GitHub's first metered billing cycle hands developers bills up to $3,000, exposing the true cost of agentic AI — and that's just one of today's signals. From Google DeepMind's talent drain to Meituan's trillion-parameter model claim, the artificial intelligence landscape is shifting fast.
Audio is available on Spreaker — see link below.
A GitHub Copilot bill of seven hundred and fifty dollars. That's what some developers are looking at after the first full month of metered pricing closed.
The talent story at Google DeepMind is getting harder to explain away. At least six senior researchers have left since February, including John Jumper, who co-created AlphaFold.
China's Meituan has open-sourced a one-point-six-trillion-parameter model and claims it was fully pre-trained and run on domestic silicon. If that's verified, it's a meaningful erosion of US export control effectiveness.
The Pentagon has formally launched what it's calling its War Force initiative. Hundreds of early-stage engineers are being recruited for two-year deployments implementing the DoD's AI Acceleration Strategy.
Generative AI hit one hundred and ten billion dollars in revenue this year, scaling three times faster than the early internet did. Token prices are falling roughly ten percent, which is driving twelve to eighteen percent usage growth.
One smaller but telling development: Cursor has entered public beta on iOS. Developers can now manage agents, merge pull requests, and handle incidents from their phones via Live Activities.
Chapter summary auto-generated from the verified script. Listen to the full episode for the complete content.