AI is the most advanced technology being developed today. AI assists people, businesses, and communities in realizing their full potential, whether it's assisting physicians in detecting illnesses sooner or allowing people to access information in their native language. And it creates new opportunities that have the potential to greatly enhance the lives of billions of people. That's why, six years ago, we refocused the firm on AI, and why we regard it as the most crucial method we can fulfill our mission: to organize the world's knowledge and make it widely accessible and valuable.
Bard AI
Since then, we've continued to invest in AI across the board, with Google AI and DeepMind pushing the technology forward. Today, the most extraordinary AI calculations are doubling in size every six months, considerably surpassing Moore's Law. Simultaneously, powerful generative AI and big language models are catching people's minds all across the world. In fact, our Transformer research project, our field-defining publication from 2017, and our significant breakthroughs in diffusion models are now the foundation of many of the generative AI applications you're seeing today.
It's an exciting moment to be working on these technologies because we're able to turn deep research and discoveries into solutions that genuinely assist people. That has been our experience with huge language models. We debuted next-generation language and conversation capabilities enabled by our Language Model for Dialogue Applications (or LaMDA for short) two years ago.
We've been developing Bard, an experimental conversational AI service powered by LaMDA. And now, we're taking another step forward by making it available to dedicated testers ahead of its general release in the coming weeks.
Bard aspires to bring together the richness of the world's information with the strength, intelligence, and creativity of our massive language models. It uses online information to give new, high-quality replies. Bard may be a creative outlet and a jumping-off point for inquiry, allowing you to explain fresh findings from NASA's James Webb Space Telescope to a 9-year-old, or learn more about the finest strikers in football right now, and then receive workouts to improve your talents.
We're starting with our lightweight model version of LaMDA. This much smaller model uses much less computational power, allowing us to grow to more people and provide more feedback. We'll combine external input with internal testing to ensure that Bard's solutions reach a high standard of quality, safety, and groundedness in real-world data. We're looking forward to this round of testing since it will allow us to continue learning and improving Bard's quality and speed.
0 Comments