Microsoft is parking its tanks firmly on Google’s lawn as it plans to integrate the sophisticated ChatGPT chatbot into its Bing search engine.
The new artificial intelligence ’bot has taken the world by storm since it was launched last November, impressing users with its flawless and often seemingly thoughtful responses to questions and directions.
Microsoft believes that it could give those searching for information a better experience than the current list of links provided by most search engines, according to sources cited by The Information website.
Google has cornered the market in search with a more than 80 per cent share, followed by Microsoft’s Bing and Yahoo, which is powered by Bing. Google is used as a verb and was added to the Oxford English Dictionary in 2006. To show the extent of its dominance, in 2021 Google even claimed in a European court that the most searched word on its rival Bing’s search engine was “Google”.
Microsoft hopes that the introduction of ChatGPT to Bing in March could change that. It has demonstrated prowess at writing everything from business plans to jokes, plays and letters; translating languages; and even addressing philosophical questions in a human-like manner.
Advertisement
Its responses are so impressive that educationalists have expressed concern that students could use it to write their essays without being spotted, saying it could encourage cheating.
The ’bot was developed by OpenAI, a company set up in 2015 by Elon Musk and Sam Altman as a research institute with the aim of ensuring that artificial general intelligence could be created safely for the benefit of all humanity. Musk left in 2018, citing a conflict with his AI work at Tesla.
Since then OpenAI has become a “capped profit” company — profits are limited to 100 times investment — and has taken $1 billion of investment from Microsoft, which is allowed to commercialise the tech.
The company has released the chatbot for a short period as a research tool as it develops GPT, the “large language model” at its core. The ’bot will continue to learn from millions of interactions with users during this time. It was trained on billions of words from the internet but also fine-tuned by humans, who ranked its responses. OpenAI has been clear, however, that the ’bot will often be wrong, and its knowledge base does not extend beyond 2021.
Google has been working on a similar conversation technology since 2017, LaMDA, which hit the headlines when an engineer at the company claimed it was sentient. In a 2021 blog post Google said it had more to do to ensure the ’bot adhered to its standards before release. Executives called the success of ChatGPT a “code red” threat to its business, according to the New York Times.
Advertisement
Microsoft would not comment.
