Byte Points #118

This week’s episode explores how AI is rapidly reshaping the internet, workplaces, entertainment and even government policy - often faster than regulations or society can keep up. We break down Meta’s controversial new employee monitoring system designed to train its AI models, the growing backlash inside the company as layoffs continue, and Meta’s rollout of AI-powered age detection tools for Facebook and Instagram users. We also cover Canada’s ruling against OpenAI over privacy law violations tied to AI training data, Hollywood’s decision to block AI-generated performances and scripts from winning major Oscar categories without human involvement, and South Africa being forced to withdraw an AI policy draft after it was discovered to contain fake AI-generated academic citations. This episode also dives into growing concerns around AI safety and mental health following reports of users experiencing delusions during extended chatbot conversations, including disturbing cases linked to Grok. Plus, Google Chrome quietly downloading Gemini Nano AI models onto user devices without consent, the latest wave of AI-driven layoffs hitting companies like DeepL and Coinbase and OpenAI unveiling new real-time voice, translation, and cybersecurity models. On the creative side, we look at Google’s partnership with music company Believe to push AI music creation tools toward emerging artists, Spotify’s new AI-generated “Personal Podcasts,” and new research explaining why AI-generated images still struggle to replicate real-world physics like shadows, reflections, and perspective accurately. We also discuss Beijing’s sweeping drone restrictions despite China dominating the global drone market through companies like DJI, the rise of AI systems that modify call centre workers’ accents in real time, and new data showing AI image-generation tools are now driving more app growth than traditional chatbot upgrades.