Compared to ChatGPT, Claude, and all the other cool kids on the block, Google isn’t getting as much attention for Gemini, its AI assistant. However, the recent pace of innovation and productization should convince people (and the tech press) to reconsider their priorities. Gemini is in a very good position to win this race. Here is why: * **Mission critical:** Google is probably *the* company that would benefit the most from AI, given its mission *“to organize the world's information and make it universally accessible and useful.”* What could be a bigger leap for their mission than an LLM containing the world's knowledge? * **Agentic capabilities:** An AI agent is only as useful as its integrations to the tools that you need it to control. Anthropic is trying to solve this through the Model Context Protocol (MCP). OpenAI is going down the path of desktop-based integrations into other apps (an approach that won’t work on mobile or on the web). Google already owns the tool stack that would benefit from agentic AI capabilities, and it’s well aware of this advantage. Gemini is already integrated with Gmail, Drive, YouTube, Docs, Calendar, and many more, making it the perfect AI Agent within the Google universe. Google’s longstanding investment in connectors is paying huge dividends in the AI era. * **Distribution:** Google is everywhere, allowing them to add Gemini to their existing footprint. Chrome, Android, YouTube, Gmail, Maps…and Search. These are just the most popular ones. Google can bring Gemini to billions of users without spending a single additional dollar on marketing. * **Knowledge/Search:** AI models are trained on large sets of data, and when they are released months later, their knowledge is already outdated. That’s why all the AI vendors are integrating search features into their products, some even make it their business model (Perplexity). Google doesn’t just have search capabilities, it dominates search. They have the largest search index in the world, access to news in nearly every country, the largest POI database with Google Maps, YouTube as the biggest learning database ever, and so much more. This gives them an advantage in training models, but also in making them genuinely useful. * **Productization pace:** Google is adding new AI features to its products at a rapid pace. Looking at their [updates page](https://gemini.google.com/updates) gives you a good idea of how serious they are about winning. At the time of writing this (March 2025), they have finally caught up with ChatGPT, thanks to their Canvas and Deep Research features. * **You:** AI as a digital assistant is only as good as its integration, and through that, the amount of information it has available about you. AI will be a true personal assistant when it knows your schedule, reads your emails, follows your travels, and sees the pictures you take. Google has all of that, and with a few more months of productization, it will be seamless. With the [ongoing arms race to productize AI technology](Productizing%20AI%20Technology.md), it’s too early to announce a winner just yet. But Google is clearly driving the fastest pace right now and is in the process of overtaking the competition. OpenAI, Anthropic and others are not sleeping though. Also, I’m not writing off Apple yet. They are frustratingly lagging behind, but failure is not an option for them. Interesting times ahead.With the ongoing arms race to productize AI technology, it’s too early to announce a winner. However, Google is clearly driving the fastest pace right now and is in the process of overtaking the competition. OpenAI, Anthropic, and others are not sleeping, though. Also, I’m not writing off Apple yet. They are frustratingly lagging behind, but failure is not an option for them. Interesting times ahead\!