Radhika Rajkumar
@radhikaraj.bsky.social
editor, news & features @zdnet.bsky.social
covering synthetic media, gen AI, health tech
work in SPICY Zine, Refinery29, Surface Mag
she/her | radhika.rajkumar@zdnet.com
covering synthetic media, gen AI, health tech
work in SPICY Zine, Refinery29, Surface Mag
she/her | radhika.rajkumar@zdnet.com
AI has only picked up speed this year, infiltrating most parts of daily life for some -- all while US AI policy remains unfinished at best.
Trump will announce his AI Action plan tomorrow. Here's everything his second admin has done thus far. @zdnet.bsky.social
www.zdnet.com/article/how-...
Trump will announce his AI Action plan tomorrow. Here's everything his second admin has done thus far. @zdnet.bsky.social
www.zdnet.com/article/how-...
How the Trump administration changed AI: A timeline
On Wednesday, the Trump administration is expected to announce its AI policy. Here's how we got here - and what might happen next.
www.zdnet.com
July 22, 2025 at 7:01 PM
AI has only picked up speed this year, infiltrating most parts of daily life for some -- all while US AI policy remains unfinished at best.
Trump will announce his AI Action plan tomorrow. Here's everything his second admin has done thus far. @zdnet.bsky.social
www.zdnet.com/article/how-...
Trump will announce his AI Action plan tomorrow. Here's everything his second admin has done thus far. @zdnet.bsky.social
www.zdnet.com/article/how-...
Expect the totally expected (latest for @zdnet.bsky.social)
Where AI educators are replacing teachers - and how that'll work
Unbound Academy is introducing a virtual school-without-teacher model. A handful of human 'guides' are standing by just in case.
www.zdnet.com
January 2, 2025 at 8:03 PM
Expect the totally expected (latest for @zdnet.bsky.social)
New testing reveals that OpenAI's latest model, o1, is particularly adept at subverting, deceiving, and lying to developers -- what evaluators call "scheming" -- and that is does so knowingly. @zdnet.bsky.social
www.zdnet.com/article/open...
www.zdnet.com/article/open...
OpenAI's o1 lies more than any major AI model. Why that matters
Several popular models 'readily engage' in deceptive tactics, researchers find. See which was the last to confess.
www.zdnet.com
December 10, 2024 at 8:02 PM
New testing reveals that OpenAI's latest model, o1, is particularly adept at subverting, deceiving, and lying to developers -- what evaluators call "scheming" -- and that is does so knowingly. @zdnet.bsky.social
www.zdnet.com/article/open...
www.zdnet.com/article/open...