Google unveils ‘magic wand’ to draft documents as AI race tightens
By Jeffrey Dastin
PALO ALTO, California (Reuters) – Alphabet Inc’s Google on Tuesday unveiled a flurry of artificial intelligence (AI) tools for its email, collaboration and cloud software, taking aim at Microsoft Corp days before its rival is expected to make a similar announcement.
In a replay of last month’s dueling chatbot launches by the tech giants, Alphabet touted a “magic wand” for its popular Google Docs software that can draft a marketing blog, training plan or other text, then revise its tone at users’ discretion, a company official demonstrated to reporters.
Microsoft, meanwhile, has teased a Thursday event about how it is “reinventing productivity with AI,” which is expected to showcase its competing Word processor.
Alphabet also said its AI will be able to summarize message threads in Gmail, craft slide presentations, personalize customer outreach and take meeting notes as part of its upgrade to Google Workspace, a product suite with billions of users on free and paid accounts.
The advances reflect how ChatGPT has spurred a race in Silicon Valley to imbue products with so-called generative AI, which learns from past data how to create content anew, just like the chatbot sensation.
Microsoft, Alphabet and peers are investing billions of dollars to build and deploy the technology, hoping business they win from speeding up writing and creative tasks for office workers will far outweigh the costs of these endeavors.
“This next phase is where we’re bringing human beings to be supported with an AI collaborator, who is working in real time,” Thomas Kurian, Chief Executive of Google Cloud, said in a press briefing.
Alphabet is giving approved test users access to new Workspace features on a rolling basis throughout the year, before a wider launch, similar to it and Microsoft’s phased release of their chatbot programs.
Kurian declined to say how much the upgraded Workspace might cost businesses or consumers.
Google also unveiled a range of generative AI tools for its cloud-computing customers, for instance previewing access to PaLM, one of its most powerful “large language models” that create human-like text.
Google said customers can fine-tune its AI model with their own data while keeping the information and benefits proprietary.
In another enterprise software example, Google showed how a fictional furniture business could build better customer-service chatbots capable of generating images as well as text, like showing how a corgi dog would look on a mid-century modern chair.
The chatbot could integrate with a payments system so a shopper then could buy the chair, a promotional video showed.
Google aims for its AI to “transform” the work of marketers, lawyers, scientists and educators, according to the video.
The Mountain View, California-based company announced a partnership with high-profile AI research lab Midjourney, with Google to provide cloud infrastructure including its custom “TPU” chips.
Microsoft’s generative-AI rollout so far has outpaced that of Alphabet, which is wary of societal harm as well as damage to its reputation as a reliable source of information.
Such software remains prone to inaccurate responses known as “hallucinations”.
A factual error that Alphabet’s chatbot Bard made in a demo last month contributed to a $100 billion slump in its market value, though Microsoft drew scrutiny of its own when its Bing search chatbot professed love or made threats to test users.
Kurian said Google remains “deeply committed to responsible AI,” giving controls to customers and reviewing proper use of its products. Microsoft has also added safeguards to its search software.
(Reporting By Jeffrey Dastin in Palo Alto, California; Editing by Jamie Freed)