I'm an AI artist working in Hollywood. Here's my advice for others looking to use the tech to boost their careers.
Minta Carlson, 35, who goes by Araminta K professionally, represents an emerging class of labor in Hollywood. She's a senior creative AI architect at Moonvalley, one of the prominent AI companies working in Hollywood. Moonvalley, which just raised $84 million, pitches an ethical AI model it calls Marey and owns AI film studio Asteria Film Co., which was cofounded by filmmaker and actor Natasha Lyonne and filmmaker Bryn Mooser.
Carlson shared her career path in this as-told-to essay and how other creative people in Hollywood can adapt to AI. It's been edited for length and clarity.
After studying theater and working in graphic design, I taught myself how to train AI models.
In the Hollywood context, what we do is called "fine-tuning" models. Let's say you wanted a specific dragon in a production. You could train a model to understand who that dragon is, how it moves, what it looks like — what it looks like really close up and really far away. And when you point the dragon this way and you tell it to fly upward toward the sunset, it'll understand what that means.
I started working at Asteria full time early this year. Sometimes, I'm working on animated shorts in-house, and other times, I'm working with studios on VFX or background pieces. I'll work with creative directors and others to flesh out how a project's characters and styles should look. We recently helped a studio augment a party scene that otherwise would have been cut for budgetary reasons.
I have not seen a lot of people like me before starting this job. Anyone with a technical and art background can do it. It takes less than an hour to explain how to curate a dataset in the context that we need. What's really difficult to teach someone is how to have taste and how to have a really critical eye. With people who don't have an art background, it can be weeks of going back and forth.
I could definitely see technical artists, traditional animators, and colorists doing this work, as well as people with less traditionally technical roles, like a creative director or a concept artist — anyone who has to communicate visual content to other people.
The most misunderstood idea is that AI is this amorphous blob that can replace an artist. At the end of the day, AI doesn't have a perspective, vision, or opinion about anything. If you're just prompting a model, whatever results you're getting are just that model's bias unless you know what you're asking for and how to ask it.
I still lean on my traditional skills
I was born in Berkeley and moved to New England when I was young. I went to school for theater directing and playwriting. After college, I worked as an independent graphic designer and honed art skills on my own. I also built websites.
I started working with Stable Diffusion in 2022 when AI was starting to blossom. I wasn't happy with the results, so I started to learn how to train the model. It was very technical, so I just muscled through. It's a very artistic and curatorial process.
In the case of Marey, you're usually putting in 20 or 30 images or videos to fine-tune the model, so each one has so much impact, and you need to understand each one's strengths and weaknesses and how each one relates to the other. Knowing how to look technically at art and think about a whole piece is something I had to do in school for theater, and those skills came back in full force.
For people already working in Hollywood, it's really easy to get overwhelmed by all of the different tools. I tell people, don't test every single tool. Come up with something you're trying to solve, like making someone walk across the screen or drawing multiple angles of an animated character. I like sites like Replicate or Fal that let you test a lot of open-source models.
Knowing how to illustrate and animate are still valuable, but I recommend people prioritize creativity and not just technical skills. Otherwise, you might never think of all of those different movements that you need to show the AI. Rather than drawing a million of the same character to try to perfect it, I would draw a million different characters to create a perspective and unique style.
Artists have a future in Hollywood
When I was consulting, I had conversations with people in marketing who wanted to try to replace artists with AI, and it didn't really work. It was very obvious that the person asking didn't understand how AI and art worked, and they didn't have self-awareness around their own standards. Even when I saw colleagues try to follow through with those requests, it was never good enough.
Now, I'm rarely involved with a project if we're not taking direction from an artist or a director.
I think the fear of AI impacting jobs in Hollywood is understandable, but I'm seeing that AI is going to unlock a lot of productions that may be struggling to get funding or forward momentum.
