A few years ago, “artificial intelligence” sounded like something out of science fiction. Today, it corrects our letters, picks vacation locations, recommends routes, and even explains complex articles. And it’s not some mysterious neural network that we can’t keep up with anymore, but services we know – Google search, Gmail, YouTube, and Android. It turns out that AI has been with us for a long time, and is so intertwined with our lives that we’ve stopped noticing it.
Google was one of the pioneers in creating language models, or so-called transformers, but it lost in the speed of their commercialization and introduction to the market. However, after launching its own chatbot Gemini, the company rejoined the race with artificial intelligence. But unlike OpenAI or Anthropic, which started from scratch, Google has the advantage of a huge ecosystem of tools used by a billion people. And it was they who began to rapidly receive “smart” functions: text generators, summaries instead of search results, image editing, personal tips and advice.
This article is a brief tour of the world of Google services, which as of 2025 have become much smarter, and in some places almost telepathic. What of this is actually useful, what is controversial or doesn't really work, and what could change the way we work and communicate online?
Google Search – AI in search
Google is rapidly and deeply integrating artificial intelligence into its core product, search. It now goes beyond just showing links to formulating generalized answers, giving suggestions, selecting options, and even planning for the user.
In 2024, the company introduced AI Overviews, which are generative answer blocks that appear at the top of the results page. They summarize key information from multiple sources and provide a concise answer right away, without the need to click through links. Previously, this feature was only available through Search Labs, but now it is gradually being rolled out to a wider audience, starting with the US, and now it is also available in Ukraine.
These reviews are especially useful when a user asks a complex or multi-part question. For example, you might type something like, “Find a Pilates studio near you that has morning classes, a great price, and parking.” Previously, you would have to click through 5-6 sites to find this. Now, search provides a concise summary of the best options based on the context of the query.
The user can simplify the answer – for example, ask for an explanation in simpler words. Or vice versa – add more details. This actually turns the search into a dialogue – you can refine your queries, and the system already understands the previous context.
However, some users complain about the inaccuracy of the answers in AI Overviews - there have been examples when the system gave incorrect or even absurd information, which is not surprising for AI. However, in general, the system works, and Google itself admits that the technology is not yet perfect and is improving it.
The new approach is also the structuring of the results page. Google now groups content into thematic blocks – for example, “Prices”, “Reviews”, “Most Popular” – so that the user immediately sees relevant categories without having to delve into each source individually. This is especially useful when choosing products, planning trips or searching for ideas.
In addition to text, the company is also taking a step towards visual understanding of queries. Another innovation is the ability to search by video. A user can shoot a video, for example, of a faulty device, and upload it to Google - the search will generate step-by-step instructions or an explanation of what exactly is happening in the video. This brings the system closer to multimodal understanding, which can work not only with text, but also with visual content.
In 2025, Google also began rolling out AI Mode, a new search mode powered by the more powerful Gemini 2.0 engine. In this mode, the user gets even deeper answers, enhanced integration with other Google services, the ability to interact via images, text, and voice simultaneously. It also boasts better situational awareness in programming.
Another issue worth highlighting in this case is the reduction of clicks to third-party sites. When the answer is right there in the search results, fewer people open the original source.
This is causing concern among media and website owners, as organic traffic is decreasing. Google, for its part, is trying to maintain a balance – adding links to sources, highlighting quotes, etc.
Gmail and Google Workspace: AI as a secretary and editor
Despite the loud announcements, most of the "revolutionary" AI features in Gmail still remain behind the scenes for Ukrainian users. Neither Help me write, nor Smart Compose, nor automatic tone suggestions — all of this is more about the future than reality for now.
In Google Docs, the situation is a little more interesting. There, autocomplete of phrases works (in English, though), and you can also activate templates of structured documents - for example, for resumes or reports. This is not AI in its pure form, but it is something. For the Ukrainian language, these functions are, to put it mildly, incomplete - but even basic tools can speed up work if you get used to them.
Yes, it’s not a “virtual editor” that will write everything for you. But as a set of smart habits, it works well. And perhaps that’s the essence of good digital tools: they don’t do everything for you, but they keep pace.
