Amazon launches new chatbot with AI tech behind ChatGPT

Amazon on Thursday said it has launched its own chatbot using artificial intelligence that is similar to ChatGPT and has rolled out new tools that allow businesses to build their own chatbots.<br><br>The tech giant has also built out its bedrock AI technology called Titan with the addition of a new model called Gemini. <br><br>The move highlights Amazon’s foray into what is considered the next era of computing guided by generative AI. Rivals including Google, Microsoft and Meta have been working on guided AI for years. <br><br>The new Amazon chatbot is called Tricia and is considered an "AI companion" by the company. It uses Gemini to power its capabilities. <br><br>Gemini is based on Titan and has been built by the company's OpenSearch division. The powerful model is available to its cloud computing division, called AWS, and can be accessed by companies who are looking to build their own chatbots. <br><br>The OpenSearch division is a group that is working across the company on building AI capabilities, including generative AI tools like Gemini. <br><br>"Titan and Gemini models are part of a larger suite of tools that we're building as part of Project Gemini," Swami Sivasubramanian, a vice president of AWS, said in an interview. "We're building several things around Gemini." <br><br>Gemini is part of a new set of tools that are being rolled out, including a framework called Gemini Studio. Gemini Studio is designed to help companies build their own chatbots. That includes a makerspace that allows companies to test out different features on their bots. <br><br>"We're going to continue to build out Gemini and make sure it's the most accurate model," Sivasubramanian said. <br><br>Amazon is also launching three new tools that companies can use to build out their own chatbots with Gemini, including an editor, maker space and a deployment studio. <br><br>While Tricia can do a range of tasks with Gemini, companies that build their own chatbots with Gemini Studio can customize them to meet specific needs, like building out a customer service chatbot for a company website. <br><br>"It really depends on what you want," Sivasubramanian said of the capabilities of chatbots being built with Gemini Studio.