From 2 Megapixels to 150: What Kodak Teaches Us About AI Skeptics
In the mid‑1990s, digital photos were terrible.
Grainy. Low‑resolution. Weird colors. Serious photographers laughed at them, and the film giants—especially Kodak—treated digital as a toy, not a threat.
We know how that story ended.
Kodak, the company that literally helped invent digital photography, underestimated it because early quality looked unimpressive compared to film. While they clung to their existing profit machine, digital kept quietly compounding in the background—going from <1 megapixel to 2 to 5 to 10 to 40 and beyond. By the time digital clearly “won” on quality, convenience, and cost, it was too late for Kodak to regroup.
Film didn’t decline gradually. It fell off a cliff.
Today, some are repeating that same mistake with AI.
Many people dismiss AI, treating it as though it’s still at the “2 megapixel” stage of its development. They point to mistakes, artifacts, hallucinations, generic writing, or awkward images and say, “See? This stuff will never replace real creatives, real marketers, real strategies.”
But the problem isn’t AI.
The problem is where we (those developing and those using AI) are on the curve.
Yes, developers continue expanding its capabilities. Meanwhile, the whole of AI is being judged based on a range of quality that’s the result of many producing AI content/images with the “toy” level of tools. Free ChatGPT is in rampant use, just as .8 to 2.0 megapixel cameras were used by the general population. But no one judged professional photographers using cameras with higher pixel values the way experts using AI today are judged because of the amateurs using inadequate AI tools.
We’re judging a compounding technology while it’s still early in its trajectory—exactly like judging 1990s digital cameras against the best 35mm film.
Kodak’s Blind Spot: Quality as a Mirage
- Film was high quality; digital was low quality.
- Film was profitable; digital was risky and cannibalizing.
- Customers loved prints; digital seemed like a niche gadget.
AI is today's "Digital Camera"
Whether we’re talking about images or content, we’re now living through the AI version of the 2-megapixel era in terms of general population use.
- AI images sometimes get hands wrong.
- AI text can be bland or generic.
- AI chatbots still hallucinate or miss nuance.
So the skeptics say, “This will never replace real creatives or real marketers.”
They’re making Kodak’s mistake.
They’re comparing today’s AI to their ideal standard and concluding that because it’s imperfect, it must be irrelevant. But AI, like digital imaging, is on an exponential curve:
- Models are rapidly improving in reasoning, style, and accuracy.
- Tools are becoming more integrated into workflows, not just “nice to have” extras.
- Interfaces are getting easier and more natural, reducing friction for non‑technical users.
If today’s free AI versions are equivalent to 7-megapixel cameras, professional versions are already sprinting through 20, 50, and heading toward “150+ megapixels” in human‑like capability. It’s not standing still while we debate it.
The real danger for businesses isn’t that AI isn’t perfect.
The danger is assuming that today’s imperfections will limit tomorrow’s potential.
Another factor to consider about the AI quality debate is that even with 150-megapixel cameras, the quality was only as good as the person pointing and shooting the camera. Once the 10 to 12-megapixel cameras were on the scene, social media was packed with freelance photographers. Many of whom positioned themselves as competent enough to be hired for projects.
I know of several people whose weddings were ruined by freelance photographers whose abilities were not as advertised. I don’t blame the camera – even if it was the wrong type of camera for the job. I blame the overzealous freelancer who didn’t understand the principles of and requirements for quality photography.
AI and the Future of Marketing: Dominate or Be Dominated
Marketing is particularly vulnerable to this shift because AI goes straight after the bottlenecks that have always constrained the field. Instead of waiting days for a campaign concept, AI can generate ideas and draft content in minutes, dramatically shrinking timelines. It can sift through massive datasets that humans could never manually process, uncovering patterns in customer behavior that used to stay hidden. And it can personalize messages at the level of the individual, tailoring content for “segments of one” at a scale that would be impossible for a human-only team.
Right now, AI may still feel like a clumsy intern—fast, eager, and sometimes impressive, but definitely in need of supervision. That’s the marketing equivalent of the 2‑megapixel phase. Yet even at this early stage, the businesses willing to lean in are already seeing tangible advantages. They’re able to ship far more variations of ads, emails, and landing pages with the same team size. They’re testing and iterating faster because content has become cheaper and easier to produce. And they’re reading customer behavior and trend signals more quickly than competitors who are still relying on manual analysis.
As AI quality improves, a clear divide will open up. On one side will be the companies that intentionally build their marketing around AI. They will use it to power hyper‑personalized campaigns, constantly refine every piece of content based on data, and free their human teams to focus on strategy, storytelling, and relationships instead of repetitive production work. On the other side will be the companies that cling to old methods. They’ll move slowly because every asset is created from scratch by hand. They’ll spend more money but achieve less reach and relevance. And they’ll watch as competitors with the same headcount and budget produce ten times the output.
In other words, AI in marketing is following the same pattern we saw with cars overtaking horses and digital overtaking film. You can either learn to master it and build on top of it, or you can find yourself competing against those who did—and in that world, the gap will only widen.
Practical Ways to Lean In
(Without Losing Your Soul)
- Use AI as a first draft, not the final say Let AI generate starting points—headlines, angles, email drafts, image variations—then add human insight and brand voice.
- Blend human strategy with machine execution Humans set the narrative, positioning, and ethics. AI scales the output, testing, and personalization.
- Systematize learning Every AI‑assisted campaign is a chance to collect data: what messages work, what segments respond, what designs convert. Feed that learning back into both your human and AI processes.
- Build AI literacy in your team The goal is not for one “AI person” to own everything. The real advantage comes when strategists, writers, designers, and analysts all know how to wield these tools.
The Call to Action: Don’t Be Kodak
Kodak didn’t fail because digital was bad. Kodak failed because digital was bad until it wasn’t—and its leaders assumed that first impression was permanent. They treated an early, imperfect version of a technology as though it were the finished product, and by the time they realized their mistake, the market had moved on without them.
Many of today’s AI objectors are falling into that same trap. They’re looking at the rough edges of this moment and making long‑term strategic decisions based on short‑term limitations. But if you lead a business, work in marketing, or create content, your real decision isn’t about waiting for AI to become perfect. It’s about whether you’re willing to start learning, experimenting, and building now, so you’re ready when AI crosses that invisible threshold from “interesting gadget” to “default infrastructure.”
When that happens—and it will, just as surely as cars replaced horses and digital photography replaced film—it won’t be a debate or a trend to watch. It will simply be the environment you operate in, the baseline expectation of how modern businesses communicate, create, and compete.
So the question becomes less about fear and more about opportunity: do you want to be the company that learns to drive early, testing the roads, gaining confidence, and discovering new routes before anyone else? Or the one still proudly polishing the horse and buggy, not because it’s the best way forward, but because it’s familiar, while the rest of the world is already miles down the road?