Confusing conversations by the synthetic uh
Weekly thoughts on human-AI co-performances
These thoughts here are often triggered while listening to a podcast, and they are tech-related often—also this week. However, I made a note while driving (and pausing listening), and I planned to make sense of my thinking and dictate later. As a side note, this ritual is becoming increasingly common: capturing spoken thoughts with voice recognition, which can be both in Dutch or English, and when at the desk again, asking ChatGPT or Claude through Lex to make a readable text. Creating paragraphs, finetuning the grammar, and more. But explicitly not by adding new information from the unreliable generative brain.
But anyhow, away from this side note: in this case I was not triggered by the topic of the podcast (still interesting to listen, the Every podcast with Nashilu Mouen-Makoua from The Browser Company, discussing new stuff explected in de Arc browser like AI). I was triggered by a new phenomenon (at least for me) where the way she talked, using some uhs and uaha’s, was almost as if I was using an AI voice. Not that she is like an AI voice, not at all, but the way that these new AI voices - with the OpenAI iteration - try to mimic real people by adding these non-functional words.
So, is that what we are doing now? Are we creating a reality that breaks down our instincts regarding what is real and what is not? And will we adapt our behavior to it? Even before it is necessary, that is often described as the Chilling Effect; you change your behavior to sync with the perceived behavior. It is also my firm belief that we will see an acceleration in “co-performance” with machines as soon as we adapt our default behavior to the characteristics of the intelligence we gonna live with. We expect the AI to adjust to us, but it works both ways.
This weekly “Triggered Thought” is written as part of the Target is New newsletter, which offers an overview of captured news from the week, paper for the week, and interesting events. Find the full newsletter here.
About the author; Iskander is particularly interested in digital-physical interactions and a focus on human-tech intelligence co-performance. He chairs the Cities of Things foundation and is one of the organizers of ThingsCon. Target is New is his “practice for making sense of unpredictable futures in human-AI partnerships”.