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Fast Company
3 minutes ago
- Fast Company
Investing in early-in-career talent is vital to win the AI race
advertisement AI is fundamentally changing how we work. People will increasingly oversee more AI agents, changing the way we think about teams. Business leaders must shape what's next—not shrink from it. From job elimination to job evolution EIC employees are AI natives who are already leading the transformation. They intuitively engage with tech, bring creative agility, and have the curiosity needed to thrive in fast-changing environments. According to the World Economic Forum, job loss between 2025 and 2030 will be more than offset by new roles, leading to a net gain of 78 million jobs. As some roles and tasks phase out, new ones emerge that require skills like AI and data fluency, creative thinking, resilience, and curiosity. Subscribe to the Daily newsletter. Fast Company's trending stories delivered to you every day Privacy Policy | Fast Company Newsletters If we don't protect and modernize the EIC pipeline, we risk widening the skill gaps and stalling the impact and ROI of AI solutions. EIC talent will be tomorrow's leaders, so we need to build pathways for them today. The demographic and leadership imperatives The talent pipeline is narrowing just as the pace of transformation is accelerating. U.S. birth rates are declining. Fewer 18-year-olds are entering the workforce. Higher education costs are skyrocketing, and many high school graduates are choosing two-year and technical degrees or trade jobs. That makes every EIC hire even more valuable. HR leaders help define the structure of the workforce and manage payroll—the largest line on the profit and loss statement—so where we invest matters. EIC roles are often the smartest entry point for workforce planning. We need to build AI-first cultures rooted in continuous learning, with roles that fuel business and personal growth. That means doubling down on equipping early-career talent with the skills, creativity, and adaptability to lead AI-powered organizations. And our succession pipelines must prioritize leadership capabilities like AI fluency, orchestration, and human-centered change management. That means focusing on these key steps: Reimagine strategic workforce planning As leaders, we must identify the skills AI won't replace and the skills that matter most to our businesses—from programming and UX design to collaboration, creative problem solving, and empathy. Then we should map those skills to evolving roles. For example, if AI handles research, an entry-level role could evolve into a prompt engineer or curator. Other future roles could include AI safety and ethics coordinators and AI agent trainers for front line workers. Design new rotations and exposure Companies that invest in internships build future-ready talent pipelines. Internships today are table stakes. To stand out, we need to build rotational programs, apprenticeships, and real-world experiences that give EIC hires exposure across the business. Reverse mentoring, for example, could give EIC talent a chance to connect directly with senior leaders, while giving those leaders a window into AI-native thinking. The goal is to retain top talent by creating a culture of growth, mobility, and connection. With clear goals, meaningful work, strong managers, and real learning experiences, EIC talent has the chance to thrive and drive innovation. At ServiceNow, 95.6% of our interns accepted our full-time offers in 2024, proof of meaningful investment. Embrace AI-first learning for growth and retention Retaining top talent, especially early-in-career talent, starts with listening followed by meaningful action. Sixty-five percent of EIC workers say they'd stay at least four years at a company if it offered robust development opportunities. We need to show EIC talent how they can grow, and design learning that matches their curiosity. EIC employees expect learning to be personalized, bite-sized, and built into the workflow. That's why we launched ServiceNow University—to train our employees and the broader technology ecosystem. It's working: EIC hires at ServiceNow have a 7% lower attrition rate in their first two years than their peers. The long game: Invest in young talent and AI Leaders don't need to decide between cutting costs and investing in the future. They can do both when they focus on transforming the workforce. Organizations that lead with intention—those that rethink roles, invest in AI enablement, and reimagine EIC talent—will attract the best minds and shape the next era of innovation. We all have a lot to learn in this new world, and we should evolve our strategies as we go. But EIC employees are essential. Their fluency with technology, drive to learn, and creative edge are exactly what we need to build the future. We can't afford to sideline them. Committing to EIC talent will require a lot of hard work and vision, but with the right strategy, it is possible. Jacqui Canney is chief people and AI enablement officer at ServiceNow.
