srl

Probably dominated by death or incapacitation, which at 85 years is around 15% per year (assuming nothing is known about the individual's health, which is probably close to right for Khamenei). Everything else seems unlikely: assassination would be extremely counterproductive for Israel (and Israel has demonstrated the ability to behave rationally); fleeing the country would imply a revolution, with a per-year probability less than 5%, resignation (absent incapacitation) would require a solid succession plan which doesn't seem to be in place.

With only a little more than a month left, all of these should be below 10% just because the per-year probability is less than 100%. Kuwait, and Tunisia seem more hostile, SA and Oman less. Qatar's behavior is bizarre and I don't know what to think. I'll give the hostile ones a 10% per-year probability (that's the minimum anyway), SA and Oman 20%, and Qatar 15% (rounded to 20%).

50/50 on this happening by 2040, with most of that concentrated before 2030, so let's say 30% by 2030. That's five years, giving 3% in the next six months. My vague understanding is that the probability should be even lower for the near term because PLA doesn't have all the material it needs prepared (and I can't imagine what advantage would be gained from an isolated attack). Happy to put 2%.

Naive base rate calculation (assuming Poisson): six tests in the last 20 years gives an expectation of 0.15 tests per six months.

Taking presidential election timing into account doesn't increase the probability that much. Taking the same window around inauguration, the only test that qualified was February 2013. There have been four inaugurations since the first test (2009, 2013, 2017, 2021), so that gives a 25% probability, with large error bars.

Difficult to see what NK would gain from this sort of timing of a test, but also I don't think of NK's behavior as being entirely rational. Sticking with base rate.

I don't see what advantage Iran would have by withdrawing. Lose some diplomatic benefits and relief from sanctions, and simultaneously lose the ability to blame the U.S.

Were Khamenei to lose power in the next 40 days, this would still be overwhelmingly unlikely. A reformist government would seek a return to compliance and an expansion of JCPOA; a hardline government would still want to remain in JCPOA.

Iran's government has shown an ability to think somewhat strategically, so the fact that this is clearly a negative for the country _and_ the government is strong evidence that it won't happen.

Add that to timing considerations.

Flattening of scaling laws should tighten competition, making the outcome more random and reducing the impact of the advantages that U.S.-based companies have.

The catalog linked in the question reveals that 15 certificates have been issued so far. Plenty of OTICs have not yet issued any certificates:

• ritt7layers and Singapore in Asia
  
• Berlin, Madrid, Paris in Europe
• All of the NA ones

Overall, that's about half of OTICs excluding the NA ones. If I include NA ones that previously existed, then maybe 1/3rd of OTICs have issued a certificate. It seems almost certain that in the next few years, at least one NA OTIC will issue a certificate.

Another observation. About 7 certificates were issued in 2023. Then I expect 2 or 3 to be issued in the remaining 4.5 months. Each has (at best) a 1/3 chance of coming from an NA OTIC, which results in a ~50% probability of one of them being from NA. This assumes nothing's correlated which is kind of silly, but I'm not at all sure which way the correlation should go!

The major RAN vendors are apparently Ericsson, Huawei, and Nokia, which together have about three-quarters market share. Both Nokia and Ericsson are members of the O-RAN alliance. I can't find when they joined. Ericsson has been a member at least since late last year.

ZTE is sometimes listed, along with Huawei, as a telecom company that causes security concerns for the west. ZTE is a member of the O-RAN alliance. This has been true at least since 2020. The alliance was founded in 2018.

The resolution criteria are pretty broad as well.

It seems like a 50% chance that Huawei is a member at the end of the decade, leaving something like a 10% chance that it's a member in a year. That's a sort of lower bound on this probability, and then I have to worry about the chance of a "collaboration", which must be at least as likely.

Since the crowd forecast is so low, I'll go with a lowest probability I can justify.

Going (slightly) higher than the crowd here. Increased volatility (Israel, Trump winning) pushes probabilities away from the extremes.