in AI companionship, so long as people understand AI is not sentient. But she notes that vulnerable people—particularly children, or those with mental health issues—may be exploited when sharing their deepest feelings with AI tools that are not designed for therapy. Still, with usage growing, Devlin thinks society is resilient enough to find a way through this period of change. “We’re very good at being human,” she says.
Tracking wildlife
To keep tabs on wildlife, conservationists use a network of motion-triggered cameras in some of the world’s most remote corners. But the torrent of images they capture can take weeks to sift through for a single, noteworthy sighting. Wildlife Insights, an AI tool born out of a collaboration between the World Wildlife Fund, Google, and others, scans millions of images, pinpointing animals with 99.4% accuracy—freeing researchers to do “more important work,” says Abby Hehmeyer, WWF’s program lead. In March, they open-sourced the tool, trained on 65 million images, so conservationists anywhere could use it. “The threats that are facing biodiversity are so great,” Hehmeyer says. “This tool allows us to react at the speed that these threats come at us.”
Education
While educational institutions scramble to contain AI, its adoption is becoming widespread. In July, U.S. teacher unions partnered with OpenAI, Microsoft, and Anthropic on an initiative to train 400,000 school teachers to use AI over the next five years. In August, OpenAI announced a partnership with the Indian government to distribute its AI to schools nationwide. For Rose Luckin, a professor at University College London with deep experience in AI and education, students’ widespread AI use has disrupted systems already failing to prepare them for the future. But with these changes come fresh opportunities, although regulation and an understanding of AI’s limits are necessary to protect vulnerable populations, she says. “The fundamental thing we’ve got to get our head around is what relationship do we want between human and artificial intelligence?” she says.
Circular recycling
More than 350 million tonnes of plastic waste are generated annually. But AI-designed enzymes can “un-bake the cake—turn [plastic] back into the chemical compounds that are used to make it—unlock[ing] infinite recycling applications for all the materials that go to landfill or incineration today,” says Jacob Nathan founder and CEO of London-based Epoch Biodesign. It has designed enzymes for the three major plastic groups using machine learning, which could help break down not just textiles, but packaging, and more. The startup, which spun out of Nathan’s high school project in 2019 and has since raised $18.3 million, will complete its first production-scale facility later this year, which is expected to be capable of processing 150 tons of waste annually, Nathan says.
Medical diagnosis
AI tools that help diagnose strokes, cancers, and other medical conditions are becoming a fixture in healthcare systems around the world. Qure.ai, which uses AI to interpret diagnostic images like x-rays and ultrasounds, is being used at over 4,500 sites in over 100 countries. Viz.ai’s diagnostic algorithms are being used in over 1,800 hospitals across the U.S. and Europe. AI systems also show promise as clinical companions. In a pilot study run by Kenyan primary healthcare provider Penda Health and OpenAI, Penda’s AI system, designed to supplement clinicians’ knowledge and used with thousands of patients, reduced medical diagnostic errors by 16%. With adoption growing, global healthcare is poised to improve.
Children’s toys
AI toys are a nascent but growing market. Companies like Curio sell AI-powered plushies that speak with young children in age-appropriate language. One was designed and voiced by musician Grimes. Others, like Miko 3—an “interactive robot that teaches and listens”—are made to assist learning. In June, OpenAI partnered with toymaker Mattel; their first product is expected to be announced later this year. Without safeguards, these toys could interfere with crucial development windows during childhood, cautions Anne-Sophie Seret, executive director at everyone.AI, a global initiative working with governments and companies to ensure children are protected when using AI. She says standards, guidelines, and proactive engagement with companies designing these toys are essential. “Once a teddy bear is under the tree on Christmas,” she says, “it’s so hard to take it back.”
Mapping Earth
In July, Google announced an AI system that acts like a “virtual satellite”, fusing optical, thermal, radar, climate, and other data to accurately characterize Earth’s terrestrial land and coastal waters. The resulting unified dataset “gives scientists and researchers a globally consistent understanding of the planet,” explains Christopher Brown, senior research engineer at Google DeepMind. The computationally efficient custom maps enabled by this model—already in use by over 50 organizations—can be used by governments, companies, nonprofits, and academics for a range of tasks, from mapping ecosystems and supply-chains to analyzing how the planet is changing over time. Brown is optimistic that, by removing barriers to entry, the technology could lead to “an explosion in earth science.”
Automating factories
While industrial robots have been a feature of factory floors since the 1930s, their “brains” have recently started catching up with the brawn. Xiaomi now operates a so-called dark factory in Beijing—so automated it can run with lights dimmed—producing a flagship smartphone every six seconds with only a handful of technicians overseeing the system. Meanwhile, German technology giant Siemens is pioneering AI-driven industrial copilots that help engineers direct robots using natural language and autonomously identify mechanical issues, driving up to 30% efficiency gains, says chief technology and strategy officer Peter Koerte. He adds that much of AI’s future impact hinges on transforming the industrial world. “After all, it’s the majority of the economy.”
Autonomous military drones
A lost signal—often from sophisticated jamming—used to mean a drone’s demise. But when some of the 117 drones involved in Ukraine’s “Operation Spiderweb” in June went dark, on-board AI took the reins, according to the Ukraine’s Security Service. The operation, which destroyed 12 Russian bombers, is the most “vivid” example of AI’s battlefield capabilities, says Mykhailo Fedorov, Ukraine’s Minister of Digital Transformation, who spearheaded Brave1, a state push to connect Ukraine’s defense and tech industries. He says while today’s drones can follow pre-defined targets, future models are marching towards greater autonomy. “In essence, the situation on the battlefield depends on who wins at this battle of autonomous drones,” Fedorov says.
The post 10 AI Applications Shaping the Future appeared first on TIME.
