Happy Holidays!

Hello Friend,

Thank you for staying by our side through 2025. We hope you enjoyed every update and found it as interesting as we did creating it. We look forward to building a much stronger relationship and achieving amazing things together in 2026.

Merry Christmas and Happy New Year in advance from all of us at The AI Colony.

LG Electronics is preparing to show off its new home robot called CLOiD at the upcoming Consumer Electronics Show in Las Vegas from January 6-9. The robot represents the company's push into home robotics and its vision of what it calls the "Zero Labor Home," where technology handles daily chores so people can spend more time on things that matter to them.

CLOiD is designed to handle a variety of household tasks inside the home. The robot has two articulated arms that can move in seven different directions, similar to how human arms work. Each hand has five individually controlled fingers that can grip and manipulate objects with the kind of precision needed for real household tasks. One teaser image shows CLOiD grabbing a towel, which might seem simple but is something most home robots can't actually do yet.

The robot's brain is housed in its head, which contains a chipset along with a display, speaker, camera, and various sensors. These components allow CLOiD to communicate expressively with people in the home. The robot runs on what LG calls Affectionate Intelligence, which is designed to sense its surroundings, interact naturally with people, and learn from repeated interactions over time. The goal is to provide increasingly personalized support for everyday tasks as the robot gets to know the household.

The teaser footage shows the robot using its five-fingered hands to pick up household items and interacting with people through gestures such as a fist bump. LG says CLOiD can recognize and learn its environment, adapt to people's schedules and lifestyles, and control connected LG smart home devices to provide customized care.

LG views robotics as an important part of its future growth strategy. The company recently established the HS Robotics Lab within its Home Appliance Solution division specifically to develop advanced robotics technologies and strengthen its capabilities in this area. LG is also working with robotics companies in Korea and around the world through joint research projects and strategic partnerships. The company has invested in robotics startups including Figure AI in the United States and AGIbot in China.

Visitors to CES 2026 will be able to see CLOiD in action at LG's booth, where the company plans to demonstrate various "Zero Labor Home" scenarios showing how the robot works in real-life situations. While LG hasn't announced a release date or price for CLOiD, the robot signals the company's serious commitment to building practical home robotics rather than just concept products.

Amazon-owned Zoox has issued a voluntary software recall for its self-driving vehicles after discovering problems that caused the robotaxis to cross into oncoming traffic lanes and block crosswalks. The recall affects 332 vehicles that provide free rides to passengers in San Francisco and Las Vegas.

The company filed documents with federal safety regulators on Tuesday explaining that while no crashes have occurred, the software issues could increase the risk of accidents. Zoox identified situations where its vehicles made moves that human drivers might commonly make but didn't meet the company's own safety standards. Sometimes the robotaxis would stop in a crosswalk to avoid blocking an intersection at a red light. Other times they would make late turns that became wide turns, crossing partially into the lane meant for oncoming traffic.

The problem first came to light on August 26 when a Zoox robotaxi made a wide right turn and briefly stopped while partially in the opposing travel lane. After that incident, the company reviewed its data and found 62 similar lane crossings near intersections between August 26 and December 5. Zoox has been working with the National Highway Traffic Safety Administration throughout this period to discuss how often these incidents happened and what caused them.

The company released software updates on November 7 and again in mid-December to fix all the identified problems. In a statement, Zoox said it successfully identified and deployed targeted software improvements to address the root causes. The company emphasized that transparency and safety are foundational to its operations, which is why it chose to submit a voluntary recall to be open with both the public and regulators about how it continues to refine its technology.

This isn't the first time Zoox has recalled its software this year. Back in March, the company recalled vehicles to fix unexpected hard braking issues after two motorcyclists crashed into the back of Zoox vehicles. Federal investigators had opened a preliminary investigation into those incidents. Then in May, Zoox filed two more software recalls related to concerns about how well the system could predict the movements of other people using the road.

The latest recall covers all Zoox vehicles that operated on public roads between March 13 and December 18.

Google's parent company Alphabet has agreed to buy Intersect Power for $4.75 billion in cash, plus taking on the company's existing debt. The deal, announced Monday, represents a major move to secure the massive amounts of energy needed to power artificial intelligence systems.

Intersect Power is a company that develops data centers alongside clean energy sources. The acquisition will help Alphabet expand its power generation capacity without having to depend on local utility companies that are struggling to keep up with the surging energy demands from AI companies. Getting reliable access to electricity has become one of the most critical challenges in training and running AI models, which consume enormous amounts of power.

This isn't Alphabet's first involvement with Intersect Power. The company already held a minority stake after Google and TPG Rise Climate led an $800 million strategic funding round in the company last December. That earlier partnership set an ambitious goal of $20 billion in total investment by 2030.

The acquisition includes Intersect's future development projects, but existing operations will be bought out by other investors and run as a separate company. Intersect builds what it calls data parks, which are essentially facilities located right next to wind farms, solar installations, and battery storage systems. These new data parks are expected to start operating late next year, with full completion scheduled for 2027.

Google will be the main customer using these facilities, but the campuses are designed as industrial parks that can accommodate other companies' AI chips alongside Google's own hardware. This approach gives Alphabet direct control over both the computing infrastructure and the power supply needed to run it, solving two problems at once.

The transaction is expected to close in the first half of next year, pending regulatory approvals. The deal reflects how tech giants are increasingly taking matters into their own hands when it comes to energy, rather than waiting for traditional power grids to catch up with their needs. As AI technology continues to advance and requires even more computing power, securing dedicated energy sources has become just as important as building the data centers themselves.

Waymo appears to be working on adding Google's Gemini AI chatbot to its self-driving taxis, according to findings by technology researcher Jane Manchun Wong. The feature would give riders an AI assistant to chat with during trips and help them control certain features inside the vehicle.

Wong discovered the plans while digging through the code of Waymo's mobile app. She found a complete set of instructions, over 1,200 lines long, that tells the AI assistant exactly how to behave inside a Waymo vehicle. When asked about the feature, Waymo spokesperson Julia Ilina said the team is always experimenting with ways to make riding with Waymo more pleasant and useful, but wouldn't confirm whether this will actually launch.

According to the instructions, the assistant would be able to answer rider questions, manage some in-car functions like temperature and lighting controls, and reassure passengers if needed. The bot is told to use clear, simple language and keep responses brief, usually one to three sentences.

When riders activate the assistant through the in-car screen, Gemini would greet them by their first name and could access information like how many Waymo trips they've taken. The assistant would be able to control temperature, lighting, and music, but not volume, route changes, seat adjustments, or windows.

The instructions make an interesting distinction between Gemini the AI assistant and the Waymo Driver, which is the actual self-driving technology. The assistant is specifically directed not to comment on or explain real-time driving actions. If a passenger mentions a video they saw of a Waymo incident, the bot is told to deflect rather than answer directly.

The assistant would be allowed to answer general knowledge questions about things like the weather or store hours, but wouldn't be able to take real-world actions like ordering food or making reservations.

Building AGI with our mission Intelligence with Everyone. Global leader in multi-modal models and AI-native products with over 150 million users.