Google Search Becomes a Chatbot: Gemini 3 Now Answers Everything

Remember when Google was just a list of ten blue links? Those days have officially moved into the history books of the internet. Today the search giant is taking its boldest step yet toward making sure you stop visiting third-party sites altogether. Google is updating its AI Overviews, building in Gemini 3 and adding the ability to ask follow-up questions right in the middle of a search. This is no longer a search engine in the usual sense. It is a full chatbot that uses the entire internet index as its knowledge base.

It all comes amid growing panic in Mountain View. After OpenAI announced SearchGPT and Perplexity began taking market share from power users, Google had to speed up. AI Overviews used to be a static box at the top of the results that often got things wrong and once advised putting glue on pizza. Now the company promises that Gemini 3 will make answers smarter and, above all, more human. You ask a question, get a summary, and if something is unclear, you simply keep talking. That is the new AI Mode, as described by Robby Stein, vice president of product.

Google's logic is simple: why should a user leave for some blog or news site if the algorithm can retell the article right here? The transition between a short overview and a deep dialogue is now seamless. You just scroll down and keep chatting with the neural network. For the average person, that is convenient. For site owners who spent years optimizing content for Google's algorithms, it looks like an elaborate betrayal. Google takes their traffic, using their own content to train its models and generate answers that remove the need to click.

Interestingly, the company positions this as a way to answer absolutely any query that pops into your head. If search used to be a tool for finding the path to information, it is now becoming the information itself. That is a fundamental shift in the business model. If people stop clicking links, what happens to the contextual advertising that feeds Google? We are probably heading into an era in which ads are woven natively into chatbot answers, making them even more intrusive and harder to tell apart from reality.

Technically, Gemini 3 in search is supposed to be faster and more accurate than its predecessors. Google claims the model understands the context of complex queries better. For example, if you are planning a trip and ask about the weather, then follow up about the best restaurants nearby, the system will understand that you are still talking about the same location. This creates the illusion of talking to a thinking assistant rather than a cold ranking algorithm.

The only question is how reliable these answers will be. We have already seen neural networks hallucinate, passing off fiction as fact. By turning search into a dialogue, Google takes on even more responsibility for the truth of every word. But in the race for survival against OpenAI, reliability seems to have taken a back seat. The main goal now is to keep the user at any cost, even if that means finally burying the open web as we knew it.

The bottom line: Google no longer wants to be a door to the internet. It wants to be the internet itself. Can the company maintain answer quality at this scale?