How has an algorithmic arms race created an environment where controversial and incendiary content is routinely amplified on our feeds?
On this episode, Marianna discusses her latest investigation, which provides a paper trail showing how social media giants ignored internal warnings about the design of their platforms.
She’s heard from whistleblowers about problematic moderation practices, how companies deal with so-called borderline content, and the troubling reality that algorithms may now be beyond our control.
Also this week, we look at the website ‘your ai slop bores me’ which launched two weeks ago and is already claiming over a million unique visitors. It may look like a chatbot, but this is actually humans cosplaying as generative AI and completing mundane tasks for one another.
Matt explains how this relates to what’s becoming known as the ‘meat layer’, where humans are used to prop up the shortcomings of artificial intelligence.
Top Comment is hosted by Marianna Spring and Matt Shea.
Chapters:
00:00 – Inside the Rage Machine: Algorithm Whistleblowers
00:30 – What is the ‘algorithm arms race’
02:45 – Meta’s research into the harm of reels
05:00 – TikTok’s prioritisation controversy
09:00 – Engineers admit algorithms are now a black box
12:30 – CrowdTangle and the collapse of transparency?
13:45 – TikTok and Meta respond
14:20 – Musk’s X Algorithm
15:30 – Your AI slop bores me: Humans pretending to be AI
17:20 – “People are sick of AI slop”
18:40 – The meat layer of AI: Hidden human labour
19:30 – Human fleet operators for driverless cars
20:35 – What is Moltbook? AI‑agent social network
23:00 – Why AI companies benefit from AI fear narratives