Google has released a new feature which will allow people to search the internet by taking a video. Video search will let people point their camera at something, ask a question about it, and get search results. Android and iPhone users globally will gain access to the feature from 1700 GMT by enabling "AI Overviews" in their Google app, but it will only support English at launch. It is the latest move from the tech giant to change how people search online by utilising artificial intelligence (AI). It comes three months after ChatGPT-maker OpenAI announced it was trialling the ability to search by asking its chatbot questions.

Google has released an open AI model called Gemma, which it says is created using the same research and technology that was used to build its Gemini AI models. The company says Gemma is its contribution to the open community and is meant to help developers "in building AI responsibly." As such, it also introduced the Responsible Generative AI Toolkit alongside Gemma. It contains a debugging tool, as well as a guide with best practices for AI development based on Google's experience. The company has made Gemma available in two different sizes -- Gemma 2B and Gemma 7B -- which both come with pre-trained and instruction-tuned variants and are both lightweight enough to run directly on a developer's laptop or desktop computer. Google says Gemma surpasses much larger models when it comes to key benchmarks and that both model sizes outperform other open models out there.

Deep learning's recent impressive breakthroughs have largely been based on huge quantities of manually annotated training data. Because appropriate large labelled datasets are often simply not available and manual labelling takes a lot of time, deep learning researchers are always on the lookout for ways to achieve the desired results with smaller amounts of data. Moreover, solving this data challenge can enable AI practitioners with limited resources to perform faster and cheaper model customization. Few-shot classification is a crucial component in the struggle to teach models using limited information. Although this is a hot research field, previous benchmarks could not reliably evaluate different models, which has hindered research progress. In a paper published at ICLR 2020 this month, Google AI researchers introduce Meta-Dataset, a large-scale and diverse benchmark for measuring the ability of few-shot classification models.

Since the explosion of AlexNet paper onto the computer vision scene, machine learning results have improved greatly due to deeper models with high complexity, increased computational power and the availability of large-scale labelled data. The room to refine neural networks still exists as they sometimes fumble and end up using brute force for lightweight tasks. To address this, the researchers at Google, have come up with a RigL, an algorithm for training sparse neural networks that use a fixed parameter count and computational cost throughout training, without sacrificing accuracy. So, what is sparse in the context of neural networks? Each layer of neurons in a network is represented by a matrix. Each entry in the matrix can be thought of as representative of the connection between two neurons.

Google is working on some nifty new features for Google Maps, including a short list of your favorite places, the possibility of a'virtual positioning system' and more. Assistant is coming to Google Maps in a big way, with a ton of new shortcuts, as well as the ability for the digital assistant to text your friend when you're on your way. Google is rolling out a tool called'Your Match', which uses machine learning to determine your location and interests, serving up targeted suggestions for new businesses opening up in your area and more.

Google is launching a new hardware and software kit aimed at developers and hackers who want to build products that incorporate computer vision… on a budget. It's called the AIY Vision Kit, and it's up for pre-order from Micro Center for $45, with an expected ship date of December 31st. You'll need to spend a few more bucks to actually use the kit though: it's designed to work with a Raspberry Pi Zero W mini computer ($10) and a Raspberry Pi Camera V2 ($30). And you'll need to supply your own SD card. The total cost is still under $100 though, which is pretty impressive for a set of tools that you can use to develop projects that use what Google calls an "intelligent camera."

You name it, and Google launched it. Two sizes, almost the same phone.Google introduces the Pixel 2 series -- and it's really proud of the camera After all those leaks, the Pixel 2 and Pixel 2 XL are now a reality, and both ditch the headphone jack. They're both (a little) squeezable, which will coax the phones to launch Google Assistant for your questions and search queries, while also packing water resistance and a camera that's apparently better than last year's well-regarded Pixel phone. Google was keen to sidestep the tech spec nitty-gritty and talk about the AI smarts it's stitching into its phones. You might recall Google's Lens visual search from earlier in the year -- well that's coming to these phones before the end of 2017.

Recently, Google announced another massive push into artificial intelligence, using a new approach to machine learning where neural networks are used to build--well--better neural networks. These artificial networks are designed to mimic the way the brain learns until AI is ultimately teaching itself. Google says its new technology, called AutoML, can develop networks that are more efficient and easier to use. Google CEO Sundar Pichai demonstrated the new technology at Google I/O, the annual developer conference that Google throws for app coders. "The way it works is we take a set of candidate neural nets, think of these as little baby neural nets, and we actually use a neural net to iterate through them until we arrive at the best neural net," explains Pichai.

Marking Google's 15th birthday, Hummingbird is the biggest change to the inner workings of the world's most popular search engine since Google's "Caffeine" update in 2010, which sped up Google's indexing of sites and delivery of search results. The Hummingbird update focuses more on Google's Knowledge Graph – an encyclopaedia of about 570m concepts and relationships that allows Google to anticipate facts and figures you might want to know about your search term. Hummingbird isn't an overhaul that Google search users will instantly notice, however. "In general, Hummingbird – Google says – is a new engine built on both existing and new parts, organised in a way to especially serve the search demands of today, rather than one created for the needs of 10 years ago, with the technologies back then," said Danny Sullivan of the search blog Search Engine Land. It will benefit those using more modern forms of search, such as conversational or voice search, where you ask Google a question rather than typing keywords into the search box.

For all the bells and whistles of Google's new smartphone, the biggest point of emphasis is tucked away in small type on the back of the device: "Phone by Google." Today, most of Google's smartphone software runs on devices manufactured by companies like Samsung, LG Electronics and Lenovo, with Google's presence often relegated to the background. But with its new Pixel smartphone, introduced Tuesday at a press event, Google is front and center and selling a phone that it created from the industrial design to the components. Google's new strategy of controlling both hardware and software for its devices puts the company more directly in competition with Apple and many of its own Android partners. It's a necessity, Google says, because of the advent of artificial intelligence.