By Michael Scott
Artificial intelligence is witnessing an extended drama involving Stability AI just days after Microsoft-backed OpenAI experienced a bumpy experience nearly a fortnight ago.
Will Stability AI Replicate Altman Ouster and Return at OpenAI?
Stability AI, famous for Stable Diffusion, seeks a sale as disagreement whirls around Emand Mostaque. It almost replicates the ouster of Altman from the company he co-founded alongside Greg Brockman, who has now resigned. The board accused Sam Altman of inefficient communication, leading to his ouster before staging a Microsoft-engineered return.
A Wednesday report shows that due to calls for chief executive officer Emad Mostaque to resign and company vice presidents to exit the company due to its utilization of copyrighted information, Stability AI, a UK-founded artificial intelligence (AI) image developer, might be seeking a purchaser. The firm is behind Stable Diffusion, a famous generative AI model, which incorporated video elements last week.
Technology Investment Firm Challenges Mostaque’s Management
In October, Coatue Management, a U.S. technology investment firm, sent a letter to Stability AI requesting information regarding top executives’ pay, which included Mostaque. The company challenged Mostaque’s management, claiming he had made several senior managers resign.
A report by data analytics site PYMNTS shows that despite no one coming forward to acknowledge the ownership invitation of Stability, two firms that have emerged as potential purchasers are Cohere, a Canadian artificial intelligence developer, and Jasper, an artificial intelligence marketing tool developer. Despite a request for comment, Stability AI failed to respond.
Stability Diffusion was unveiled in August last year and utilizes generative artificial intelligence to generate images based on user prompts.
PYMNTS shows that Stability AI’s value was $1B at the time. Like Dall-E and Midjourney, Stable Diffusion has been contentious since these kinds of tools have been utilized to generate artificial intelligence-generated deepfakes of events and international leaders.
Mostaque Steers Stability AI to Safety-Oriented AI Development
Three months ago, Cohere and Stability AI signed onto the Biden Administration’s vow to develop artificial intelligence safely. Additionally, Stability took part in November’s United Kingdom Safety Summit in Bletchley Park in Buckinghamshire, England.
During the meeting, the European Union (EU) and 29 nations signed a declaration to avert the unregulated development of artificial intelligence. Despite Stability signing the pledge, the firm’s position on what entails fair utilization resulted in some senior officials resigning.
Stability AI Suffers Key Executive Resignation as Leadership Plunged into Conflict
Earlier in November, Ed Newton-Rex, Stability’s VP of Audio, left because of the firm’s reasoning for training artificial intelligence models with copyrighted information.
On X (formerly Twitter), Newton-Rex said his exit from his leadership role at Stability AI was because of disagreeing with the firm’s stance that using copyrighted works to train generative artificial intelligence models is ‘fair use.’
In an interview with a media outlet, Newton-Rex claimed his decision to resign from Stability and slam the practice of utilizing copyrighted material without consent was to highlight a vast industry problem.
Newton-Rex cited a 22-page comment regarding generative artificial intelligence submitted by Stability AI to the United States Copyright Office. The comment called upcoming technology ‘a transformative, adequate, and socially helpful utilization of content safeguarded by fair use.’
According to Newton-Rex, this objection was not against Stability since it embraces a similar tactic that most generative artificial intelligence firms take. The issue is a cross-industry stance that he disagrees with. He also said he was leaving a big group of firms that take a similar approach.
Meanwhile, Stability AI realized partial victory in a court case exploring innovation and copyright. The plaintiffs alleged Stability AI paid another entity to train Stable Diffusion by relying on billions of copyrighted images without seeking artists’ authorization.
The case development echoes Newton-Rex’s perspective, which emphasized the need for AI to portray consciousness to artists and rethink fair use provisions when incorporating creative processes.
Editorial credit: rafapress / Shutterstock.com