The News: On October 10, 2023 during Adobe MAX, Adobe announced three new AI models: Adobe Firefly Image 2 Model, Adobe Firefly Vector Model, and Adobe Firefly Design Model. Together, they comprise the next major release in the company’s growing family of generative AI models.
- Firefly Image 2 Model: This next generation of the original Firefly Image Model generates higher-quality images including improved human rendering, better colors, and greater ability to control output. (These improvements are due to Adobe increasing the data on which Firefly is trained). Also, text-to-image capabilities are being added to the Firefly web app; it is notable that 90% of Firefly web app users are new to Adobe products. Also new: Generative Match enables users to apply the style of a user-specified image to generate new images at scale. The model includes improved text prompt capabilities, such as recognizing more landmarks and cultural symbols, suggestions for improved prompts, and a feature called Prompt Guidance, which teaches users to expand or reword prompts more effectively.
- Firefly Vector Model: Adobe says this is the world’s first generative AI model for vector graphics designed to generate “human quality” vector and pattern outputs. This model includes Generative Match, editable vector gradients, and seamless patterns that allow design patterns to be repeated infinitely without visible gaps.
- Firefly Design Model: This model supports the generation of template designs via Text to Template, which combines layout technology with Firefly Image Model, Adobe Stock, and Adobe Fonts. Users can generate images in Firefly, carry over the prompts to continue refining in Express, deploy automated designs, and more.
Read the full Adobe press release on the next generation Firefly models here.
Adobe Firefly: Blazing a Generative AI Application Trail
Analyst Take: As of today, Adobe Firefly is the most successful generative AI product ever launched. Since it was introduced in March in beta and made generally available in June, Firefly users have generated more than 3 billion images. Adobe says Firefly has attracted a significant number of new Adobe users, making it hard to imagine that Firefly is not aiding Adobe’s bottom line. Here are the key lessons to learn from Firefly’s success:
How Does Deploying AI Make a Company Better?
One of the primary reasons that Firefly is such a hit is that image generation is a significantly practical and logical use case for the creative industries, which have been Adobe’s primary target market and primary customer base for years. Adobe obeyed the first rule of AI, which is:
Every Company Should Ask Itself: What Problem Are We Trying to Solve With AI?
Image generation enables creative professionals to work more quickly, automating routine graphics work. What is more, Adobe says that the work of creative content generation must expand throughout organizations, and go beyond creative professionals to reach a wider range of marketers. Firefly is a direct way for Adobe to help companies accomplish that goal.
Does a Company Have the People and Processes to Execute AI?
From The Futurum Group’s June 2023 Research Note on Firefly:
It is difficult for companies to duplicate Adobe’s AI experience. Many Adobe competitors are likely not as far along in their AI experience. Some might be thinking about AI for the first time because of the Generative AI phenomena. To leverage AI, companies must step through the AI lifecycle process – what is AI, how does this AI help us, should we use this AI, what are the risks to using this AI, etc.
To further elaborate, Adobe has lived with AI for nearly 10 years. The company has had the opportunity to figure out, by trial and error, what was needed in terms of the people and processes to make AI work. By lifecycle, we mean that Adobe, through experience, understands the infrastructure needed to execute on AI. That has more to do with organizing people into the right teams, smartly setting up data management and data governance, and lastly, creating a framework to manage AI risk. None of those things have to do with the technology itself, but rather with developing the people and processes around the technology to point it in the right direction.
Time and Experience with AI Matter
Finally, experience with AI enables a company to move fast. Experience means a company has a better understanding of AI issues, particularly how they impact communities of interest.
We can point to many examples, the first regarding speed. The leaders of AI tech within Adobe recently told The Futurum Group how much they learned after the March beta launch of Firefly, particularly regarding heavy use. One lesson: Adobe needed to figure out how to run Firefly AI inference at scale. That finding was part of the impetus for Firefly Image Model 2, which launched in June and was followed by this October update―speedy responses to lessons learned.
In terms of understanding issues affecting communities of interest: As Firefly has rolled out, feedback from creators has poured in. Adobe knew that creators would need to feel compensated and protected when image generation rolled out. So, Adobe went to work on multiple fronts. Most obviously, Adobe is working with The Content Authenticity Initiative (CAI) to make sure creators who do not want their work to be part of generative AI training are protected (Note: Adobe Stock creators have signed agreements allowing their work can be used for AI training). In terms of compensation, a prickly issue, Adobe has a path through the Generative Credits program.
Adobe has also made indemnification pacts with its creators, pledging it will cover any legal costs for IP or copyright disputes. The company’s AI Art Indemnification Policy will “protect, defend and hold harmless” customers who license Adobe generative AI art that infringes on third-party copyrights. Further, the company has proposed new federal legislation called the FAIR Act, which would protect creators from people misusing AI to deliberately impersonate their work. Adobe is supporting inclusion of that language in potential AI regulatory frameworks being drafted by the White House and Congress.
Conclusions
The runaway success of Adobe Firefly is not a fluke. It is the result of a long-term investment by Adobe in the promise of AI. The methodology, culture, and process that Adobe brings to bear for leveraging AI are the methodology, culture, and process other enterprises should aspire to if they wish to leverage AI to their own benefit.
Disclosure: The Futurum Group is a research and advisory firm that engages or has engaged in research, analysis, and advisory services with many technology companies, including those mentioned in this article. The author does not hold any equity positions with any company mentioned in this article.
Analysis and opinions expressed herein are specific to the analyst individually and data and other information that might have been provided for validation, not those of The Futurum Group as a whole.
Other insights from The Futurum Group:
Firefly, Sensei GenAI Ascendent: Adobe Excels at Generative AI
Adobe and Google To Bring Photoshop and Express to Chromebook Users
Key Trends in Generative AI – The AI Moment, Episode 1
Author Information
Mark comes to The Futurum Group from Omdia’s Artificial Intelligence practice, where his focus was on natural language and AI use cases.
Previously, Mark worked as a consultant and analyst providing custom and syndicated qualitative market analysis with an emphasis on mobile technology and identifying trends and opportunities for companies like Syniverse and ABI Research. He has been cited by international media outlets including CNBC, The Wall Street Journal, Bloomberg Businessweek, and CNET. Based in Tampa, Florida, Mark is a veteran market research analyst with 25 years of experience interpreting technology business and holds a Bachelor of Science from the University of Florida.
