DeepMind, Google’s artificial intelligence (AI) subsidiary, has announced that its upcoming AI system will surpass OpenAI’s ChatGPT in capability.
DeepMind’s Gemini AI: A Leap Ahead
DeepMind’s CEO, Demis Hassabis, detailed in a Wired report that the company’s Gemini AI is a Large Language Model (LLM) that processes text similarly to ChatGPT but boasts added abilities such as planning and problem-solving. The forthcoming system builds on technology and techniques from DeepMind’s early AI system, AlphaGo, developed in 2016.
Hassabis explained, “At a high level, you can think of Gemini as combining some of the strengths of AlphaGo-type systems with the amazing language capabilities of the large models.” He hinted at “new innovations,” terming them “pretty interesting.” The first unveiling of Gemini occurred at Google’s I/O developer conference in May, alongside several other AI products.
Awaiting Gemini’s Official Launch
Although DeepMind has not yet fully launched Gemini, Hassabis says the system is just a few months away. The development of Gemini is likely to cost tens or even hundreds of millions of dollars. Despite the hefty investment, Hassabis expressed confidence, seeing no reason why their work with Gemini “won’t work.”
DeepMind’s CEO also emphasized the pressing need for further research into potential risks of more powerful AI systems. He predicted the emergence of human-level AI before 2033 in earlier statements.
Google Bard: Another Contender in the AI Space
News of Gemini’s development comes shortly after Google hurriedly launched its own AI chatbot, Bard, rivaling OpenAI’s ChatGPT. However, Bard’s debut faced criticism and concerns for societal impact from former Google employees.
In fact, regulators in Ireland have blocked the launch of Google Bard within the European Union due to Google’s alleged failure to submit the required documentation.
