In an interview, Pichai discussed the potential of Gemini, its comparison with OpenAI's GPT-4, and the challenges and opportunities of AI technology. He emphasized the importance of responsibly deploying AI technology, addressing its potential negative impacts, and ensuring its accessibility. Pichai also touched on the legal landscape around AI, the future of search, and the expected changes in user experience with the introduction of AI-powered products.
Key takeaways:
- Google has released the first phase of its next-generation AI model, Gemini, which reflects years of efforts from inside the company, overseen by CEO Sundar Pichai.
- Gemini is a natively multimodal model, learning from text, audio, and code, and is state of the art in 30 of the 32 leading benchmarks.
- The model will be integrated across Google's line of products throughout 2024, and is part of a larger shift towards AI, which Pichai believes is a profound platform shift, bigger than web or mobile.
- Despite the potential benefits, Pichai acknowledges the challenges and potential negative impacts of AI, including job displacement and safety concerns, and emphasizes the need for responsible deployment and societal adaptation.