Google in January: Circle to Search and the End of School Geometry

While everyone waited for Google to finally wake up and give OpenAI a worthy answer, the company decided to simply bury us in updates within a single month. January was all about "catch up and overtake," and it must be said that the search giant has started to pull it off. If Google spent last year making excuses for half-baked releases, now we are seeing an attempt to build AI in where it is actually needed: our smartphones and everyday tasks.

They started with the sorest spot: search. Circle to Search, launched alongside the Samsung Galaxy S24 lineup, looks like an admission: we have become too lazy to switch between apps. You no longer need to copy text or take a screenshot to find the sneakers from an Instagram video. You simply circle the object with your finger right on the screen, and Google's magic returns a result. This is an important UX shift, because search stops being a separate action and becomes part of the system interface. Google understands that if search is not instant, chatbots will take its place.

But the real heat came in the research labs. Google DeepMind introduced AlphaGeometry, a system that solves highly complex school Olympiad geometry problems at the level of gold medalists. Why does it matter? Because ordinary language models like GPT-4 are catastrophically bad at strict logic and spatial reasoning. They simply guess the next word. AlphaGeometry combines the intuition of a neural network with the strict logic of a symbolic engine. It is a huge step toward AGI, where a machine does not just imitate human speech but actually understands the rules of the physical and mathematical world.

Creativity was not forgotten either. While the internet buzzed about Sora's capabilities, Google quietly showed Lumiere. It is a video generation model that tackles the industry's main problem: jittery footage and lack of coherence. Thanks to a new Space-Time U-Net architecture, the neural network creates all the video frames at once rather than one after another. The result is smooth motion and physics that does not fall apart before your eyes. For now it is only a research project, but it makes one thing clear: startups will not have a monopoly on high-quality video content.

All this January activity says one thing: Google has moved from panic to systematic work. It is no longer trying to simply build "its own ChatGPT." It is embedding AI into Android, Chrome and its cloud services, making it unobtrusive but ubiquitous. Integrating Gemini Pro into Samsung smartphones is only the beginning of a big game to capture the mobile market for AI features. Google owns the operating system, and that is its main trump card, which neither Microsoft nor OpenAI has.

The bottom line: Google is no longer a follower but an aggressive player, ready to change user habits to keep its lead. Will Apple be able to offer something comparable at WWDC in June?

*Meta has been designated an extremist organization and is banned in the Russian Federation.