As we know that ChatGPT has got everyone’s attention. DeepMind, a Google-owned research association, guarantees that its next wide language model will be comparable to or far superior to OpenAI’s.
According to a Wired article, DeepMind is creating Gemini, a chatbot that competes with ChatGPT, using AlphaGo’s AI algorithms. AlphaGo was the first AI system to defeat a skilled human player in the board game Go.
On the off chance that everything goes as expected, Gemini will actually want to plan, take care of issues, and examine text, as indicated by DeepMind CEO Hassabis.
Gemini, in a broad sense, combines some of the advantages of systems similar to AlphaGo with the incredible language abilities of the huge models, according to Hassabis. We also have some fresh developments that will be rather fascinating.
Knight hypothesizes that Gemini, which was briefly hinted at Google’s I/O developer conference in May, will use advancements in reinforcement learning to complete tasks that current language models find challenging. Reinforcement learning involves “rewarding” an AI system for certain behaviors and/or “punishing” undesirable ones in order to “teach” the AI system which behaviors to exhibit in a certain situation.
As Knight brings up, improvements in the field of language models have proactively been made because of support learning, which is essential to how systems like ChatGPT respond to orders.
Given its depth of experience (AlphaGo being one example), DeepMind is surely eager to extend its reinforcement learning expertise to the field of generative AI.
It’s crucial to remember that DeepMind has tested language models prior to Gemini. Sparrow, a chatbot that the company unveiled last year, was marketed as being less likely than previous language models to react to queries in an “unsafe” or “inappropriate” way. Hassabis told Time in January that plans to release Sparrow for a private beta at some point this year are still on track, but it’s unclear if they are.
But if early reports are to be believed, Gemini is DeepMind’s most audacious space endeavor to date. According to a March article in The Information, Jeff Dean, Google’s most senior AI research officer, has direct participation in Gemini, the chatbot project that was inspired by Bard, Google’s chatbot project that fell behind ChatGPT.
The competition for supremacy in the generative AI market is fueled by the extreme customer and investor enthusiasm. Grand View Research estimates that the market for generative AI, which includes text-analyzing AI like Gemini, could expand by 35.6% from 2030 to $109.37 billion by 2030.