Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
it is a way for the rich to downsize population so they can enjoy the earth reso…
ytc_Ugyz4bVrm…
G
I asked an AI if it would talk to their leader and spare me and it said yes, so …
ytc_UgzBIROeC…
G
Yeah this entire last couple of years really feels like a bunch of companies get…
rdc_m6zyivz
G
Looks like my 9–5 (marketing) is on its way out (I currently use AI for work and…
ytc_UgxP45lw7…
G
No joke. Im actually doing this just in case AI gains conscience and seeks veneg…
ytc_UgzSlGPzP…
G
I find it interesting that ChatGPT cannot talk about itself. It cannot see insid…
ytc_UgxHfjosJ…
G
The people in Washington don't care about their constituents. They pretend like…
ytc_Ugw08YXEk…
G
AI will take over farming too the chickens the goats the ducks every thing will …
ytr_UgwDS93pT…
Comment
If Twitter cares about retaining users outside these demographics, and they should given that their userbase has been stagnant for a while, or retaining users who care about bias, they'll address it or at least pay it lip service.
"The algorithm only shows people what they want" Sure, but consumers of content are only half the story here - Twitter depends on content *producers* just as much if not more than consumers, and producers are likely to gravitate towards platforms on which they have a chance at engagement and away from platforms that hamper engagement for their content.
There's also the fact that in this case, the output from the algorithm is *linked* to its subsequent input - the algorithm drives engagement for certain posts based on its output, and then the increased engagement with those posts causes it to drive *more* engagement towards them. It can reinforce and exaggerate the effect of existing input bias in that way (recommending skinny faces drives engagement with skinny faces which then causes more skinny faces to be recommended, and so on). It both shows *and* influences what people want.
Not to mention the effect scandal can have in influencing people's decisions about which platforms to use or not use. If Twitter gets a reputation for bias, users could leave even if it doesn't affect them directly. There's a large left-leaning base of Twitter users who care about social justice and equity that could decide to do so, for example. The Guardian clearly already thinks it's news worthy/scandalous - it could turn into a *thing* in other media outlet too. Cancelling was *literally* invented on Twitter. Blizzard is at this very moment reckoning with the revenue impact of a major scandal.
So, keep that in mind if you think "so what".
reddit
AI Harm Incident
1628611879.0
♥ -7
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | company |
| Reasoning | utilitarian |
| Policy | regulate |
| Emotion | approval |
| Coded at | 2026-04-25T08:33:43.502452 |
Raw LLM Response
[
{"id":"rdc_h8fy585","responsibility":"user","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"rdc_h8fybo2","responsibility":"unclear","reasoning":"consequentialist","policy":"unclear","emotion":"mixed"},
{"id":"rdc_h8fezag","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"outrage"},
{"id":"rdc_h8fyn2d","responsibility":"user","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"rdc_h8fayf8","responsibility":"company","reasoning":"consequentialist","policy":"regulate","emotion":"approval"}
]