Fast Company
34 minutes ago
- Fast Company
The future of manufacturing depends on empowering workers
There is a very real anxiety around automation, especially in America—where once-thriving communities have been gutted by waves of displacement, offshoring, and the unchecked application of technology. The narrative around manufacturing automation has long been framed as a zero-sum game: Either humans do the work, or machines do. But that binary thinking is outdated and dangerous. Not only does it threaten millions of livelihoods, but it jeopardizes innovation and efficiency. It overlooks a more powerful and sustainable vision for manufacturing's future, one where AI empowers people instead of replacing them. AI is key to a more resilient workforce and more competitive industries. And when people leverage manufacturing intelligence platforms, they can outperform full automation. Manufacturing at the crossroads America's manufacturing sector is at a crossroad. On one hand, we face a growing shortage of skilled labor amid major sector growth. Companies are announcing new factories in the United States every month, but the National Association of Manufacturers projects that by 2030, more than two million manufacturing jobs could go unfilled due to a lack of qualified workers. On the other hand, we're in the middle of an extraordinary wave of technological advancement—AI, augmented reality, robotics, and more. By using AI to make expertise more accessible, repeatable, and scalable, we can build a more efficient and safer workforce—and actually fill our empty jobs. Today's factory jobs are more art than science, where years of built-up knowledge are critical to getting things done the right way. They rely on software, sensors, and split-second subjective decisions. But the tools on the factory floor haven't evolved past a checklist and a hope. We're asking workers to operate with 20th-century instructions in a 21st-century environment. Manufacturing intelligence platforms aid operators Empowerment in a 21st-century environment starts with manufacturing intelligence platforms. It involves equipping operators with the context and support they need to thrive on the factory floor. A new hire should be able to perform a task with the confidence of a 10-year veteran. Seasoned workers should be free to focus on high-value tasks where their expertise has the greatest impact. And those on the frontline of work should have the tools and agency to shape the future of AI—imagining potent use cases no boardroom could conceive. This vision is not just better for operators. It's better for business. The most forward-thinking companies aren't chasing full automation. They're investing in tools that help their people work smarter, faster, and with more precision. They recognize that machines can't replicate everything humans do. The real opportunity lies in upskilling and empowering people, not replacing them. And this vision is better for national competitiveness. As stress on global supply chains intensifies, the United States faces a defining challenge: to rebuild and reimagine industrial strength. The future of American manufacturing depends on revolutionary efficiency and innovation for the people actually doing the work. Human adaptability is a competitive advantage, not a liability, and human intelligence enhances artificial intelligence just as much as the reverse. To deliver that efficiency where it matters most and truly expand domestic capacity, we need to harness the power of AI to accelerate human potential. Make progress There's a broader economic and cultural truth here, too. Manufacturing isn't just about making things. It's about making progress. And progress requires us to unlock the full capabilities of our people. America's strength in the 20th century didn't come from labor automation. It came from the transformation of it. From the GI Bill to labor protections to the rise of technical education, we invested in our people. And now, as we enter a new industrial era, we have a chance to do it again. This time with smarter tools, more connected systems, and a deeper understanding of what humans are best at: problem solving, adaptability, and creativity.
Forbes
36 minutes ago
- Forbes
The AI Bubble Paradox: Why OpenAI's $500 Billion Valuation Proves The Opposite
It's an odd spectacle: Sam Altman warning about an AI bubble while his company, OpenAI, rockets toward a $500 billion valuation. If there's more obvious proof that we're not in traditional bubble territory, it's hard to imagine what it might be. Speaking to The Verge, Altman acknowledged what many observers already suspected. "If you look at most of the bubbles in history, like the tech bubble, there was a real thing. Tech was really important. The internet was a really big deal. People got overexcited. Are we in a phase where investors as a whole are overexcited about AI? My opinion is yes," he said. But here's where Altman's bubble warning breaks down into paradox. In the same conversation where he cautioned about investor overexcitement, he casually mentioned expecting OpenAI to spend "trillions of dollars on its data center buildout in the not very distant future." His attitude toward the inevitable criticism? "You should expect a bunch of economists wringing their hands, saying, 'This is so crazy, it's so reckless,' and we'll just be like, 'You know what? Let us do our thing.'" This isn't the language of someone truly concerned about