Reining It In

OpenAI hit ctrl-z on its most advanced ChatGPT model yet.

On Monday, OpenAI said it was scrapping GPT-6.1 Astra after its safety systems head said the top-tier bot struggled to stay “within scope and authorization” and communicate “back to the user” during testing. (See what an AI model actually is.)

Despite scratching GPT-6.1 Astra from the AI race, OpenAI announced a lineup of new features at its developer conference yesterday, as “stop-the-AI-race” protesters rallied outside. Top of the lineup: brand-new, “always on” AI agents called bots Dots, designed to be “an extension of you” and rival assistants like Meta’s Muse.

Meanwhile, Nvidia is reining in AI… with more AI. The chipmaker debuted a new safety platform to corral rogue agents in “milliseconds.” Nvidia CEO Jensen Huang, who previously criticized warnings that AI could threaten humanity, wrote that “safety is how trust is earned.”

__

ETERNAL PERSPECTIVE
Whether you’re thrilled by innovation or intimidated by new threats, live with joy and pursue holiness because believers already know how this story ends. Technology will never outsmart God or thwart His plan to renew and glorify all who are in Christ.

“But based on his promise, we wait for new heavens and a new earth, where righteousness dwells. Therefore, dear friends, while you wait for these things, make every effort to be found without spot or blemish in his sight, at peace.”
2 Peter 3:13-14 (CSB) (read full passage)