scripod.com

Are AI Glasses Over?, Big Technology Audience Questions, Alex Stamos on AI Cybersecurity

In this episode of the Big Technology Podcast, recorded live at the Big Technology AI Summit, host Alex Kantrowitz and Ranjan Roy from Margins dive into the latest tech news, focusing on Snapchat's new AR Spectacles and the broader implications for AI devices. They also explore the challenges of planning in a rapidly evolving AI landscape and the ethical responsibilities of AI companies.
The discussion begins with Snapchat's AR Spectacles debut, which led to a stock drop. Ranjan Roy shares his experience with the 2021 version, praising the AR tech but criticizing the current form factor. The debate centers on whether AI glasses are over, with arguments for and against their success. Voice dictation is explored as a new computing interface, alongside the challenge of planning amid rapid AI changes. The panel discusses AI companies using personal data for advertising, noting the unprecedented depth of data collection. The conversation shifts to US-China AI competition, with the rise of DeepSeek as a cost-effective option. Alex Stamos joins to clarify cybersecurity implications, debating the White House's order for Anthropic to take down its Fable and Mythos AI models. Stamos argues that restricting Fable is pointless since other models can perform similar tasks, and warns that jailbreak-proof AI is impossible, forcing the U.S. AI industry to consider Chinese models as backups.
05:30
05:30
AR glasses trend may be over
15:28
15:28
Agents already work autonomously for many tasks.
20:45
20:45
AI knows far more than traditional search.
45:41
45:41
Attackers only need a few bugs, defenders need full coverage.
53:21
53:21
Jailbreak-proof AI is impossible