Could AI Have Predicted Amazon in 1994?
An Experiment in Curiosity
What if you had the tools to spot the next big thing before anyone else? Could AI have predicted Amazon back in 1994?
This one’s for my entrepreneur friends—the risk-takers, dreamers, and visionaries bold enough to tackle the impossible. I posed this very question during a college AI class I taught last week and ran a live experiment to find out. Using GPT-4o and other engines, we recreated the world of 1994, analyzed the zeitgeist, and asked AI to uncover groundbreaking business opportunities.
Spoiler: GPT-4o suggested a business model that bore an uncanny resemblance to Amazon—with an intriguing twist. Meanwhile, other engines offered creative alternatives, each targeted at solving real consumer pain points and leveraging early innovations. The key takeaway? AI provides options for the curious, not answers for the lazy.
Could you build the next Amazon? Someone’s going to. Why not you?
Here’s how the experiment played out—and what we discovered.
Step 1: Mapping the Zeitgeist of 1994
To start, I asked GPT o1 pro’s “Deep Research” to build a snapshot of the U.S. cultural, political, and economic landscape in 1994. Here’s the exact prompt I used:
"I am interested in understanding the zeitgeist in the U.S. in 1994, including politics and major events, pop culture (movies, music, TV, fashion, celebrities), social movements and public sentiment, economic climate and business trends, technology and innovation, and sports and entertainment."
The response was a fun time capsule:
Politics: The Clinton administration was focused on economic growth and healthcare reform.
Pop Culture: Grunge music and CDs dominated, Forrest Gump and The Lion King ruled the box office, and cable TV was on the rise.
Technology: The World Wide Web was emerging but largely confined to universities and tech enthusiasts.
Public Sentiment: There was growing optimism about technology, tempered by skepticism about whether the internet could ever go mainstream.
Business Trends: Physical retail—especially chain bookstores and music stores—was thriving but increasingly inefficient.
After reviewing the output, I removed references to Amazon and e-commerce (to not ‘tip it off’), saving the cleaned-up version as 1994 Zeitgeist.docx (included for your reference).
Step 2: Analyzing the Books and Music Market in 1994
Next, I explored the industries Amazon would eventually disrupt: books and music. Using GPT-4o, I asked:
"I’m interested in understanding the market for books and music in 1994—retail stores, buyers of these products, the current economy, media and formats, consumer preferences, etc. Imagine I’m looking for innovations that would be relevant for these two industries in any way. Also, identify the market headwinds, tailwinds, and overall conditions."
GPT-4o’s analysis revealed several key pain points:
Physical inventory limited variety, preventing consumers from accessing niche products.
Chain stores dominated but relied on outdated promotional strategies, like bestseller lists and in-person browsing.
High operational costs restricted experimentation with new formats.
Tailwinds included rising consumer interest in convenience and personalization driven by early digital technology.
As before, I saved the edited document as 1994 Book and Music Market.docx.
Step 3: Putting AI in 1994’s Shoes
To bring everything together, I uploaded both documents into GPT-4o with the following prompt:
"I’d like you to imagine that we’re in the U.S. in 1994. Read the attached '1994 Zeitgeist.docx' for context about what was going on, how people were feeling, and what they were concerned about in the U.S. at that time. I’d also like you to understand how people were thinking about books and music in 1994 by reading the attached '1994 Book and Music Market.docx'."
This allowed GPT-4o to analyze conditions as if fully embedded in 1994.
Step 4: Identifying Big Problems to Solve
Next, I asked GPT-4o to surface major problems in the books and music industry:
"Pretend it’s still 1994. I’m looking for a big problem that I might be able to solve as a business in the book and music industry that connects to the zeitgeist."
The AI identified pain points such as:
Limited access to niche or hard-to-find products.
Consumer frustrations with browsing inefficiencies in physical stores.
Curiosity about technology, but limited practical digital tools for books and music.
Step 5: Exploring Business Opportunities with AI Responses
Finally, I asked GPT-4o to suggest potential business opportunities:
"Given these conditions in 1994, what business opportunities or innovations could emerge at the intersection of books, music, and technology?"
The full list of business ideas generated is provided, and here are some standouts:
The Book & Music Interactive Kiosk
Opportunity: Install kiosks in stores where customers can preview albums, read book excerpts, and scan barcodes for reviews and recommendations.
Why It Works in 1994: This bridges physical retail with emerging digital curiosity.
“The Culture Pack” Subscription Box
Opportunity: Pair a monthly book and music CD around curated themes (e.g., sci-fi novels with electronic music).
Why It Works in 1994: It simplifies consumer choices and builds excitement around curated experiences.
The Early Digital Bookstore
Opportunity: A CD-ROM-based virtual bookstore where consumers can:
Read summaries, sample chapters, and listen to audiobook excerpts.
Watch author interviews and video trailers.
Order physical copies via phone or mail-in forms (pre-e-commerce).
Business Model:
Sell CD-ROMs via bookstores, electronics retailers, and mail-order catalogs.
Charge publishers a listing fee for featuring books.
Include digital extras like Q&As and exclusive content.
Why It Works in 1994: It solves the problem of book discovery in a pre-internet world, offering a richer experience than a paper catalog. It also lays the groundwork for digital e-commerce—just before Amazon’s rise.
This idea closely mirrored Amazon’s eventual model, while offering an early bridge to digital retail via CD-ROMs.
The Big Takeaway: Innovation for the Curious, Not the Lazy
The key insight from this experiment? AI provides starting points for the curious—it doesn’t hand you fully formed solutions. Its strength is helping visionaries uncover hidden opportunities and imagine what’s possible.
When paired with curiosity and ambition, AI becomes the ultimate partner for entrepreneurs seeking transformative ideas.
Why Not You?
Someone will build the next Amazon. Why not you?
For those bold enough to ask the right questions, the future is wide open. AI offers the tools—now it’s time to bring the curiosity and vision. I can’t wait to see what my students—and dreamers like you—build next.
Footnote on Other AI Models
Curious about what other AI models predicted? I’ve included their actual responses as links below.
TIE FOR FIRST PLACE WITH GPT 4o: Grok 3’s insights: Touts the “Cyber Shelf,” a bare-bones web-based store with direct shipping to customers. Paints a picture of text-heavy pages, forums for reviews, and door-to-door delivery—a strong mirror of early Amazon.
THIRD PLACE: ChatGPT o1’s insights: Focuses on an online discovery and mail-order system, showcasing both books and music. More of a broad aggregator that connects indie retailers and consumers, with a strong emphasis on “secure transactions” and easy shopping.
FOURTH PLACE: ChatGPT 4.5’s insights: Proposes a dial-up-friendly “sampling” site where folks try short clips or chapters, then purchase online. Envisions bridging physical pickup options with a curated web experience, aiming to ease buyer uncertainty.
FIFTH PLACE: Claude 3.7’s insights: Focuses on personalized media recommendations (like early Amazon’s “people who bought this also bought…”). Mixes it with interactive kiosks and subscription clubs, leaning more on human curation than purely on e-commerce.
Footnote on Ella Ai
Ella, powered by the latest AI models (currently the ‘winning’ GPT-4o in this example), combines the storytelling and clarity needed to refine visionary ideas. Tools like Ella make all the difference in moving an idea into reality.
