Google's New AI Experiment: Disco

Google has introduced a new AI experiment called Disco, powered by Gemini. This tool transforms open browser tabs into custom web applications, creating what Google calls 'GenTabs.'

Main Features:

• GenTabs: Proactively suggests interactive web apps to help with tasks related to your browsing.
• Custom App Creation: Allows users to build their own apps using written prompts.
• Task Assistance: Helps with tasks like studying, meal planning, and trip planning.
• Gemini Integration: Uses Gemini 3 to build custom experiences on the fly using information from your browser and Gemini chat history.
• Refinement: Allows users to refine apps using natural language commands.
• Source Linking: Generative elements link back to original sources.

Availability:

The feature is initially available to a small number of testers through Google Labs. Users can join a waitlist to download the app, starting on macOS.

Future Plans:

Google suggests that interesting ideas developed through Disco may find their way into other larger Google products. GenTabs is noted as the first feature being tested, with more to come over time.

Rivian's AI Assistant Launch

Rivian's AI assistant will launch in early 2026, integrating with all existing EVs in its lineup. This assistant will allow drivers and passengers to control climate settings and other infotainment system tasks. It will also connect vehicle systems with third-party apps using an agentic framework, with Google Calendar being the first third-party app to launch within the AI assistant.

Technical Integration

The AI Assistant will be augmented by Google Vertex AI and Gemini for grounded data, natural conversation, and reasoning. Rivian has developed a model and platform-agnostic architecture branded as Rivian Unified Intelligence (RUI), which includes custom large language models and an orchestration layer to ensure seamless operation of various AI models.

Vertical Integration

Rivian's commitment to vertical integration is evident in its AI & Autonomy event in Palo Alto, California. The company has developed new software and hardware, including a custom 5nm processor built in collaboration with Arm and TSMC, to expand its hands-free driving assistance system. This work has been ongoing for years, with significant updates to the R1T truck and R1S SUV in 2024.

Future Applications

Beyond the AI assistant, RUI will be used for vehicle diagnostics, acting as an expert assistant for technicians by scanning telemetry and history to pinpoint complex issues. This reflects Rivian's push towards becoming more vertically integrated and redefining how apps integrate into cars.

OpenAI's GPT-5.2: A New Frontier in AI

OpenAI has launched GPT-5.2, its latest advanced model, designed for developers and professionals. The model comes in three versions: Instant for routine queries, Thinking for complex tasks, and Pro for maximum accuracy.

Key Features and Improvements

GPT-5.2 excels in creating spreadsheets, building presentations, writing code, perceiving images, understanding long context, and linking complex projects. It sets new benchmarks in coding, math, science, vision, and tool-use.

Competition and Market Dynamics

GPT-5.2 is positioned against Google's Gemini 3, which has been leading in most benchmarks. OpenAI aims to reclaim leadership amidst concerns about declining ChatGPT traffic and market share loss to Google.

Enterprise Focus and Future Plans

OpenAI is targeting developers and the tooling ecosystem, aiming to become the default foundation for AI-powered applications. Despite internal calls for a delayed release, GPT-5.2 is seen as a push to enhance enterprise opportunities.

Challenges and Strategic Moves

OpenAI faces high stakes with significant investments in AI infrastructure. The focus on reasoning models, though costly, is a strategic move to maintain leadership in the AI arms race.

Google's AI Try-On Feature: Virtual Fitting with Just a Selfie

Google has updated its AI try-on feature to allow users to virtually try on clothes using just a selfie. Previously, users needed to upload a full-body picture. The new feature uses Nano Banana, Google's Gemini2.5 Flash Image model, to generate a full-body digital version of the user for virtual try-ons.

Users can select their usual clothing size, and the feature will generate several images. They can choose one to make it their default try-on photo. Users still have the option to use a full-body photo or select from a range of models with diverse body types.

The new capability is launching in the United States today. Google first launched the try-on feature in July, allowing users to try on apparel items from its Shopping Graph across Search, Google Shopping, and Google Images.

Google has been investing in the virtual AI try-on space, with a separate app called Doppl designed to help visualize how different outfits might look on you using AI. Earlier this week, the tech giant updated it with a shoppable discovery feed that displays recommendations so users can discover and virtually try on new items.

The discovery feed features AI-generated videos of real products and suggests outfits based on your personalized style.

Ford and SK On End Joint Venture

Ford and SK On have agreed to terminate their joint venture, which was established four years ago to build factories in Tennessee and Kentucky for producing batteries for electric F-Series trucks.

Division of Assets

Ford will take ownership and operation of the twin battery plants in Kentucky, while SK On will operate the factory at the BlueOval SK campus in Tennessee.

Strategic Partnership

Despite the end of the joint venture, SK On will maintain a strategic partnership with Ford centered on the Tennessee plant.

Industry Context

The joint venture was formed during a period of significant investment in electric vehicle production. However, EV sales have not met industry projections, and the end of federal EV tax credits has further dampened sales.

