michal_dubrawski

Michał Dubrawski
About:
Show more

-0.07018

Relative Brier Score

88

Forecasts

454

Upvotes
Forecasting Activity
Forecasting Calendar
 

Past Week Past Month Past Year This Season All Time
Forecasts 6 11 59 54 186
Comments 8 15 111 95 442
Questions Forecasted 6 6 10 8 19
Upvotes on Comments By This User 10 25 214 192 800
 Definitions
New Prediction
michal_dubrawski
made their 2nd forecast (view all):
Probability
Answer
Forecast Window
13% (-5%)
Yes
Nov 21, 2024 to May 21, 2025
87% (+5%)
No
Nov 21, 2024 to May 21, 2025

I am slightly reducing based on great feedback from @NukePirate (here) and @geoffodlum (feedback in team discussion + his forecast here). I created a judgmental poll to learn more how our team reads the situation. 

To do that I collected some statements from my own forecast and other comments - would you like to share how likely do you think they are true? Here is my judgmental poll as a google form:

https://forms.gle/RGEmDAmBxVet8xE46

If you don't have a clue about one of them, you can skip the question - only the one about one's current forecast is set as required to answer.

Results will be visible here: https://docs.google.com/spreadsheets/d/1WtzZ58z-Fb8OBOJidJDJpSngIyYgy5sAt_CnbrIHg44/edit?usp=sharing


Here is my list of statements to be assessed (they are not necessary my own views):


The kiloton-to-kilogram efficiency of North Korea’s nuclear warheads is likely comparable to US weapons developed in the 1960s and 1970s (0.1-3 kt/kg)


Russian expertise is enhancing North Korea’s nuclear weapons development, including yield and design improvements.


New designs of nuclear weapons require testing to ensure reliability and performance.


North Korea is working on a new nuclear design with a yield higher than 250 kilotons.


Russian support is accelerating North Korea’s progress in nuclear development.


China really (not just officially) wants North Korea to refrain from conducting any further nuclear tests.


North Korea is determined to conduct another nuclear test at some point in the future.


Past warnings about nuclear tests were not necessarily false alarms but reflect a gradual development process.


China would impose significant pressure on North Korea if test preparations were discovered.


North Korea could successfuly use techniques similar to India’s to conceal nuclear test preparations from international observation, including from China and the US.


North Korea’s current geopolitical strategy makes a nuclear test within the next six months advantageous from their perspective


North Korea is working on new nuclear weapon designs.


The probability of a nuclear test increases over time.


North Korea relies heavily on its nuclear program to maintain international relevance.


North Korea is in a stronger position now due to support from Russia and China, reducing the need for saber-rattling.


The absence of visible test site activity, a reliable indicator of no imminent test.


In the short term (at least 6 months) North Korea will focus on missile launches or other provocations rather than nuclear tests.


North Korea’s collaboration with Russia makes a nuclear test less likely in the short term (at least 6 months).

I also asked a fellow forecaster and nuclear physicist if he could answer some technical questions we have, I hope to get a better understanding  soon based on this work and possible further feedback. 

Files
michal_dubrawski
made a comment:
@michalbod thank you very much for your feedback and insights. Great point about China's impact on Russia and their likely ability to prevent Russia from technological transfer related to nuclear warheads' technology. Your example of Russia not providing S400 and Su35 to Iran likely because of Saudi Arabia makes sense, and it shows the complexity of the international situation in which Russia currently is. As someone else mentioned in the comment to their forecast, if Russia counts on their gains from negotiations with Trump over Ukraine, such test from North Korea may not be good for them. I wonder, if Trump would turn out not that willing to negotiate with Putin after all (which I am afraid is the less likely outcome), and if the Trump administration will continue the US support for Ukraine, maybe then the North Korean test could be valuable for Putin as another mean of putting pressure on the US and the West? The test is not a threat itself, but if there is also a narrative from Russia of supporting North Korean nuclear arsenal development in case yet another of their so-called "red lines" is crossed or if they won't get something which they want. I am not saying that it would be effective, just considering if that could be something they would try.
Files
New Prediction
michal_dubrawski
made their 1st forecast (view all):
Probability
Answer
Forecast Window
18%
Yes
Oct 31, 2024 to Apr 30, 2025
82%
No
Oct 31, 2024 to Apr 30, 2025
Why do you think you're right?

