AI’s Next Phase: Faster Chips, Realistic Robots, and the Energy Race

Week 5: Dec. 22 – 26

AI Unpacked

Welcome back to AI Unpacked, your weekly dose of AI news. We hope your Christmas was relaxing, festive, and only mildly painful for your wallet. This week, we cover the biggest AI stories from December 22–26, from Nvidia’s move involving Groq to the future of humanoid robots and beyond.

Nvidia's Licensing Deal With Groq

Nvidia is making another strategic move to strengthen its position at the center of the AI boom. The company has licensed AI inference technology from startup Groq, signaling how important fast and efficient AI inference has become as models move from training labs into everyday products.

Inference refers to the process of running AI models once they are trained, which powers tools like chatbots, search engines, and real-time recommendations. While Nvidia already dominates AI training with its GPUs, this agreement shows the company is focused on improving performance and efficiency as demand for large-scale inference grows. By licensing Groq’s technology, Nvidia can enhance how its systems handle real-time AI workloads without building everything from scratch.

As part of the deal, several Groq executives are joining Nvidia to help integrate the technology, while Groq continues operating independently. The agreement is non-exclusive, meaning Groq can still work with other partners, but it gives Nvidia access to specialized expertise at a critical moment in AI adoption.

For everyday users, this matters because faster and more efficient inference translates into quicker responses, lower costs, and more reliable AI-powered services. For investors and the broader market, the deal highlights a key shift in AI hardware. The next phase of competition is not just about training the biggest models, but about running them efficiently at massive scale.

Are Humanoid Robots A Threat To Your Job?

Even as billions of dollars flow into humanoid robot startups, many of the companies actually building these machines believe the technology has been overhyped and remains far from widespread real-world use. Leaders in the space say that while robots with human-like forms attract attention and investment, the reality is that most current systems can only perform very limited, narrow tasks under controlled conditions. Some robots are beginning to find specific roles in warehouses, such as moving objects or assisting with repetitive tasks, but complex capabilities like doing laundry or household chores are still distant goals. Experts at industry gatherings have pointed out that engineering reliable, safe, and cost-effective humanoid robots presents major technical challenges. These include the difficulties of locomotion, manipulation, balance, and handling diverse environments that humans navigate easily. As a result, many insiders caution against expecting humanoid robots to replace human labor soon, and suggest that specialized robots designed for particular jobs may be more practical in the near term.

Tesla's Optimus Robot. Tesla

Alphabet Moves to Power AI with Major Energy and Data Center Push

Alphabet, the parent company of Google, has agreed to acquire Intersect, a U.S.-based provider of data center and energy infrastructure solutions, in a definitive deal worth $4.75 billion in cash plus assumed debt. The move is part of Alphabet’s strategy to keep pace with the rapidly growing energy demands driven by artificial intelligence and cloud applications.

Intersect specializes in building energy generation and data center capacity in tandem, including projects that pair power plants directly with computing facilities. Under the agreement, Alphabet will bring Intersect’s team and multiple gigawatts of energy and data center projects in development or under construction into its broader infrastructure ecosystem, while allowing Intersect to continue operating under its own brand led by CEO Sheldon Kimber.

By acquiring Intersect, Alphabet aims to accelerate the rollout of critical power and data center capacity at a time when AI services are placing unprecedented strain on existing grids and traditional infrastructure. The deal builds on a prior minority investment by Google in Intersect and underscores the growing importance of energy strategy alongside chips and software in the AI era.

This acquisition reflects a broader trend among tech giants to secure reliable and scalable energy solutions as foundational support for AI development and cloud growth. Alphabet expects the transaction to close in early 2026.