Artificial Intelligence

OpenAI’s social video experiment has drawn warnings from Hollywood and opt-outs from Disney, but its climb up the App Store suggests the debate around the future of generative social video is here to stay.

The fact that a new generative video platform from a major AI lab is an IP free-for-all should surprise nobody. The fact that it seems to be overcoming cultural inertia, and that it brings day-one tools for users to model themselves and their friends, asks some more unexpected questions.

According to the World Trade Organisation, AI could massively scale-up global trade in goods and services within fifteen years. But these results would come from discerning the difference between the volume of ‘work’ generated and the actual utility of it - a skillset that lies largely outside consumption markets.

Although they’re concentrated in the coding domain for now, the rise of the “AI cleanup specialist” is a reminder that generative AI can bring ideas to life, but actually implementing them demands expertise.

A new proposed standard for content licensing for AI training and inference, put forward by the publishing sector, could offer a template for beauty and fashion to follow.

The use of AI in campaign and editorial imagery keeps landing in that space between workflow efficiency and cultural resistance, and the reaction is anything but straightforward.

Google's new AI image generation model delivers consistency in a way that’s so far been reserved for models that had undergone industry and brand-specific fine-tuning and reinforcement, but wider availability is likely to pour fuel on fashion’s deeper ethical and process questions when it comes to AI images.