Google (GOOGL) and Microsoft (MSFT), the two tech giants dominating the artificial intelligence space, have been in an arms race since the beginning of the year to be a leader in generative A.I., a term popularized by the success of OpenAI’s ChatGPT. Yesterday (April 25), both companies reported better-than-expected earnings for the first three months of 2023. During calls with analysts afterward, Google CEO Sundar Pichai and Microsoft CEO Satya Nadella provided much-anticipated updates on where their respective A.I. efforts stood at the end of March and the plans for the future.
Google is building chatbots that are smart enough to replace coders.
On the new Bard: Google introduced the Bard chatbot, its answer to ChatGPT, in February and released an updated version earlier this month to include coding capabilities as requested by users. The new Bard is powered by Google’s Pathways Language Model (PaLM), which is trained on a larger database than the previous language model, LaMDA.
“Bard can now help people with programming and software development tasks, including code generation. Lots more to come,” Pichai said on the earnings call.
Pichai said Google’s future A.I. strategy will focus on building better language models and creating A.I. applications for both developers and business clients. He said a number of organizations are already using Google’s language models across Google Cloud, Google Workspace and its cybersecurity products.
On A.I. in search: There is a lot of speculation around when Google will incorporate generative A.I. in its search engine. The New York Times reported on April 16 Google is secretly working on A.I. search features in a project codenamed “Magi” and could have an initial rollout as soon as May.
Pichai didn’t go into detail when asked about his plan with Google Search, but said the company will “continue to incorporate generative A.I. advances to make search better in a thoughtful and deliberate way.”
Microsoft sees user growth following OpenAI partnership.
On OpenAI collaboration: Instead of building its own language model, Microsoft’s A.I. approach focuses on its partnership with OpenAI. Microsoft was already a major investor in OpenAI because ChatGPT went viral. In January, the tech giant followed up with a $10 billion, multi-year partnership with OpenAI and launched two new products, a new bing search engine powered by GPT and Azure OpenAI Service, a cloud platform that hosts A.I. applications built with OpenAI’s models.
Nadella said Azure OpenAI has gained more than 2,500 customers since launch in January. “We have the most powerful A.I. infrastructure, and it’s being used by our partner, OpenAI, as well as Nvidia and leading A.I. startups like Adept and Inflection to train large models,” Nadella said on yesterday’s earnings call.
On GPT-enhanced Bing: Nadella said Bing currently has 100 million daily users and downloads of the search engine have jumped since Microsoft incorporated GPT into the product. Daily installs of the Bing mobile app have grown four times since January, he said.
On developer-facing A.I. tools: Like Google, Microsoft is also developing A.I. applications that include coding capabilities. Nadella said he will share how the company is “building the most powerful A.I. platform for developers” at its Build conference, scheduled for May 23-25. He also hinted at an image generator using OpenAI’s Dall-E model.
“We continue to innovate with first-of-their-kind A.I.-powered features, including the ability to set the tone of chat and create images from text prompts, powered by Dall-E,” Nadella said.