Runway's GWM-1 World Model

Runway has launched its first world model, GWM-1, which simulates the world through frame-by-frame prediction, understanding physics and real-world behavior. This model is designed to train agents in various domains like robotics and life sciences.

Key Features of GWM-1

GWM-1 includes specific versions: GWM-Worlds, GWM-Robotics, and GWM-Avatars. GWM-Worlds allows interactive project creation with an understanding of geometry, physics, and lighting. GWM-Robotics uses synthetic data for robot training, and GWM-Avatars focuses on simulating human behavior.

Enhanced Video Model

Runway has also updated its Gen 4.5 video model, adding native audio and long-form, multi-shot generation capabilities. This update enables users to generate one-minute videos with character consistency, native dialogue, background audio, and complex shots from various angles.

Future Plans

Runway plans to merge Worlds, Robotics, and Avatars into one model. The company is also in active conversations with robotics firms and enterprises for the use of GWM-Robotics and GWM-Avatars.

Rivian's Autonomy & AI Day Highlights

Rivian unveiled its plans to enhance the autonomy of its electric vehicles, including new hardware like lidar and custom silicon. The company aims to expand its hands-free driver-assistance software to over 3.5 million miles of roads in the USA and Canada, with plans to eventually include surface streets. This feature, called 'Universal Hands-Free,' is set to launch in early 2026 and will be available on second-generation R1 trucks and SUVs.

Rivian's CEO, RJ Scaringe, described a point-to-point navigation feature that allows drivers to input a destination and have the vehicle drive them there completely autonomously. This feature is part of Rivian's goal to reach 'personal L4' autonomy, where vehicles can operate without human intervention in specific areas.

Scaringe hinted at Rivian's potential entry into the ride-hail market, competing with companies like Waymo. The company is also developing a 'large driving model' and a custom 5nm processor in collaboration with Arm and TSMC. This processor powers the third-generation 'autonomy computer,' or ACM3, which can process 5 billion pixels per second and will be featured in the upcoming R2 SUV in late 2026.

Rivian plans to couple the ACM3 with a lidar sensor to provide three-dimensional spatial data and redundant sensing. The company believes this combination will enable a higher level of autonomy in its vehicles, aiming for 'superhuman' sensing capabilities.

Ford and SK On End Joint Venture

Ford and SK On have agreed to dissolve their joint venture, which was formed four years ago to build factories in Tennessee and Kentucky for producing batteries for electric F-Series trucks.

Asset Division

Ford will take ownership and operation of the twin battery plants in Kentucky, while SK On will operate the factory at the BlueOval SK campus in Tennessee.

Strategic Partnership

SK On will maintain a strategic partnership with Ford centered on the Tennessee plant.

Industry Context

The joint venture was created during a period of significant investment in electric vehicle production. However, EV sales have not met industry projections, and the end of the federal EV tax credit has dampened sales pace.

Spoor's AI Revolution: Protecting Birds and Expanding Horizons

Founded in 2021: Spoor launched with the mission to use computer vision to reduce the impact of wind turbines on bird populations.

Technology: Spoor's software tracks and identifies bird populations and migration patterns within a 2.5-kilometer radius using off-the-shelf high-resolution cameras.

Applications: Wind farm operators use this data to plan locations and adjust turbine operations during migration periods.

Founder's Vision: Co-founder and CEO Ask Helseth was driven by the lack of effective tracking methods in the industry, despite strict regulations.

Technological Advancements: Since its seed raise in 2024, Spoor has doubled its tracking range to 2.5 kilometers and improved bird identification accuracy to 96%.

Global Reach: Spoor operates across three continents and collaborates with over 20 of the world's largest energy companies. It has also partnered with Rio Tinto to track bats.

Future Prospects: Interest in Spoor's technology is growing, with potential applications in airports, aquaculture farms, and tracking other objects. Helseth predicts increased demand due to regulatory crackdowns on wind farms.

Funding: Spoor recently raised an €8 million Series A round led by SET Ventures, with participation from Ørsted Ventures, Superorganism, and strategic investors.

Mission: Spoor aims to enable industry and nature to coexist, with plans to become a global leader in addressing these challenges.

TIME Honors AI Pioneers as Person of the Year 2025

Each December, TIME Magazine names a Person of the Year, highlighting individuals who have significantly influenced the news and the world. In 2025, TIME has chosen to recognize a group of people known as the 'Architects of AI,' comprising CEOs shaping the global AI race from the U.S.

The selection includes notable figures such as Nvidia's Jensen Huang, Tesla's Elon Musk, OpenAI's Sam Altman, Meta's Mark Zuckerberg, AMD's Lisa Su, Anthropic's Dario Amodei, Google DeepMind's Demis Hassabis, and World Labs's Fei-Fei Li. These individuals have been instrumental in advancing AI technology, influencing government policy, and intensifying geopolitical competition.

TIME's decision reflects the profound impact AI has had on various aspects of life, including the information landscape, climate, and livelihoods. The magazine's announcement was leaked on prediction market Polymarket before the official release.