Meredith Martin
Researcher with public evidence across AI research.
- AI signals
- 13 past 30d
- Sources
- 12 distinct domains
- Discussions
- 29 past 30d
- Latest signal
- 5d ago
Articles & links
finally had a moment to read this. "why should I waste my time and my sanity debating kooks? In short, because the kooks are now in power, and they are using that power to dismantle the institutions designed to keep the rest of us safe, healthy, and free." www.techpolicy.press…
arxiv.org/abs/2510.08831 updated version of our article from last year @wouterhaverals.bsky.social and I updated the title & experiments. Check out this strengthened version, esp. if you are thinking abt cultural evaluation & benchmarks. @arnicas.bsky.social @dmimno.bsky.socia…
- Thirteen AI evaluators graded passages labeled 'AI-generated' 34.3 percentage points lower than identical passages labeled human, versus 13.7pp for 556 human readers.
- The bias held across a 14x14 matrix of AI creators and evaluators at +25.8pp, indicating it operates across model architectures.
- Attribution labels caused evaluators to 'invert assessment criteria,' rating identical features positively or negatively based solely on perceived authorship.
Recent commentary
Authors who are filing claims in the Anthropic lawsuit -- are you all going 50/50? I can't find the contract for my 2012 book. I know I can ask the press but I'm also just curious how folks are doing this.
In Meredith Martin's orbit
Center = Meredith Martin. Left = members they follow (green edges). Right = members who follow them (blue edges). Top = mutual follows (orange edges, slightly larger). Drag any node to reposition; click to open that profile.
Are you Meredith Martin? Show it.
Add the Who’s Who of AI badge to your site or bio. It links back to this profile.
Markdown: [](https://aiweekly.co/whos-who/person/mmvty-bsky-social)