The word 'innovation' requires no evidence to work. Merriam-Webster defines it as simply a new idea, method, or device, but the moment organizations attach that label to a product or policy, it inherits the credibility of proven progress. This linguistic sleight of hand is the article's central argument, and it is a precise one. Innovation describes novelty. Progress describes improvement. The two are not the same, and conflating them is not an accident.
The examples here do real work. The Reliant Robin, a three-wheeled 1970s British microcar, was marketed as an innovative response to fuel costs and tipped over on sharp turns. Antibiotics saved millions and generated antibiotic-resistant bacteria. Smartphones transformed commerce and decimated attention spans. Each case shows the same mechanic: the innovation label arrived first, the consequences arrived later, and the people who deployed the word were rarely the ones absorbing the costs. The article argues this is structural, not incidental.
Generative AI is the live case study. The article names specific industry phrases, 'minimum viable product,' 'it's a feature not a bug,' 'iteration is part of the process,' as the vocabulary that insulates developers from liability when models hallucinate facts, infringe on intellectual property, and reach millions of users before any regulatory framework exists. Read the full piece for the breakdown of how innovation language functions as a legal and ethical buffer, not just a marketing strategy.
[READ ORIGINAL →]