Last week at its annual developers conference, Google I/O, Google announced several artificial intelligence (AI) products, pointing towards its commitment to expanding its dominance in the space. There’s the new MusicLM AI that can make music out of text, Sidekick to help you compose better documents in Google Docs, and Magic Compose, which makes your text more concise, professional, or even make it sound like William Shakespeare.
One of the big reveals was a competitor to OpenAI's GPT-4, called Gemini. This is Google’s next-generation large language model (LLM) and comes next in line after its existing PaLM 2 system, with an improved machine-learning model.
According to Google, Gemini was created from the ground up to be multimodal, highly efficient at tool and API integrations and built to enable future innovations, like memory and planning. In a blog post, Google said that it is already seeing impressive multimodal capabilities not seen in prior models.
“Once fine-tuned and rigorously tested for safety, Gemini will be available at various sizes and capabilities, just like PaLM 2, to ensure it can be deployed across different products, applications, and devices for everyone’s benefit,” Google said in the blog post.
PaLM 2, introduced at the conference, is a “state-of-the-art language model with improved multilingual, reasoning and coding capabilities.” It has been heavily trained on multilingual text, with more than 100 languages. It can generate and translate nuanced text, including idioms, poems and riddles, across a wide variety of languages.
Gemini is considered to up Google's AI game and give tough competition to the one that started the race, Open AI’s ChatGPT. While the popular ChatGPT is a text-only model, meaning that it can only understand and generate text, Gemini is multimodal which means that it is not limited to only text and could respond to images. Integrated with Google search, it could enhance the way people work with the popular search engine.
More detailed information on Gemini is yet to be released. For now, it looks like Gemini could outperform ChatGPT and Bing AI, pulling Google ahead in the AI race. As Google CEO Sundar Pichai said at the conference, “Seven years into our journey as an A.I.-first company, we are at an exciting inflection point.”