AI sometimes suggests phrases that sound like something out of a business style book, and this distorts authenticity. However, it is difficult to call it unnecessary – how long does it take to get used to writing business letters at the same speed without prompts?
Google Photos and Google Lens: AI sees more
It seems that the only one who remembers our photos better than us is Google Photos. It has long been organizing its image archive using AI - you can easily find all the pictures with friends or family. The search by image content is impressive: enter "dog" or "sea", and AI will show all the relevant photos. The service even creates short video memories from your favorite photos - it selects frames and puts music on top (although sometimes it gets annoying - I get sent such a video with photos of the neighbor's cat every month).
And Magic Editor is generally fantastic: remove an unnecessary object, replace the sky with a sunny one, or even change a facial expression – all this AI does in seconds. Most of us have tried editing vacation photos, and the result looked natural. But aren't we "spoiling" our photos by using artificial intelligence for editing? Will much of our "authorship" remain?
Google Lens goes even further: it recognizes text in a photo, translates it in real time, and identifies objects. It’s indispensable when traveling – take a picture of a menu at a restaurant, and you immediately know what to order. It’s also a godsend for students: take a picture of a formula or text, and AI will help you figure it out.
Of course, you can't blindly rely on it (but we do it from time to time!). Mistakes in recognition happen, and in critical situations they cost a lot.
Gemini instead of Google Assistant: a personal AI assistant
Gemini is gradually replacing Google Assistant on Android, allowing you to switch to an assistant interface that adapts to your habits, it also offers text generation directly in applications and integration with third-party programs. For example, you write a message in a messenger, and AI offers ready-made phrases or even entire paragraphs. This makes Android not just an operating system, but a real assistant. But isn't it penetrating our lives too much? The data that AI collects can tell us more about us than we are willing to show.
And isn't this another step towards losing control? Google promises privacy settings, but in practice they're often hidden deep in menus. As a user, I want convenience, but not at the cost of my freedom.
YouTube, recaps, and AI
YouTube is not just about videos, it’s an entire ecosystem where AI plays a key role. Automatic subtitles and their translation make content accessible to people with hearing impairments or those who don’t know the language of the video. It’s a revolution: imagine a Ukrainian watching a Japanese vlog with Ukrainian subtitles – and all thanks to AI. I often use this feature myself when I’m looking for something specific on foreign channels, and it’s amazing.
Another area where Google is testing Gemini’s capabilities is automatic video summarization. It all sounds convenient: instead of watching an hour-long lecture video, you’ll get a few paragraphs with a concise retelling.
In practice, it all looks like a crude experiment so far. Summaries are often superficial, overly general, or simply repeat phrases from subtitles. They save time, but rarely provide enough information.
The idea is clear: give the user a tool for quickly navigating content. But the technology is not yet at a level where it can truly replace viewing or provide meaningful presentation of the material. Rather, it is a direction that is interesting to observe — and which could potentially be useful when it becomes more accurate, more contextual, and better adapted to different video formats.
Google Bard → Gemini: the evolution of chat from Google
The transition from Bard to Gemini is not just a rebranding, but a real evolution of Google’s AI. Gemini competes with GPT-5: it generates texts, code, analyzes documents. It can create plans, theses, and even poems. There are paid and free versions, and integration with the Google ecosystem makes it available everywhere – from Gmail to Docs.
But there is also a question for the paranoid: isn't this becoming another way for Google to collect our data? Every request to Gemini is information about our interests, needs, and thoughts. Although, it's probably too late to worry about privacy, providing access to everything and everywhere...
Bottom line: What does “Googled AI” mean for us?
Google’s AI is no longer the future, but our present. It makes us more productive: it writes letters, edits photos, translates texts, suggests ideas. But along with comfort comes addiction. We save time, but we lose uniqueness – our texts, photos, even thoughts become partially “artificially intelligent”. Data control is another problem: Google guarantees security, but can we expect it forever?
AI is expected to penetrate even deeper into Google services soon. Google Maps could become a personal guide that knows your preferences, and Google Search could predict your queries before you formulate them. It’s convenient, but are we ready to give AI that much power?