Adobe expands AI models training with help from artists, photographers

Getting your Trinity Audio player ready...

Adobe (NASDAQ: ADBE) has announced plans to begin the development of an artificial intelligence (AI) model to rival industry first-movers but seeks help from a network of photographers and artists.

Adobe is keen on developing a text-to-video generator to add to its growing list of AI products. While details for the new video generator are sparse, it appears that model training has taken off with Adobe using its library of stock media for the initiative.

To stand apart from its competitors, Adobe is seeking to expand the scope of training data for the model by turning to its network of photographers. The firm is willing to pay contributors for content to augment any gaps that may crop up in training its AI model.

For each photograph submitted, Adobe will pay anywhere between 6 cents and 16 cents, while each minute of video content may fetch up to $2.62. Video content can earn as much as $7.25 per minute, with the software company keen on building its repository of simple shots of feet, hands, and other facial features.

The videos to be submitted range from the mundane like walking or emotional expressions of joy and anger, to complex actions like operating machinery.

Despite seeking a large pool of content, Adobe says submitted images or videos should not contain offensive content or nudity. The company is also keen on protecting copyrights, learning from the pitfalls of industry first movers that are currently plagued by a slew of intellectual property (IP) claims.

Adobe’s incoming video-based generator is not the company’s first rodeo with AI, as its frontline model, Firefly, has seen significant action since its launch in 2023. The firm’s flagship products, Reader and Acrobat, have received their fair share of AI features as the competition heats up.

“Early adopters like IBM are putting Firefly at the center of their content creation processes,” said Wadhwani. “IBM used Adobe Firefly to generate 200 campaign assets and over 1,000 marketing variations in moments rather than months.”

SORA sends ripples down the ecosystem

OpenAI’s announcement of a text-to-video generator may have forced Adobe to hasten the development of its offering. Experts opine that OpenAI’s Sora could upend Adobe’s dominance in the sector, with Sora being touted as a mainstay for the film industry.

Despite OpenAI’s pioneering move, Sora faces a barrage of questions over its training data, and given the recent SAG-AFTRA strike, the video generator could have an uphill climb in Hollywood.

Adobe’s waltz with emerging technologies has seen it dabble with blockchain technology and non-fungible tokens (NFTs) in a brave attempt to stay ahead of the curve.

In order for artificial intelligence (AI) to work right within the law and thrive in the face of growing challenges, it needs to integrate an enterprise blockchain system that ensures data input quality and ownership—allowing it to keep data safe while also guaranteeing the immutability of data. Check out CoinGeek’s coverage on this emerging tech to learn more why Enterprise blockchain will be the backbone of AI.

Watch: Does AI know what it’s doing?