We're in 1905: Why Electricity (Not Dot-Com) Is the Right AI Analogy
The article argues that comparing today's AI boom to the dot-com bubble is misleading, suggesting instead that AI's impact is more akin to the historical adoption of electricity in factories. Early adoption of electric motors did not yield productivity gains until factory layouts and workflows were fundamentally redesigned. Similarly, true AI transformation will require rethinking organizational structures and processes, not just applying AI to existing systems.
- ▪Electricity was introduced in factories in the 1880s, but productivity gains did not materialize until the 1920s when factory designs were restructured.
- ▪Factory redesigns enabled by electricity included single-story layouts and unit drive systems, which optimized the use of electric power.
- ▪The data industry has repeatedly adopted new technologies without changing underlying architectures, such as using AI to accelerate legacy code migrations without transforming systems.
- ▪Many companies today are merely bolting AI tools like LLMs onto existing workflows without altering organizational design or decision-making processes.
- ▪Conway’s Law suggests that organizational structure deeply influences system design, meaning real AI transformation requires structural change, not just technological upgrades.
Opening excerpt (first ~120 words) tap to expand
We're in 1905: Why Electricity (Not Dot-Com) Is the Right AI AnalogyThe Weekend Windup #29 - Cool reads, events, links, and moreJoe ReisApr 19, 202657713ShareA 1920’s era speculation on electrifying the farm. Source: Paleofuture.Everyone loves comparing AI to the dot-com bubble. It’s the most recent tech revolution we can point to, and the parallels are obvious: insane valuations, companies raising money by slapping “AI” on their pitch deck, the general pandemonium and silliness of it all. I was around during the dot-com era, and it was f*cking bananas. Money made from thin air…and also vanishing into thin air. Much like today.But I think it’s the wrong analogy for where we are with AI today.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at Substack.
Discussion
0 commentsMore from Substack
