Can it change the situation? Adobe launches AI video generator to compete with OpenAI and Meta
Adobe launched a video editing tool with integrated generative artificial intelligence at its annual product conference in Miami, allowing users to extend video clips. The tool is designed to compete with rivals such as OpenAI and Meta, making Adobe the first major software company to widely offer such generators to customers. Ely Greenfield, Chief Technology Officer of Adobe Digital Media, stated that customers want to incorporate AI capabilities into existing applications, and Adobe's new model is specifically designed for real work processes
According to the financial news app Zhitong Finance, Adobe (ADBE.US) has launched an AI tool for creating and editing videos, joining large tech companies and startups in tapping into the demand for emerging technologies.
Adobe announced at its annual product conference in Miami on Monday that its video editing software Premiere now integrates a feature that allows users to expand video clips using generative artificial intelligence. Other online tools enable users to create videos based on text prompts and existing images.
While OpenAI, Meta Platforms (META.US), and Google have all showcased AI video generators, Adobe is the first major software company to widely offer this generator to customers. Some startups like Runway AI have already publicly released their video generation products.
Ely Greenfield, Chief Technology Officer of Adobe Digital Media, commented on competitors' technologies, saying, "When we talk to customers, what we hear is that these technologies are very cool, but they can't use them."
Greenfield stated that customers want AI features integrated into the applications they already use, and Adobe's new video model is "designed specifically for real workflows and tool integration."
Over the past year, Adobe has been focused on adding generative AI features to its software suite for creative professionals, including flagship products like Photoshop and Illustrator. The company has released tools for generating images and illustrations from text, which have been used billions of times to date.
Adobe has been working to differentiate its models as "business-safe" by training data carefully and conducting strict reviews. For example, if a user tries to generate videos of certain faces, Adobe will block those faces. Competitors have faced criticism for extensively scraping internet data to build AI models.
Greenfield mentioned that Adobe's video model is primarily trained on videos and photos from its vast stock media library provided to marketers and creative agencies. He added that in some cases, the company uses public domain or licensed data. Adobe has proposed purchasing videos from its network of creative professionals at a rate of approximately $3 per minute.
Earlier this year, OpenAI showcased its video generation model Sora, sparking concerns among investors about Adobe potentially being disrupted by new technologies. As of last Friday's close, Adobe's stock price has fallen by 17% this year. At the time of writing, the stock rose over 2% in early trading.
Adobe has not yet charged for AI features beyond standard subscription fees. Greenfield mentioned that each user will receive a certain amount of AI substitute points, but most plans do not enforce these restrictions. Company executives indicated that in the future, Adobe may charge higher fees for using AI focused on videos compared to similar tools for photos Adobe also announced improvements to other software at the conference, such as making 3D content viewing in Photoshop easier. The company is also committed to developing AI models that can generate 3D graphics