Head of Google DeepMind claims that the next chatbot of Gemini lab will surpass ChatGPT
ChatGPT may have grabbed the world’s attention, but Google’s DeepMind research lab claims that its next big language model will be able to rival and possibly surpass OpenAI.
According to information published in Wired, DeepMind uses techniques from AlphaGo, DeepMind’s artificial intelligence system that was the first to beat a professional go player, to create Gemini, a chatbot that can compete with ChatGPT.
If all goes according to plan, Gemini will be able to plan, solve problems, and analyze text, DeepMind CEO Demis Hassabis told Wired in an interview.
“At a high level, you can think of Gemini as a combination of some of the advantages of systems like AlphaGo with the impressive language abilities of large models,” – Hassabis said. “We also have some new innovations that will be quite interesting.”
Will Knight, the author of the article, suggests that Gemini, which was briefly mentioned at the Google I/O developer conference in May, will use the latest advances in reinforcement learning to perform tasks that today’s language models have difficulty with.
It is worth noting that Gemini is not DeepMind’s first step in the field of language models. Last year, the company introduced Sparrow, a chatbot that the lab claimed was less likely than other language models to give “dangerous” or “inappropriate” answers to questions.
Hassabis told Time in January that DeepMind was considering releasing Sparrow for private beta testing this year; it is unclear whether these plans remain in place.
However, Gemini is DeepMind’s most ambitious work in this area to date, at least according to initial reports. The Information reported in March that senior Google executives, including Jeff Dean, the company’s senior executive director of artificial intelligence research, were assigned to the Gemini project, which was sponsored due to the failure of the Bard project, to develop a Google chatbot to compete with ChatGPT.
The struggle for dominance in the field of generative artificial intelligence is taking place against the backdrop of enormous enthusiasm from investors and users. According to Grand View Research, the market for generative artificial intelligence, including text-analyzing AI like Gemini, could reach $109.37 billion by 2030, representing a CAGR of 35.6% from 2030.