This seems like a straightforward thing to do to ensure trust in reporting
2.029466
Relative Brier Score
0
Forecasts
0
Upvotes
Forecasting Activity
Forecasting Calendar
No forecasts in the past 3 months
| Past Week | Past Month | Past Year | This Season | All Time | |
|---|---|---|---|---|---|
| Forecasts | 0 | 0 | 0 | 0 | 6 |
| Comments | 0 | 0 | 0 | 0 | 6 |
| Questions Forecasted | 0 | 0 | 0 | 0 | 6 |
| Upvotes on Comments By This User | 0 | 0 | 0 | 0 | 0 |
| Definitions | |||||
New Prediction
Probability
Answer
81%
Yes
19%
No
Why do you think you're right?
Files
Why might you be wrong?
I don't know fully what's involved to get compliance.
Files
New Prediction
Probability
Answer
73%
Yes
27%
No
Why do you think you're right?
The bar is low for "use" as defined here. I don't think a fully automated solution will be deployed but LLMs are being used ad hoc all the time now, so even manually creating bots and using an LLM on the side to generate bio info would count, and I can easily imagine IRA workers doing this.
Files
Why might you be wrong?
Even though I am fairly confident LLMs will be involved in most aspects of generative text, their use in the modest ways I suggest may not be detectable by Meta.
Files
New Prediction
Probability
Answer
18%
Yes
82%
No
Why do you think you're right?
I think this is coming but I think rollout on a massive level in 2023 is too soon
Files
Why might you be wrong?
I don't really know what's involved in calculating C2PA; maybe it's easy to do and then it could be quickly rolled out
Files
New Prediction
Probability
Answer
3%
Less than 4 million
59%
More than or equal to 4 million but less than 6 million
38%
More than or equal to 6 million
Why do you think you're right?
People don't love signing up for things like this but 4 million is not too high a barrier
Files
Why might you be wrong?
Even six million isn't too high and if some government requires WorldID then there would of course be a sudden spike
Files
New Badge
New Prediction
Probability
Answer
81%
Yes
19%
No
Why do you think you're right?
AI detection is a hot issue and the research labs of these companies will want to get ahead of the curve
Files
Why might you be wrong?
it's really hard to detect AI-written *short* messages, which a lot of these platforms encourage
Files
Why do you think you're right?
Why might you be wrong?