I’m going to be blunt: most of the conversations I hear about AI and art are missing the point. They’re either hysterical rants about robot artists taking over, or they’re naively optimistic takes that ignore the very real ethical minefields. I’ve spent the last two years and a significant amount of my own money going deep down this rabbit hole, and I’m here to tell you that the reality is far more nuanced and interesting.
As someone who has been in the trenches of Silicon Valley for over a decade, with a couple of successful exits under my belt and over 200 angel investments in companies like Anthropic and OpenAI, I’ve seen my fair share of hype cycles. But what’s happening with generative media is different. This isn’t just a new feature or a product; it’s a fundamental shift in how we create and consume visual information. And frankly, a lot of people are getting it wrong.
I didn’t write this to give you a list of 99 things to check off. These are the hard-won lessons that I believe will save you from the same costly mistakes I’ve seen others make. Whether you’re an artist, a founder, or an investor, this is my attempt to cut through the noise and give you a real-world framework for thinking about the ethics of AI in art.
The Myth of Originality and the Reality of Data
Let’s start with the elephant in the room: copyright. The number of times I’ve had to sit in a boardroom and listen to lawyers debate whether an AI can be an “author” is enough to make your head spin. The whole conversation is a red herring. The real issue isn’t about whether a machine can be an artist; it’s about the data we feed it.
I once invested in a promising little startup that was building a tool for generating hyper-realistic product images for e-commerce. The tech was brilliant. The problem? Their training data was scraped from all over the web, without much thought given to the source. They ended up in a legal nightmare, and the company imploded. It was a painful lesson, but a valuable one.
Here’s the truth: all art is derivative. Every artist, human or otherwise, learns by studying the work of others. The difference with AI is the scale. A model like Midjourney or Stable Diffusion has “seen” more art than any human could in a thousand lifetimes. This doesn’t make it a thief. It makes it a very, very good student. The ethical question we should be asking is not “did the AI copy this?” but “was the data used to train this model sourced ethically?”
- Lesson 1: Your model is only as ethical as your dataset. If you’re building in this space, you need to be obsessed with your data supply chain. Where did it come from? Who owns it? Have the original creators been compensated?
- Lesson 2: Stop talking about “originality.” It’s a dead end. The conversation we should be having is about attribution and compensation. How do we build systems that reward the artists whose work contributes to these massive datasets?
- Lesson 3: The legal frameworks are a decade behind the technology. Don’t wait for the courts to figure this out. If you’re a founder, you need to be proactive and build your own ethical guidelines.
The Bias in the Machine and the Ghost in the Shell
I remember seeing a demo from a company that was using AI to generate headshots for corporate websites. The results were… unsettling. All the executives were white men. The AI had learned the biases of the society it was trained on and was spitting them right back out. This isn’t a new problem, but it’s one that’s amplified to a terrifying degree with generative AI.
The thing about these models is that they are black boxes. We don’t fully understand how they work. We can’t just pop open the hood and fix the “bias” bug. It’s baked into the very fabric of the system. This is what I call the “ghost in the shell” problem. The biases of our world are the ghosts that haunt these machines.
I’ve made it a personal rule to never invest in a company that doesn’t have a clear plan for addressing this. It’s not just about ethics; it’s about building a good product. A biased product is a flawed product.
- Lesson 27: If you’re not actively fighting bias, you’re perpetuating it. It’s not enough to say you’re aware of the problem. You need to have a concrete strategy for mitigating it. This could mean curating your datasets, building in manual overrides, or creating feedback loops that allow users to flag biased content.
- Lesson 28: The “black box” is not an excuse. I’m tired of hearing founders tell me they can’t fix the bias because they don’t understand the model. If you can’t control your own creation, you have no business selling it.
- Lesson 29: Diversity in your team is your best defense. The more diverse your team is, the more likely you are to spot these biases before they become a problem. This isn’t just about race and gender; it’s about diversity of thought, experience, and background.
The Artist’s New Role: From Creator to Curator
I’ve heard a lot of artists express fear that AI will make them obsolete. I think that’s the wrong way to look at it. AI is not a replacement for artists; it’s a new tool. And like any new tool, it will change the way artists work.
I have a friend who is a concept artist for video games. She used to spend weeks creating a single character design. Now, with tools like Midjourney, she can generate hundreds of ideas in an afternoon. Her job hasn’t been eliminated; it’s been transformed. She’s no longer just a creator; she’s a curator, a director, a visionary. She’s the one with the taste, the eye, the story to tell. The AI is just a very powerful assistant.
This is the future of art. The artists who thrive in this new world will be the ones who can master these tools and use them to express their unique vision. The ones who resist will be left behind.
- Lesson 52: The most valuable skill in the age of AI is taste. With infinite content at our fingertips, the ability to curate, to select, to find the signal in the noise, will be more valuable than ever.
- Lesson 53: Stop thinking of AI as a competitor. It’s a collaborator. Learn to work with it, not against it.
- Lesson 54: The artist is not dead; the artist is evolving. The definition of what it means to be an artist is expanding. Embrace it.
Investing in the Future of Art: My Playbook
As an investor, I’m constantly being pitched new “AI art” companies. Most of them are garbage. They’re either a thin wrapper around an existing API or they’re a solution in search of a problem. But every now and then, I come across a company that gets it. A company that understands the nuances of this space and is building something truly innovative.
So, what do I look for? Here are a few of the things on my checklist:
- A defensible data strategy: As I mentioned earlier, your data is your moat. I want to see a clear, ethical, and sustainable plan for sourcing and managing data.
- A focus on workflow, not just generation: The real opportunity is not in building another image generator. It’s in building tools that integrate into the artist’s workflow and make their lives easier.
- A strong point of view: I want to invest in founders who have a clear vision for the future of art and are not just chasing the latest trend.
- A commitment to ethics: This is non-negotiable. I need to see that the company has thought deeply about the ethical implications of their technology and has a plan to address them.
I believe we are on the cusp of a new renaissance in art. The tools we are building today will empower a new generation of artists to create things we can’t even imagine. It’s a messy, complicated, and often frustrating space to be in. But I wouldn’t have it any other way. The future is being built today, and I’m thrilled to have a front-row seat.
Frequently Asked Questions
What experience informs this perspective?
This perspective comes from over a decade of building companies in Silicon Valley, two successful exits (RemoteTeam to Gusto, MovieLaLa to Gfycat), and investing in 200+ startups including Anthropic, OpenAI, and Scale AI. I write about what I've lived.
How can I apply this thinking to my own situation?
Start by identifying the core principle behind the opinion, not the specific example. Then ask yourself: does this principle apply to my context? If yes, test it in a small, low-risk way before going all in.
Do all experts agree with this view?
No, and that's fine. The best ideas in business are often contrarian. I share my perspective based on my experience and data, but I encourage you to seek out opposing viewpoints and form your own conclusions.
What's the most common pushback you get on this?
People often push back by citing exceptions or edge cases. And they're usually right that exceptions exist. But building a strategy around exceptions rather than patterns is a losing game for most founders.