Google is expected to show the world Gemini AI in the fall of 2023. The new AI model will be the answer to ChatGPT and GPT-4 when developing LLM features, for example. about creating images.
According to new data, Google is preparing for the grand premiere of the AI Gemini tool - the debut of the product should take place this year. Gemini is set to compete and even beat ChatGPT in the realm of conversational AI tools based on large language models.
Google Gemini AI - premiere in 2023?
An anonymous source The Information revealed additional details about the Google Gemini AI project. According to the source, the AI model should debut in the fall of 2023 and compete with the company’s main competitor, the GPT-4 model behind ChatGPT.
Google Brain and DeepMind specialists are responsible for the project. The main goal of engineers seems to be to combine in one AI model both the capabilities of large language models (LLMs responsible for the operation of AI chatbots) and algorithms that generate images from text commands.
Thus, Google wants to provide a comprehensive tool that performs as many tasks as possible, while currently users have to use several separate systems. According to the source, Gemini is currently being trained on transcripts of YouTube videos, albeit only those that are not protected by copyright.
Gemini AI is how Google wants to defeat ChatGPT
As we wrote in Wprost.pl, DeepMind CEO Demis Hassabis speaks about Gemini with great courage. This is a new foundation that can be used to create new AI tools. For example - ChatGPT uses the GPT-4 model and it is this model that allows the tool to create correct answers to numerous user questions.
Why does the Google Twins have to be better than the competition? The manager says that it will be a huge advantage for him to include another system in the model, which was previously developed by DeepMind. We are talking about AlphaGo, an artificial intelligence that many years ago became famous for defeating the best Go players.
The new component should bring to the developed artificial intelligence system the ability to plan and solve very complex tasks. AlphaGo is based on the reinforcement learning method. This means that after each failure, the machine receives feedback on the result and gradually changes its actions, all the while improving them. This is supposed to be the key to gradually surpassing ChatGPT.
Source: Wprost
I have been working as a news website writer for over 10 years. I have also authored several books on technology news. I am highly knowledgeable in the field of technology and enjoy sharing my insights with others. I am an active user of LinkedIn and use it to connect with other professionals in my field. I am always interested in learning new things and sharing my knowledge with others.
