Today, OpenAI has released authorized "plugins" with selected partners, which are meant to augment its language model with "eyes and ears" that connect with recent or business-specific content. This reminds me of the way, over the last decades, a few digital companies have managed to set standards upon which other businesses had to develop or share their own assets, and where the standard owner—Google, Apple, and others—monetizes its intermediation. It happened so with operating systems, first on computers then on mobile (app stores) and other nomad endpoints (CarPlay, Android Auto). It also happened on web search engines—ruling whether and how much a content matters or not—or in superapps.
The rise of conversational interfaces trained over such a massive amount of data is likely to change what digital consumer experience is, whether on end-user side (visitors expecting more conversational interfaces and reluctant to search content at the root) or on publisher side (under pressure to make sure their content or products remain visible in these new challenges). It also raises ethical concerns about how a user will find his way between trustworthy content and biased one.
