The race to develop powerful language models continues to heat up as DeepMind, the renowned research lab owned by Google, sets its sights on creating a chatbot that could rival or even surpass OpenAI’s popular ChatGPT. Named Gemini, this new AI language model incorporates techniques derived from DeepMind’s groundbreaking AlphaGo system, which famously defeated a professional human player in the ancient board game Go. Gemini represents an exciting fusion of AlphaGo’s strategic planning abilities with the remarkable language capabilities of large-scale models.
DeepMind CEO Demis Hassabis revealed to Wired that Gemini would not only possess the capacity to analyze text but also exhibit problem-solving skills. “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,” Hassabis stated. “We also have some new innovations that are going to be pretty interesting.” The forthcoming Gemini chatbot, briefly previewed at Google’s I/O developer conference in May, is expected to leverage advancements in reinforcement learning, a technique that rewards and penalizes AI systems to teach them desired behaviors. This approach has already yielded significant improvements in the language model domain, as seen in ChatGPT’s response generation. With DeepMind’s expertise in reinforcement learning, honed through projects like AlphaGo, the lab aims to push the boundaries of generative AI.
It’s important to note that Gemini isn’t DeepMind’s initial venture into language models. Last year, the lab introduced Sparrow, a chatbot specifically designed to minimize “unsafe” or “inappropriate” responses. DeepMind’s CEO mentioned the possibility of releasing Sparrow for private beta testing this year, although its current status remains unclear. Gemini represents DeepMind’s most ambitious effort in the language model space thus far, according to early reports.
The Information reported in March that Gemini emerged in response to Google’s chatbot project, Bard, struggling to keep pace with ChatGPT. Significantly, Gemini has garnered direct involvement from top-level Google executives, including Jeff Dean, the company’s most senior AI research executive. The race for dominance in the generative AI field is driven by soaring investor and customer enthusiasm. According to Grand View Research, the generative AI market, encompassing text-analyzing AI such as Gemini, is projected to reach a staggering $109.37 billion by 2030, marking a 35.6% increase from 2023.
As DeepMind continues to push the boundaries of language models with Gemini, the stage is set for an intense competition between AI giants seeking to revolutionize the way machines understand and generate human language. With its unique blend of AlphaGo-inspired techniques and cutting-edge reinforcement learning, Gemini promises to be a significant contender in the race for the most advanced chatbot.