New AI features
Google wants to compete with ChatGPT
02/07/2023, 05:10 am
The text robot ChatGPT is currently causing a worldwide sensation. In response, Google is now presenting several new AI applications. These will soon be made available to the public.
With a comprehensive initiative, Google wants to present its applications with artificial intelligence to a broad public. This was announced by CEO Sundar Pichai in a blog entry. With the AI offensive, the search engine giant is reacting to the success of the start-up OpenAI, which has attracted the attention of the tech world in recent weeks with its text robot ChatGPT. Google’s AI initiative consists of three parts: a chatbot called Bard, new AI capabilities in Google Search, and the provision of programming interfaces (APIs) that can be used to develop AI applications.
The AI-controlled language model Lamda (Language Model for Dialogue Applications) from Google will play a central role in this. The group management had celebrated Lamba 2021 as a “breakthrough in conversation technology”. After that, the model disappeared from public view. Bard is based on a variant of Lamda and is described by Google as an “experimental AI service for conversations”. Bard will open to “trusted testers” starting immediately before opening to the public in the coming weeks. With the name of the service (German: “Barde”), Google alludes to the poet William Shakespeare. England’s national poet is often referred to as the “Bard of Avon”.
“Bard seeks to combine the breadth of world knowledge with the power, intelligence, and creativity of our large language models,” Pichai wrote. It draws on information from the Internet to provide up-to-date, high-quality answers. “Bard can be an expression of creativity and a springboard for curiosity – whether it’s explaining the new discoveries of NASA’s James Webb Space Telescope to a nine-year-old or learning more about the best forwards in football today.” Google also wants to offer a glimpse into several AI-powered features in Google Search. This can help users to summarize findings for questions to which there is no single correct answer. You will soon see functions in the search “that convert complex information and different points of view into easily digestible formats”.
Beware of fraud and abuse
Google has long been developing software based on artificial intelligence that can talk to people. As early as spring 2018, the group demonstrated a program that called restaurants to make a reservation – and was not recognized as a computer. Criticism was immediately raised that such technology could be misused. Google has had its voice software used internally by employees for the past few years, but shied away from a broad market launch due to the risks. In November 2022, OpenAI made its ChatGPT software public, which delivers texts in a few seconds that can hardly be distinguished from answers written by real people – also in German.
The technology is causing a stir, but it is also causing concern: after all, you can try to cheat with it at school or university or create false information on a large scale for distribution on the Internet. ChatGPT also sometimes gives wrong answers, but this is not recognizable for users. ChatGPT also strategically pressured Google to compete with Microsoft. The arch-rival, which has had moderate success with its search engine Bing, has invested billions in OpenAI and will integrate the start-up’s software into its successful Azure cloud platform.