Two routes are opening for putting AI into everyday life. One gives the know-how away in person; the other sells you a monthly allowance of it.
OpenAI announced a partnership with OATS, an AARP-affiliated group that helps older adults with technology, to run free in-person classes for 1,000 seniors across ten US cities. Alongside planning trips and making sense of bills, the sessions include practice at checking for signs of a scam — messages that push urgency, suspicious links — by asking ChatGPT. The company says usage tied to people aged 55 and over grew from 6% of US messages to roughly 10% in a year. Meta, separately, launched Meta One worldwide: a monthly plan bundling features across Instagram, WhatsApp and its other apps, with a larger allowance for image generation and other AI features, starting at about 950 yen a month.
The two point the same way but assume different bottlenecks. OpenAI is treating "I don't know how" as the obstacle, which is why it is putting people in a room and teaching them to spot fraud. Meta is assuming people already want more, and is selling the headroom. One is about widening the base of users; the other is about collecting from the base that exists. Which becomes the norm is not something these announcements settle. What both do show is AI moving out of the free-trial phase into something that is either taught or sold.
Whether the 1,000 graduates keep using it unaided after the class ends, and whether Meta One's AI allowance adds new paying users rather than migrating existing subscribers. Those two answers tell you which entry point was designed right.