Google Introduces New Model to Enhance AI Chatbot’s Understanding of Multimedia
In a bid to enhance its AI chatbot, Bard, Google has unveiled a new model called Gemini. This model aims to improve the chatbot’s understanding of video, audio, and photos, allowing it to mimic human communication abilities like speech and imagery.
Gemini has been released in dozens of countries, but currently, it is only available in English. This new technology enables the chatbot to perform complex tasks such as summarizing documents, reasoning, and even writing programming code.
Google has divided Gemini into three versions: Nano, Pro, and Ultra, each targeting different levels of computing power. Gemini Nano is designed for mobile phones and will power new features on Google Pixel 8 phones. For instance, it will summarize conversations in the Recorder app and suggest message replies in WhatsApp.
Gemini Pro, on the other hand, runs in Google’s data centers and powers a new version of the Bard chatbot. As for Gemini Ultra, it is currently limited to a test group but is expected to be available in an advanced version of Bard by early 2024.
This marks the third major revision of Google’s AI models in the field of generative AI, where chatbots create their responses based on prompts. With its training on text, programming code, images, audio, and video, Gemini efficiently processes multimedia input.
Gemini’s capabilities are impressive, as it can correctly guess the next shape in a series, connect photos of the moon and a golf ball to the Apollo astronauts’ golfing on the moon, and even convert bar charts into labeled tables. However, Google warns that the chatbot may display inaccurate information and urges users to double-check its responses.
While a demonstration video released by Google showcased Gemini’s abilities through visual data, it is important to note that the chatbot is capable of accepting spoken and video input as well.
Looking ahead, Gemini Ultra will undergo further testing for security vulnerabilities through a process called “red teaming” before its official launch in 2024. Google emphasizes its commitment to approaching AI advancements responsibly and collaboratively with governments and stakeholders to address any risks that may arise as AI becomes more capable.
With Gemini, Google aims to improve the capabilities of its AI chatbot and provide users with a more seamless and comprehensive communication experience.
“Zombie enthusiast. Subtly charming travel practitioner. Webaholic. Internet expert.”