The New York Times reports that Google will launch new AI-powered tools in September, and even more features are coming.
The New York Times reports that the new features will only be available in the US and initially will be limited to one million users. The tools are not yet known, but it is likely that they will build on Google’s Bard experimental chatbot’s conversational promise. The tools are being developed under the “Magi” codename.
Google is implementing these plans to counter the threat of new systems such as Microsoft’s Bing Chatbot and OpenAI ChatGPT. These chatbots are thought to be able to replace Google and other traditional search engines in the future, despite their shortcomings.
According to the New York Times, Google is under such threat that Samsung may replace Google as the default mobile search engine with Bing. Google is expected to earn $3 billion annually from this deal (it has a similar agreement with Apple that is worth $20 billion), but it is not known how seriously Samsung will consider the switch. Microsoft’s AI might have influenced the company, but it could also be taking advantage of Google at a time of weakness.
Google plans to radically re-build its search engine, in addition to the new features that were developed under project Magi. The Times claims that there is “no timetable for when Google will release its new search technology.”
Google is also reportedly working on a number of other AI tools. These include an AI image creator called GIFI and a language-learning system called Tivoli Tutor. Another feature called Searchalong would integrate a bot into Google’s Chrome web browser to answer any questions about the current webpage. This is similar Microsoft Bing AI’s sidebar for Edge browser.
Google is working on a chatbot which can generate code and answer questions about software engineering. Google has also tested a feature which allows people to search music via a chatbot. According to The Times the company has dedicated more than 160 staff to the redesign.
Chrome’s “Searchalong” feature is one of the new features “in various stages in development.” A chatbot could scan a webpage to provide contextual information. If you are looking for a place on Airbnb to stay, you can ask the chatbot what you should see and do nearby. Another pair of experimental features called “Tivoli Tutor” and “GIFI,” would let users ask Google Image Search for images, and then converse with a bot to learn a language. Many of these features have been demonstrated by Google in the past, or are available on other platforms such as Duolingo. Slides , for example, already supports image generation in.
Google is reportedly planning to announce Magi in the next month, before introducing new features later this fall. This timing indicates that the project will be presented at 2023. The company intends to make Magi’s features available to one million US users before expanding to 30 millions by the end the year.