After recently combining the Brain Team in Google Research and DeepMind into one unit, Google still aims to take a “safe and responsible” approach to developing its AI systems, CEO Sundar Pichai said on Alphabet’s Q1 earnings call.
The company is also adhering to its commitment to “invest responsibly and with discipline,” with machine learning being one of the main areas where it strives to operate more cost-effectively and “with greater velocity,” according to the CEO.
“We have significant multiyear efforts underway to create savings, such as improving machine utilization and finding more scalable and efficient ways to train and serve machine learning models. We are making our data centers more efficient, redistributing workloads and equipment where servers aren’t being fully used. This is important work as we continue to significantly invest in infrastructure to drive our many AI opportunities.”
Since the company released its ‘Bard‘ conversational AI in March, it has made advances such as adding its PaLM model to the Bard, owing to the consolidation of the DeepMind and Brain teams. The API allows Bard to assist in tasks relating to programming and development, including code generation.
Pichai said on the April 25 earnings call that Google has also released its PaLM API and its new MakerSuite tool for developers to allow them access to the company’s large language models and begin building new generative AI applications. He added:
“A number of organizations are using our generative AI large language models across Google Cloud platform, Google Workspace and our cybersecurity offerings.”
Pichai said the company has added generative AI in its cloud offerings and made them available to Cloud customers, noting that Google is the only cloud provider with NVIDIA’s new L4 Tensor Core GPU with its new G2 Virtual Machines, which are “purpose-built for large inference AI workloads, such as generative AI.”
Alphabet recorded $7.45 billion in revenues for the Google Cloud segment for Q1 2023, up from $5.82 billion in the year-ago quarter and Pichai noted that 60% of the world’s 1,000 largest companies are Google Cloud customers.
New experiences in search and beyond
The CEO said Google aims to “unlock entirely new experiences in Search and beyond” as it develops its AI capabilities and as they evolve “just as camera, voice and translation technologies have all opened entirely new categories of queries and exploration.”
Responding to a question from JPMorgan’s Douglas Anmuth on integrating Bard into Search products, Pichai said that the company has launched Bard as a complementary product to Search but expects to bring large language model (LLM) experiences “more natively into Search.”
However, he noted that the company plans to roll such experiences out in an “incremental way” in order to test it, create and innovate.
“I think overall, I think it can apply to a broad range of queries. So I think I’m excited that it can allow us to better help users in a category of queries, maybe in which there was no right answer, and they are more creative, et cetera. So I think those are opportunities. But even in our existing query categories, where we get a chance to do some heavy lifting for the users and we use AI to better give — guide them, I think you will see us exploring in those directions as well. It’s early days, but I think there’s a lot of innovation to come.”
Pichai expressed excitement about the potential for AI to help people and businesses and said the company will share updates at Google I/O on how it is using AI across its products, including the Pixel devices, and will also share some new developments for Android.
“We have used AI to open up access to knowledge in powerful ways. We’ll continue to incorporate generative AI advances to make Search better in a thoughtful and deliberate way. We’ll be guided by data and years of experience about what people want and our high standards for quality. And we’ll test and iterate as we go because we know that billions of people trust Google to provide the right information.”