Google is said to be focusing maximum resources on ensuring the release of its new artificial intelligence (AI) project ‘Gemini’ this fall, catching up with competitors like OpenAI’s GPT-4.
Gemini is basically combining the advantages of self-learning systems like AlphaGo, with the language capabilities of chatbots based on LLM models. In a word, Gemini combines the same text generation capabilities as GPT-4 with the ability to generate images from text descriptions. This will allow Google to build conversational AI products like chatbots, analyze data like charts, create graphics, and even control software with natural language commands.
At the heart of Gemini is an algorithm training solution that roughly translates to reinforcement learning, which allows an algorithm or software to learn how to approach a problem in a logical and systematic way, such as choosing a country. Go next in a game of Go, or play a game.
Earlier this year, Google merged two separate AI development teams under CEO Sundar Pichai, to accelerate the development of large language models like Gemini. The market potential is huge, but also competitive because OpenAI’s GPT-4 and DALL-E models have demonstrated the capabilities of advanced AI and are beginning to show real capabilities. In addition, a big name, Microsoft, is also showing great determination in this area, most obviously the recent project to integrate OpenAI’s artificial intelligence technology into Office 365 applications.
“Googlers over the past few months have mostly been discussing how to leverage Gemini to provide features like chart analysis or graphic creation with textual descriptions, as well as control over software. software by text or voice commands. Google is betting on Gemini to reap success from services like the Bard chatbot to enterprise applications.”
Google expects Gemini to support consumer products like the Bard chatbot, while also providing the developer community access to models through Google Cloud. This can help Google Cloud better compete with Microsoft Azure in providing AI capabilities for applications.
The timing of Gemini’s release this fall is crucial, as Google faces pressure to keep up with the rapid advancement of many big names in AI. It is estimated that to create an AI like Gemini, the cost can reach hundreds of millions of dollars, not to mention the time and effort spent.