Based on I'm about 80% confident that this could be done if it were prioritized / funded (70% this happens), but, for all I know, this might get crowded out by other priorities (or as @belikewater pointed out, this could be derailed by war).
-0.603605
Relative Brier Score
Questions Forecasted
Scored Questions
282
Forecasts
59
Upvotes
Forecasting Activity
Forecasting Calendar
| Past Week | Past Month | Past Year | This Season | All Time | |
|---|---|---|---|---|---|
| Forecasts | 1 | 20 | 201 | 118 | 861 |
| Comments | 1 | 16 | 100 | 94 | 268 |
| Questions Forecasted | 1 | 15 | 43 | 30 | 120 |
| Upvotes on Comments By This User | 0 | 2 | 40 | 21 | 169 |
| Definitions | |||||
New Prediction
New Badge
Star Commenter - Jul 2025
Earned for making 5+ comments in a month (rationales not included).
New Prediction
Probability
Answer
56%
Yes
44%
No
Why do you think you're right?
Files
Why might you be wrong?
It seems like this doesn't require breakthroughs, but just engineering, but maybe this is more complicated than I'm imagining. I can think of other problems that are mostly engineering challenges that have been hard to crack (e.g. fusion).
Files
New Prediction
Probability
Answer
0%
(-1%)
Yes
100%
(+1%)
No
Why do you think you're right?
It was already unlikely to happen due to the limited time, but @404_NOT_FOUND 's recent comments convinced me that the chances are low enough that I can round them down to zero. Potentially angering the Trump administration is yet another reason why this is unlikely.
Files
Why might you be wrong?
As I've mentioned before, an announcement is a very low bar. If it happens I'll be quite surprised but every now and again a 1/250 chance happens.
Files
New Prediction
Probability
Answer
27%
(-1%)
Yes
73%
(+1%)
No
Why do you think you're right?
Lowering slightly on the lack of news. As I mentioned in my last forecast, I'm guessing the majority of Chinese military efforts between now and the end of this question will be focused on areas applicable to Taiwan.
Files
Why might you be wrong?
A focus on capabilities related to Taiwan doesn't preclude expansion in Africa, and maybe it will become a priority, particularly if investments in nearby areas are increased.
Files
New Prediction
Probability
Answer
0%
(0%)
Less than or equal to 9%
4%
(0%)
Between 10% and 19%, inclusive
78%
(0%)
Between 20% and 29%, inclusive
17%
(0%)
Between 30% and 39%, inclusive
1%
(0%)
More than or equal to 40%
Confirmed previous forecast
Files
New Prediction
Probability
Answer
0%
(-1%)
Less than or equal to 59
9%
(-1%)
Between 60 and 69, inclusive
65%
(+7%)
Between 70 and 79, inclusive
25%
(-4%)
Between 80 and 89, inclusive
1%
(-1%)
More than or equal to 90
Why do you think you're right?
Looks like we're on course to hit the middle bin. As time progresses, the range of outcomes is narrowing. The main thing that I think could change this is if there are media campaigns. For example, if we're in a campaign now and it ends abruptly we could end in the second bin. Similarly, if one starts or ramps up, there is still a chance of ending in the highest bin. From what I can gather though, it doesn't appear that the campaign model is happening and it seems there's just a slow and relatively steady stream of cases.
Files
Why might you be wrong?
As I mentioned in my comment, if the campaign model turns out to be correct it could push the answer toward the extremes.
Files
New Prediction
Probability
Answer
1%
(-2%)
Angola
1%
(-1%)
Kenya
3%
(-6%)
Ethiopia
1%
(-1%)
Nigeria
Why do you think you're right?
Time is running out, so it's time to start lowering. I don't see any news to keep me from decreasing.
Files
Why might you be wrong?
I could easily have missed something material in my searches.
Files
New Prediction
Probability
Answer
5%
(+1%)
Yes
95%
(-1%)
No
Why do you think you're right?
Increasing slightly. As per this article, "OpenAI opted to treat the new model as having a high biological and chemical capability level in its preparedness framework". They do point out that this model doesn't necessarily have a high risk-level yet, but it may come. This is mostly expected as capabilities increase, but I was probably just a smidge too low previously.
Files
Why might you be wrong?
We're a long way from 2030 and I don't want to be overconfident.
Files
Why do you think you're right?
Why might you be wrong?