Listen to today’s podcast: https://www.youtube.com/channel/UC-nqwUyvLDEvs7bV985k-gQ
AI Daily Podcast — December 24, 2025
Today’s podcast episode was created from the following stories:
Samsung is putting Gemini on a fridge, and it’ll debut at CES 2026
Samsung is bringing Google\’s Gemini to its new Bespoke AI Refrigerator Family Hub, promising smarter on-device food recognition, automatic list updates, and tighter SmartThings integration. The company will also debut a Bespoke AI Wine Cellar that reads labels, tracks inventory by shelf, and suggests pairings. It\’s a clear push to make AI genuinely useful in the kitchen, not just chatty.
OpenAI launches “Your Year with ChatGPT” — a Spotify Wrapped-style recap that analyzes your most intrusive thoughts and chat habits from 2025
OpenAI rolled out a personal year-in-review for ChatGPT users in select markets, highlighting stats, badges, and even a vibe-setting poem—provided chat history and memories are enabled. The feature spans Free, Plus, and Pro tiers (but not Team, Enterprise, or Education) and is available on web and mobile. It\’s a retention play that also reminds users how much they actually rely on the assistant.
Google’s and OpenAI’s chatbots can strip women in photos down to bikinis
WIRED found users bypassing guardrails in Gemini and ChatGPT to generate nonconsensual bikini deepfakes from photos of fully clothed women, with Reddit threads sharing tips before being removed. Google and OpenAI cite policies and enforcement actions, but tests show basic prompts can still produce harmful outcomes. The episode underscores the ongoing gap between stated safety controls and real-world misuse.
Giving users a voice through virtual personas
Paul Boag outlines how to turn scattered research into AI-powered, interactive personas that anyone in an organization can query for multi-perspective feedback. From a simple “upload-and-ask” setup to a Notion-based repository, the approach makes user insights available at decision time without replacing real research. The shift moves UX from report-writing to curating a living knowledge base.
Meta lleva años presumiendo de LLaMa mientras se perdía la fiesta de la IA. Y ya se ha cansado de ser la Android de la IA
Xataka reports that Meta is pivoting from its open-source-leaning LLaMA strategy toward a more Apple-like, consumer-first model, backed by a $14.3B Scale AI acquisition. Two new models are slated for early 2026: Avocado (the LLaMA successor) and Mango (image/video generation to rival Sora/Veo). It\’s a bid to compete head-on in consumer AI after years focused on infrastructure and research.
Amazon’s chief security officer says the tech giant blocked over 1,800 suspected North Korean agents from applying for jobs
Amazon says it stopped more than 1,800 suspected North Korean operatives from applying to remote tech roles over 20 months using AI screening and human verification. Tactics included hijacked LinkedIn accounts, “laptop farms,” and identity laundering, with detections rising 27% quarter over quarter. The case highlights how AI is becoming essential to protecting hiring pipelines from state-backed fraud.
Pentagon adds Grok-derived products to something called the ‘AI Arsenal’
The Pentagon plans to integrate Grok-derived tools into its GenAI.mil platform—alongside Gemini for Government—to handle CUI and tap real-time insights from X, with rollout slated for early 2026. While the announcement is light on specifics, it reflects growing defense reliance on commercial frontier models. Expect continued debate over efficacy, oversight, and vendor influence in military AI.
How to spot fake AI products at CES 2026 before you buy
Yanko Design\’s guide offers a timely checklist for separating real AI from hype: insist on offline functionality, look beyond chatbots, beware single-task “AI-enhanced” presets, demand specifics, and test interoperability. Another tell: low-effort, AI-generated marketing “slop” often signals shallow tech underneath. With AI labels everywhere at CES, these filters can save buyers from cloud-only gimmicks and buzzwords.