unsustainable speculation. It's the confidence of someone who sees the long-term trajectory clearly, even if the short-term valuations seem frothy. The Infrastructure Reality Check The numbers backing Altman's confidence are staggering. OpenAI just secured $8.3 billion in new funding at a $300 billion valuation round that was five times oversubscribed. Employee share sales could push the company's valuation to $500 billion, making it one of the most valuable private companies in history. These aren't the desperate funding rounds of bubble companies scrambling for survival; they're the measured capital raises of a company planning for massive scale. OpenAI isn't alone in this infrastructure binge. Microsoft plans to spend $80 billion on AI data centers this fiscal year alone. Meta is projecting up to $72 billion in AI and infrastructure investments. These aren't speculative bets — they're calculated investments in what companies see as the next fundamental computing platform. When Altman tweets that OpenAI will bring "well over 1 million GPUs online by the end of this year," he's not describing bubble behavior. He's describing the build-out of critical infrastructure for a technology that's already demonstrating real utility and rapid adoption. The Usage Explosion Adoption metrics tell a story that bubble skeptics often miss. Altman recently revealed that reasoning model usage among OpenAI's customers is exploding. Free users went from less than 1% to 7% daily usage, while Plus users jumped from 7% to 24%. This isn't the shallow adoption curve typical of bubble technologies; it's the steep trajectory of tools becoming indispensable. Our research at Futurum Group suggests AI inference workloads could account for more than 80% of computing by 2030, when factoring in on-device processing, autonomous driving, edge computing, and agentic AI systems. If accurate, we're not looking at a bubble but at the early stages of the most significant computing platform shift since mobile. Tuesday's tech selloff, triggered partly by an MIT study claiming 95% of companies see no returns from generative AI, exemplifies the market's nervous energy around AI investments. NVIDIA Corp. dropped 3.5% and Palantir nearly 10%, suggesting investors are increasingly sensitive to any hint that AI returns aren't materializing fast enough. But as I have been saying since the arrival of ChatGPT, this build out will last decades as AI will be the arbiter of the world's leading economies and all the others. We are just getting started. The expectation that enterprise AI deployments should show immediate, measurable returns reflects a fundamental misunderstanding of how transformative technologies get adopted. Companies don't typically see ROI from major technological shifts in quarters. They see it in years. The Bubble Brigade Crowd High-profile financial figures Ray Dalio and Joe Tsai have warned about AI investment pacing ahead of sustainable growth, drawing parallels to the dotcom crash. Apollo Global's Torsten Slok argues the current situation could eclipse the 1990s internet bubble, noting that the 10 largest S&P 500 companies are more overvalued relative to fundamentals than they were at the dotcom peak. These warnings deserve attention, but they conflate different types of risk. Yes, some AI startups are overvalued, and some applications overhyped. But the core infrastructure investments driving the boom are based on demonstrated utility and clear demand trajectories. Meta's approach illustrates why the smartest companies aren't treating this as bubble territory. It is optimizing for the future while continuing to be one of the best plays for real world AI consumption. The company is simultaneously laying groundwork for AGI while embedding AI into everyday products that millions use daily. This dual approach of building for the long term while monetizing current capabilities reflects confidence in AI's staying power. On Thursday, the company confirmed a Wall Street Journal report that it has paused hiring for its new AI division amid the restructuring. But it doesn't signal Meta intends a major cutback in AI spending – it is simply in digestion mode after a massive spending spree that included several acquisition-sized offers and hires in the nine-figure range. Before pouring more investment into its AI teams, Meta needs time to place and access its new talent before forging to new breakthroughs like superintelligence. Altman's apparent contradiction — warning about bubbles while planning multitrillion-dollar infrastructure investments— makes perfect sense. He's distinguishing between short-term market froth and long-term technological transformation. Investor overexcitement in the near term doesn't negate the fundamental importance of the underlying technology. The AI sector may be experiencing speculative excess in valuations and expectations. And certainly not all names that have benefitted from AI are of the same quality. But when the CEO is most positioned to benefit from that speculation, it suggests we're witnessing something more substantial than a bubble. We're watching the expensive, messy process of building the next computing platform, one that promises to be worth every overexcited dollar invested in its foundation. In bubbles, the smart money gets out early. In genuine technological revolutions, it doubles down on infrastructure. OpenAI's $500 billion trajectory signals which category we're really in.
