AI Morning Brief

September 16, 2026 · The previous day · 5 stories from 4 sources

lead · primary

Google DeepMind Introduces Gemini 3.8 Live and 3.8 Live Extended Thinking

Google DeepMind introduced Gemini 3.8 Live and 3.8 Live Extended Thinking, two new speech-to-speech models. The models are similar in shape to OpenAI's GPT-Live family.

Why it matters: The introduction of Gemini 3.8 Live and 3.8 Live Extended Thinking may improve the capability of speech-to-speech models, enhancing user experience and accessibility.

Google DeepMind →

research · voice

Simon Willison's Weblog Demonstrates Gemini Live Audio

Simon Willison's Weblog demonstrated Gemini Live audio by building a web UI for trying out the new models. The implementation uses no libraries and connects to the Google AI Generative Service WebSocket endpoint.

Why it matters: The demonstration of Gemini Live audio may increase access to speech-to-speech models, allowing more developers to experiment and build applications with the technology.

Simon Willison's Weblog →

research · reporting

Ars Technica Reports on Frontier AI Models

Ars Technica reported on Mozilla's findings that cheap open models can catch up to frontier AI models in capability. The report also noted that paying for frontier AI models can provide a 4-month head start at 5x the cost.

Why it matters: The report may influence the cost and deployment of AI models, as developers and organizations weigh the benefits of paying for frontier models against the potential for cheaper alternatives to catch up.

Ars Technica — AI →

research · reporting

The Verge Reports on Unpopularity of AI and Data Centers

The Verge reported on poll data showing that AI and data centers are incredibly unpopular, with 61% of likely voters surveyed opposing the construction of data centers to power AI tech.

Why it matters: The unpopularity of AI and data centers may impact governance and institutions, as politicians respond to public concerns and potentially implement regulations or restrictions on AI development and deployment.

The Verge — AI →

research · reporting

Meta Launches One Subscriptions with AI Usage

Meta launched its new One subscriptions, which pair standalone app subscriptions with extra AI usage. The subscriptions are available globally, with several tiers for individual users, creators, and businesses.

Why it matters: The launch of Meta's One subscriptions may affect the deployment and accessibility of AI, as users and businesses can now purchase subscriptions that include AI usage, potentially increasing adoption and usage of AI-powered services.

The Verge — AI →

The conversation

  • The growing concerns about AI safety and the need for regulations
  • The increasing adoption of AI models and their potential impact on industries
  • The rising public scrutiny of AI and data centers and its potential effects on governance and institutions

Signals to watch

  • Growing concerns about AI safety
  • Increasing adoption of AI models
  • Rising public scrutiny of AI and data centers

How this is made: Matthew Williamson defines the source mix, weights, and standing editorial rules; collection and AI summarization run automatically once each weekday. These summaries are not essays written by Matthew. Company announcements are labeled as primary sources, commentary remains attributed, and every item links to the complete original work.