AI Faces the Music, the Classroom, & the Home, but gets NSFW

Aug 28

Written By Katie Michael

ElevenLabs Expands Into AI Music

ElevenLabs, best known for its AI voice tools, has introduced a new music generation model that allows users to create commercial-use music. This marks the company's first major move beyond text-to-speech. Unlike some competitors currently facing legal trouble, ElevenLabs has taken a licensing-first approach. Its agreements with Merlin Network and Kobalt Music Group—whose catalogs include artists like Adele and Childish Gambino—ensure the model is trained only on opt-in music. A Kobalt spokesperson emphasized the deal provides “new revenue streams, revenue sharing, and safeguards against misuse.” This contrasts sharply with rivals Suno and Udio, both of which are currently being sued by the RIAA for allegedly using copyrighted content without permission. As AI-generated music faces increasing scrutiny, ElevenLabs’ rights-cleared model may offer a more sustainable path forward.

ChatGPT 4o Reinstated After User Backlash

OpenAI has reversed course on its decision to fully replace GPT-4o with GPT-5 after users criticized the new model for being less personable. Despite GPT-5's technical upgrades, many ChatGPT users described its tone as “colder,” sparking widespread calls for GPT-4o’s return. Within 24 hours of launching GPT-5, CEO Sam Altman confirmed that GPT-4o would be reinstated for Plus subscribers. The company noted that GPT-4o's continued availability will depend on usage metrics. The quick turnaround underscores how personality and emotional tone are now central to user satisfaction with AI tools. OpenAI’s move may also signal a future where users can choose from a range of models tailored to different communication styles.

Google’s AI Summaries Under Fire

AI-generated summaries on Google Search are drawing concern from media experts, who argue the new feature undermines both traffic to news sites and the accuracy of public information. Since the introduction of Google’s “AI Overview,” publishers have reported sharp drops in referral traffic. A Pew Research study showed that click-through rates fell from 15% to 8% when AI summaries were present. Experts like Jessica Johnson of McGill University and Chirag Shah from the University of Washington caution that these summaries often miss context and can’t reliably interpret jokes, satire, or outdated info—leading to errors like recommending glue on pizza. While Google claims traffic levels remain stable, many in the publishing world describe the situation as a “one-two punch” following Meta’s news ban in Canada. The concerns are particularly acute in Canada, where policymakers are still implementing the Online News Act amid growing tension with tech giants.

Apple’s AI Push Targets Smart Homes

Apple is investing heavily in AI-powered smart home products, according to a Bloomberg report. Among the new devices in development are a Siri-enabled tabletop robot—likened to a Pixar-style lamp—and an animated, visually enhanced Siri interface that could take the form of a Memoji or Finder icon. Apple is also developing a wheeled home robot and exploring the potential for humanoid designs. A smart home display, expected by mid-2026, will offer music controls, video calling, and facial recognition for personalized content. The initiative includes a suite of home security hardware, suggesting a major new ecosystem is on the horizon. Apple’s use of large language models aims to enable natural voice interaction, potentially rivaling devices like Amazon’s Astro and Google Nest Hub.

AI in Classrooms: Teachers Lead the Way

Across the U.S., educators are integrating AI into classrooms, reshaping how they teach and learn. Teachers like Jerome Ong, a 5th-grade educator in New Jersey, are using AI tools weekly to customize lessons and explore the technology’s strengths and weaknesses with students. A major training initiative—backed by Microsoft, OpenAI, and Anthropic—has pledged $23 million to help the American Federation of Teachers train hundreds of thousands of educators in AI literacy. Advocates say AI saves time and enhances lesson planning, with some reporting up to six hours saved per week. However, critics like former teacher Lois Weiner argue that real educational reform must focus on better working conditions, not just tech tools. CBS News reports that from 2023 to 2025, schools have shifted from banning AI to embracing regulated, responsible use across districts nationwide.

Grok Imagine Launches With NSFW Option

Elon Musk’s xAI has unveiled Grok Imagine, a new image and video generator embedded in the X iOS app, currently free for a limited time. Normally available to SuperGrok and Premium+ subscribers, the tool can turn text or images into 15-second videos with audio. A controversial “spicy mode” allows nudity and other NSFW content, though limits remain on generating explicit content involving celebrities. Critics note that the platform still struggles with the uncanny valley, but users have praised its speed and user-friendly interface. Musk says Grok Imagine will “get better every day,” as xAI positions itself against competitors like OpenAI, DeepMind, and Runway. The addition of auto-generation tools and broader creative options indicates a growing emphasis on multimedia AI tools within social platforms.

Sources

OpenAI’s ChatGPT Agent: Power, Price & Pushback