Artificial intelligence generated a lot of excitement this spring. Its influence spread from Silicon Valley to Hollywood, with a confluence of technologies introduced in Las Vegas at this year’s NAB Show.
Generative AI, in particular, caught the attention of content creators and consumers, manufacturers and media companies. Its promises range from improved efficiency to truer independence for creatives. In Q2 2023, we’re seeing workflows and products shaped by the potential of neural networks, while creatives are starting to take advantage of AI tools.
However, that excitement has come paired with several flashing ⚠️ caution ⚠️ signs. The ethics of AI (and we’re not just talking about deepfakes), concerns about inadequate regulatory frameworks, and ever-present worries about how machines may change the job market persist.
But one expert, ETC’s Yves Bergquist, tells us to consider the source when listening to these worries: “Are they accountable to their audience, or to their code?”
That is to say, what do those who are actually working on the algorithms and data sets and workflows think — and contrast that information with context from those who might benefit from controversy and hyperbole.
Another generative AI proponent and creative practitioner, Pinar Seyhan Demirdag, thinks these tools will ultimately find their footing in use cases determined by the four Ds: tasks that are dull, dirty, dangerous, or dear for humans to perform. (Rotoscoping comes to mind as an example of dull.)
Demirdag joined Bergquist at NAB Show to discuss generative AI workflows, as well as their thoughts for how this tech may shape the industry’s near-term future. Watch the session below.
Read on: Adrian Pennington also explores a lot of AI myths in his coverage for NAB Amplify. His breakdown of What Ifs vs. What Ares for generative AI is especially clarifying (as of April 2023, anyway).
These conversations are just beginning for the media and entertainment industry. But let’s take a quick look at some of the generative AI tools available now.
Tools on the Market
You’re most likely familiar with OpenAI’s ChatGPT and may even be a paying subscriber for GPT-4. DALL-E and Midjourney may have been in your arsenal since last summer. Many other companies have long integrated forms of AI and machine learning into their offerings, but here are a few newer generative AI tools targeted at M&E creatives.
Adobe: Firefly
Adobe announced its AI art generator ahead of this year’s NAB Show. The beta version of Firefly is focused on text and images, but the company expects to integrate its features across its video, audio, animation and motion graphics design apps in the near future.
Colourlab.ai
Colourlab promises film-quality color-grading without the drudgery. It pitches this AI tool as an assistant who frees up creatives to have more time and focus on the unleashing of our imaginations and a return to kid-like experimentation. There’s also a pitch for simplifying remote collaboration and reduced file storage for versioning (two other trends that we know aren’t going anywhere for M&E).
DeepBrain AI
Are you interested in beginning a “virtual human journey”? That’s how DeepBrain AI describes its tools for creating 2D and 3D avatars and related text-to-video solutions.
Flawless
This startup’s tagline is “Hollywood 2.0,” and it promises “magical new tools and emerging technologies” tailored for filmmaking. What does it currently offer in its generative AI toolbox?
NVIDIA & Getty Images collab: NVIDIA Picasso
NVIDIA and Getty Images announced a collaboration in late March, in which NVIDIA Picasso’s generative AI model will be trained on Getty’s image library of fully licensed images (and Getty contributors will be credited and paid accordingly). You can learn more via NVIDIA CEO Jensen Huang’s GTC keynote here.
Prisma Labs: Lensa
Lensa is described as an all-in-one imagine editing app. Its magic avatars (there’s that word again!) are a followup to a 2016 offering that went viral in 2022. Wired covers some of the features and a lot of the controversy that followed its debut.
Runway
This New York-based startup offers more than 30 of its so-called “AI Magic Tools” (are you sensing a theme in how we think about AI?) that aid creatives in generating and editing images, audio, and video content.
You may have heard about Runway during the 2022 Awards Season; some of the VFX work for “Everything Everywhere All at Once” was crafted using Runway’s editing suite (specifically the rotoscoping work for that rock scene, IYKYK).
Seyhan Lee: Cuebric
This browser-based app promises the ability to move “from concept to camera in minutes” using Stable Diffusion image generation model. Forbes’ Tom Davenport shares a good overview of Cuebric.
Synthesia
Synthesia Studio is an AI video generation platform with a lens on the corporate video space. Its home page highlights the ability to transform PowerPoint presentations into training videos with avatars, as well as use cases for sales, onboarding and more. There’s less emphasis on creativity and more focus on efficiency and budgets.