LAS VEGAS – With artificial intelligence (AI) disrupting everything from customer service to agriculture, it was only a matter of time before it reached more creative sectors like graphic design – and according to the creative software gurus at Adobe Systems Inc., their practitioners can look forward to reaping substantial benefits too.
During the Oct. 18 keynote, and a post-keynote Q&A, at the company’s annual Adobe Max conference, Adobe CTO Abhay Parasnis and CEO Shantanu Narayen said that while the company is always cognizant of the potential risks in adding ever-more-advanced AI features to its Creative Cloud platform, their ultimate goal is not to endanger users’ jobs, but to create tools that allow them to concentrate on what they do best – improving their latest projects.
“Without question, AI represents the most disruptive paradigm shift of the next decade,” Parasnis said. “From voice assistants to self-driving cars, it’s already taking over the world. And it’s bound to change how all of us work in profound ways.”
“Some in the industry believe that AI may ultimately replace humans for a lot of activities,” he continued. “At Adobe, we have a very different viewpoint… Our fundamental belief is that [AI] will amplify human creativity and intelligence, not replace it. It will free you from doing repetitive, mundane tasks.”
The comments were delivered one year after the release of Adobe’s own AI platform, Sensei, which some analysts were dismissing as poorly defined as recently as six months ago. But Sensei powered many of the new features added to what the company called its largest-ever Creative Cloud update on Wednesday, such as an Adobe Typekit feature that allows users to identify new fonts by shooting a photo of one they like, or the machine learning-based search added to its Lightroom software (which as of Wednesday has been rebranded Lightroom Classic).
It also powered many of the features presented during the company’s Adobe Sneaks event on Thursday, such as Scene Stitch, which uses Sensei to replace the unwanted part of an image with something appropriate from Adobe’s stock photo library.
Sensei’s developers are currently focused on three domains, Parasnis said: compositional creativity, which allows Creative Cloud to better assist artists and designers with their imaging, video, design, and illustration tasks; experience intelligence, which allows businesses and their brands to better personalize and measure the engagement of the great experiences they create; and content understanding, incorporating deep learning techniques to search, tag, and extract meaning from a wide range of creative assets and documents.
One of Sensei’s most important features, Parasnis and CEO Narayen said, is that when it fails to work properly – for example, by replacing the unwanted part of a Scene Stitched image via with something nonsensical – users can report it, allowing Sensei to learn from the Adobe community and ultimately creating a smarter, more accessible program.
“We’ve built Sensei with two very unique approaches,” Parasnis said. “First, it learns from hundreds of millions of assets that are managed within our creative tools. Second, it is purpose-built, with a very deep understanding of the creative and design domains, learning from all of you who use our tools every day. In effect, each of you are essentially training the Sensei creative fabric.”
“For us, AI is really… about harnessing the collective intelligence of the community,” Narayen said. “And the more we can learn from people – what features they’re using, how they’re using it, what it means to understand objects and images and video… I think we dramatically change the number of people who can use our creative tools to tell a story.”
Not The Terminator
During the post-keynote Q&A, Parasnis responded to a question about the growing capabilities of AI (related to the undeniable realism of certain Sensei-produced images, and whether it would necessitate the release of a program that could detect whether an image had been doctored) by emphasizing that Adobe’s developers are aware of the risks and ask themselves similar questions and consider the real-world implications of every feature they create.
“Sometimes internally we talk about… the Terminator fear of the future where machines are going to take over and everything will be doom and gloom,” he said, referring to the Arnold Schwarzenegger-led action movie franchise. “We actually don’t believe in that scenario.”
“Yes, AI and machine learning can create a very powerful tool to improve productivity… but internally we joke that we believe more in the Harry Potter vision of the future, where it’s more magical, experiences are more complimenting what users are capable of,” he continued.
The reality, Parasnis said, is that today’s tools can already create much of what Sensei is streamlining, and tools that keep track of image edits are already built into the Creative Cloud platform.
As for the longstanding fear that AI could decimate thousands or even millions of jobs over the next 20 years, Parasnis repeatedly made it clear that Adobe falls into the camp that believes AI will ultimately create more jobs than it destroys.
“When it comes to AI, we definitely think we have a role to play in being mindful, but it is more about helping our users, not replacing what they do today,” he said.