AI-Generated NFTs

The Promise

In March 2021, Beeple’s “Everydays: The First 5000 Days” sold at Christie’s for $69.3 million, legitimizing NFTs as an art form overnight. Almost immediately, entrepreneurs spotted an opportunity: what if the art that sold for millions could be generated automatically by AI?

AI-generated NFT projects proliferated throughout 2021 and 2022. Art Blocks launched “generative art” drops that sold out in seconds. Projects like Bored Ape derivatives used AI to create endless variations. Collections marketed themselves as “10,000 unique AI-generated artworks”—scarcity through algorithm.

The pitch combined two of the hottest narratives in tech: generative AI was getting good enough to create beautiful images, and NFTs were the future of digital ownership. Put them together, and you had a machine that printed money—literally.

The Rise

The numbers were absurd. AI art NFT collections raised millions in public mints. OpenSea’s marketplace filled with AI-generated collections, each promising unique artwork verified on the blockchain. Celebrity-backed projects combined AI art with influencer marketing.

The technology improved in real-time. DALL-E, Midjourney, and Stable Diffusion launched in rapid succession, each producing increasingly impressive images. NFT projects that had used crude GAN-generated art suddenly had access to photorealistic AI imagery. The floor was rising.

Collectors bought in, driven by the same FOMO that powered the broader NFT market. Some AI art NFTs traded for hundreds of ETH. Secondary market volume soared. The dream was obvious: buy into an AI art project early, watch the floor rise, sell to the next buyer at a profit.

The Fall

The collapse was a perfect storm of two bubbles bursting simultaneously.

First, the crypto winter hit. Bitcoin crashed from $69,000 to under $20,000. Ethereum followed. NFT trading volume collapsed by over 90% from its peak. The speculative fever that had driven NFT prices evaporated. Projects that had raised millions now had worthless treasuries. Secondary markets went silent.

Second, AI art became free. When Midjourney launched its public Discord, anyone could generate stunning images for $10/month—or free with trial credits. Stable Diffusion went open-source, allowing unlimited local generation. The artificial scarcity of AI-generated NFTs was exposed as absurd: why pay thousands for an AI image when you could generate your own in seconds?

The combination was lethal. AI art NFTs had no intrinsic value (unlike physical art), no scarcity (anyone could generate similar images), and no functioning market (crypto winter killed liquidity). Collections that had traded for millions became unsellable. The floor didn’t just drop—it evaporated.

Projects that had built entire businesses around AI NFTs collapsed. Marketplaces shut down. Discord servers went quiet. The artists who had pivoted to AI NFTs pivoted back—or out of art entirely. The intersection of two speculative manias produced nothing lasting.

Warning Signs

• Artificial scarcity of the infinitely reproducible: “Owning” an AI image when anyone could generate similar ones made no sense
• Speculation without utility: No one wanted the images; they wanted price appreciation
• Double exposure to volatility: Combining AI hype with crypto meant double the bubble risk
• Floor price as only value: Collections without community or utility depended entirely on speculation
• Technology working against you: Every AI improvement made existing AI NFTs less special

Epitaph

🪦 Infinite supply met zero demand