Pichai recalls a defining moment when he witnessed the groundbreaking work at DeepMind, an AI research lab, which focused on understanding the essence of intelligence. This research significantly influenced his belief that Google should evolve into an AI-first company.
Many questions have followed in Google’s AI journey. Questions like, “How can we power this new generation of computing?” led to the invention of Tensor Processing Units (TPUs), providing substantial performance improvements for machine learning. These TPUs played a vital role in AlphaGo, the AI system that defeated the Go world champion Lee Sedol in 2016. Simultaneously, Google released the Transformer paper in 2017, laying the foundation for most generative AI and large language models in use today.
These technological breakthroughs have driven some of Google’s most remarkable product innovations. Search results for complex queries have improved significantly thanks to large language models like MUM and BERT. Google has introduced novel ways for people to express their search intent through voice, images, and even questions about the content they see with Multisearch. Generative AI is now revolutionizing core products, from the Search Generative Experience (SGE) to the “Help Me Write” feature in Gmail. Earlier this year, Google launched Bard, an experimental tool allowing people to collaborate with generative AI.
Pichai’s most cherished moments stem from witnessing how Google’s products make a difference in people’s lives. Whether it’s a busy parent using Bard to inspire a rainy day project for their children, a traveler translating a foreign train schedule with Lens, or, more profoundly, a family hearing the voice of an ALS patient thanks to Google’s speech recognition and synthesis research, these experiences demonstrate the real-world impact of responsible technology design.
Designing products that help people at scale is both a privilege and a responsibility. Questions arise: Can we trust these new technologies? Google places great emphasis on building responsible technology from the outset, ensuring that user information is secure and protected and that people are safeguarded from malicious actors online.
This commitment extends to engaging in essential debates about how these technologies will shape our society. AI, in particular, is a focal point. Pichai acknowledges that, while Google is excited about AI’s potential benefits, the technology also poses complexities and risks. Google’s development and utilization of AI must address these challenges while fostering responsible technology advancement. The AI principles introduced by Google in 2018 serve as a guide for ensuring that AI applications are beneficial and free from harm.