David Wadhwani, senior vice chairman of digital media for Adobe, speaks throughout the launch of Adobe Creative Cloud and CS6 in San Francisco on April 23, 2012.
David Paul Morris | Bloomberg | Getty Images
Adobe on Tuesday launched a man-made intelligence assistant in its Reader and Acrobat applications that can produce summaries of and answer questions on PDFs and other documents.
The AI assistant, currently in beta, is now available on Acrobat, “with features coming to Reader over the approaching days and weeks,” based on a news release. Adobe plans to release a subscription plan for the tool after it’s out of beta.
The AI assistant will help users digest information from long PDF documents by generating transient overviews of their contents, the corporate said. The assistant can also answer questions on the knowledge in a document through a “conversational interface,” and suggest questions on the file that users might ask.
Adobe said the AI assistant can also generate citations that allow users to confirm the source of the tool’s answers, and can produce text for various formats reminiscent of emails, presentations and reports, based on the news release.
Other AI models reminiscent of ChatGPT offer PDF readers that similarly expedite analyses of lengthy documents, but those services require users to upload a PDF. Adobe’s AI assistant is a built-in feature.
In an interview on CNBC’s “Squawk on the Street” on Tuesday, Adobe CEO Shantanu Narayen said the brand new tool represents the corporate’s goal to “democratize access” to the trillions of PDFs in use.
“Just imagine you have opened a 100-page document. You must understand the summary, you desire to have a conversation with it, you desire to ask questions,” Narayen said. “You must correlate that with other documents that you would possibly have in addition to all the information that you will have in your enterprise.”
Last week, OpenAI, which developed ChatGPT, launched a latest tool that generates realistic, high-definition video off a text prompt. Responding to a matter about whether OpenAI’s model, called Sora, represents an encroachment on Adobe’s turf, Narayen said the corporate is “working on our video models as well” and intends to use that technology “in a responsible way” to “tools and workflows.”