Google I/O goes all-out with AI


At a glance

  • Video search through Google Lens. Open the app and take a video, ask questions and get answers in real-time.

  • Gemini will be able to go through the Photo Library to retrieve images or information buried in the thousands of photos you may have.

  • Workplace gets a Gemini boost, allowing AI to generate emails or reports based on the documents in your Drive.

  • Google intros Veo, its generative video AI.


GOOG.png

 

The recently concluded Google I/O event was all about AI. 

It focused on Google showing a glimpse on how AI will be incorporated across Google features, such as Chrome and Workplace. 

Let’s run through them: 

Google Lens 

One of the features is the ability to search using video through Google Lens. The idea is to use Google Lens to capture a video, ask a question, and receive an answer from Google. The hope is to get a coherent response.

It’s extremely important to note that, while it is fun to use AI, like Gemini, to conduct searches, it’s important—can't stress that enough—to know that it’s not perfect. Search results may come up totally fictional or inaccurate. If you’ve been using it for a while, it’s possible it may draw answers from previous data. 

Gemini in your Photo Library 

Okay, this one is cool. “Ask Photos” will allow Gemini to go through your Photo Library and pull up images you’ve asked about. Google CEO, Sundar Pichai, demonstrated the feature by asking for his license plate number and the results came up with all the digits and a picture to go with it. Of course, we are taking this with a grain of salt.  The event, after all, is a marketing thing. But it is still an interesting feature to try out once it rolls across Android devices later this year. Certainly, Pixel devices will get dibs on these features. We can’t confirm for sure if other Android devices with modified OS will be able to roll out these features right away.  

Gemini 1.5 Flash 

Google just loves to overwhelm us with all the AI stuff. Gemini Nano, Gemini Pro, Gemini Ultra, Gemini Advanced, Gemini 1.5, and now we have Gemini 1.5 Flash. Google says 1.5 Flash is as powerful as 1.5 Pro with improved reasoning and other capabilities. The difference with Flash is that it’s designed to be fast and efficient.  

According to Gemini itself, Flash is capable of handling tasks faster for real-time applications. It said we can consider Pro as the specialist and Flash as its efficient assistant. So, while the Pro can handle more complex tasks, it may take longer to come up with results, while the Flash may not be as complex but will be able to do things faster.  

Gemini busts into Workplace 

Surprising no one, Google has integrated Gemini 1.5 Pro into Workplace. Apps such as Gmail, Docs, Sheets, and Slide will have sidebars for Gemini. So, any files you put in your Drive will be accessible by Gemini. Ideally, you will be able to ask Gemini to generate documents that contain information from other files within your Drive.  

More things AI 

From this point on, Google explores the AI’s realm of possibilities. This includes Veo, Google’s own AI generative video. Using prompts, users will be able to create videos and adjust using additional prompts. The videos generated are realistic and capable of creating cinematic shots. It’s a rival to OpenAI’s Sora, which recently shocked the internet with its generative video. We’re hearing stories about Google pitching this to YouTube creators and Hollywood—what this means for the movie industry remains to be seen. Personally, this may be used to generate backgrounds or special effects and help blend images better. There are too many movies these days where you watch them, even as YouTube clips, you can notice how fake the backgrounds are, and yes, they are annoying. AI products like Veo, may be able to help with that. I don’t think actors are going to be replaced (yet) any time soon. But creatives in the visual and audio field may have to work much harder (as if the industry isn’t already brutal).  

This can be concerning as we live in a digital age where misinformation is running amok. Though, Google will also be releasing a watermark feature for Veo-generated videos through SynthID. Google said this will also detect if a particular content was generated by AI. 

In the US, for now, Google will also roll out AI Overviews. It creates a summary of search topics for you, so you can get information faster and in an organized format. It sounds pretty much like Arc Search, something I’ve been using every now and then when I have questions. 

Android devices will get smarter AIs as well. This includes spam detection, something we all could use. Users will soon be able to ask questions about videos running on the screen. 

Google will also be integrating Gemini Nano, the lightweight AI that is available on Android devices now, into Chrome. This way, as Google puts it, you can ask it to generate social media posts for you. 

Finally, Google talks a little about Project Astra. This is Google envisioning a future where we have AI assistants, ones capable of operating at the same level we see in science fiction movies. The kind of AI that can do a lot of things.  

Suffice to say, with all this AI from Google alone, along with rivals across the industry, the next decade is going to be both exciting and frightening at the same time.