Google Unveils Project Gemini—For a More Human AI Model

Google unveiled Gemini, an AI model trained to think and behave more like a human. Less sophisticated versions, “Nano” and “Pro,” will be incorporated into Google’s AI-powered chatbot Bard and Pixel 8 Pro smartphone.

Gemini will enable Bard to be more intuitive and better at tasks involving planning. For example, on the Pixel 8 Pro, Gemini will quickly summarize recordings made on the device and provide automatic replies on messaging services beginning with WhatsApp.

Gemini goes big next year with an Ultra model that will be used to launch “Bard Advanced,” initially offered to a test audience. “Bard Advanced” may be capable of unprecedented AI multitasking, simultaneously recognizing and understanding text, photos, and video presentations.

Google acquired London-based DeepMind a decade ago and melded it with its “Brain” division to focus on Gemini’s development. Its problem-solving skills are said to be adept in math and physics, potentially leading to scientific breakthroughs that improve human life.

“We’re approaching this work boldly and responsibly,” Google CEO Sundar Pichai wrote in a blog post. “That means being ambitious in our research and pursuing the capabilities that will bring enormous benefits to people and society, while building in safeguards and working collaboratively with governments and experts to address risks as AI becomes more capable.”

With Gemini’s rollout, OpenAI may try to prove its technology remains smarter than Google’s. For now, Google dodged questions about how Gemini compared to GPT-4 and declined to share specific technical details, promising to release more information about Gemini in an upcoming paper.