I start with 18%, I read some really great comments by other pros. Some ideas that came to my mind:
I read a good argument that they may not need testing since they already tested their nuclear bombs in the past and have something that works. In the last test in 2017 the blast was estimated to be around 250 kilotons. How much is it? For the purpose of calibrating our mental models we can use:https://nuclearsecrecy.com/nukemap/ If this will not be too disturbing for you, you can see the blast area for such a nuclear detonation in some cities which you know from your experience (have a mental representation of their size). Examples:
Washington DC: https://nuclearsecrecy.com/nukemap/?&kt=250&lat=38.895037&lng=-77.036543&airburst=0&hob_ft=0&psi=20,5,1&zm=12
London: https://nuclearsecrecy.com/nukemap/?&kt=250&lat=51.5073219&lng=-0.1276474&airburst=0&hob_ft=0&psi=20,5,1&zm=11

Seoul: https://nuclearsecrecy.com/nukemap/?&kt=250&lat=37.5666791&lng=126.9782914&airburst=0&hob_ft=0&psi=20,5,1&zm=12
Warsaw: https://nuclearsecrecy.com/nukemap/?&kt=250&lat=52.231958&lng=21.006725&airburst=0&hob_ft=0&psi=20,5,1&zm=11
Can we look at North Korean Ballistic Missiles payload capabilities to say something about warheads they intend to carry on them? 38North have estimated payload of North Korean Hwasong-16 ballistic missile payload to be about 2,500 to 3,000 kg,

How many kilotons should we expect per kilogram? This article is great in answering that:
https://blog.nuclearsecrecy.com/2013/12/23/kilotons-per-kilogram/
"all US weapons currently in the stockpile end up, in the 0.1-3 kt/kg range, especially the 2-3 kt/kg range"
So only if we use the absolutely lowest values of this ratio and the lower end of the payload estimate, we will get 250 kilotons, but as some of the nuclear weapons in the US nuclear stockpile were developed in the late 1960s and 1970s https://nuclearweaponarchive.org/Usa/Weapons/Allbombs.html and with the possibility of Russia sharing their nuclear know how along with some other technical expertise with North Korea in exchange for weapons and soldiers, I am not an expert, but I think we can expect that they will be able to achieve better kiloton to kg ratio than 0.1 kiloton/kg. So I think they aim for something bigger than 250 kilotons, especially now, when they may have access to Russian expertise. I may be wrong, since my understanding is limited, but I think they are now likely working on some new design with higher yield, rather than only making more of the current ones.
Another point: other pros were right to point out that the warnings about the test keep repeating and nothing happens. However, as I read some of the warnings from last two years, one (IIRC it was  from 2022) was saying explicitly that the test is not imminent but that North Korea is working on the technology and the test is likely in the future, and the one from 2023 said about the test "next year", So maybe some of these were not necessary false alarms, or at least we do not know yet. 
One more point that comes to my mind: I love the idea that we would be able to see preparations for the test on the satellite images so if we cannot see them now it will likely not happen in the next two or three months, but I keep thinking - if North Korea is determined to conduct the test (they may want to test improvements, especially if there is/will be know-how transfer from Russia - I assume that they will want to test it, that testing is needed for new designs or at least they will want it to be sure) they may want to try hiding the preparations for the test and avoid international pressure not to do it (I would expect pressure if everyone could read about the preparations). Is it possible? It would be hard, but India was successful. Also, the problem with any open-source intelligence methods which results are published online is that the people whom we observe can access these analyses as well and they will know what to hide from us. As a counter-argument, it is one thing to play this tactic with the West, but they also have to think about their relationship with China, which is against North Korean nuclear tests.

I think the probability slightly increases over time because this is a process that is progressing, and now they may have Russian help. At the same time, I don't think they will do the test that fast. Of course, I may underestimate the importance of political factors here, like nuclear diplomacy with the new US President for example - if Kim Jong Un decide that it may give him political benefits, then testing even the same nuclear bomb as before would be reasonable.

Files
Why might you be wrong?

I wonder if North Korean government affiliated forecasters were asked what would be a result of conducting a nuclear test before the US Elections on their result... This is not impossible.

Files
michal_dubrawski
made a comment:
@srl I noticed your high quality forecasting rationales and followed you. Back then I also noticed in your description that you are a physicist, and as I just checked, you are a nuclear physicist, right? I do not know if you are affiliated with RAND, and even if you are, I don't have any power to ask you for help as a part of your work, but I wanted to ask: Would you be interested in sharing some of your technical expertise with us?

I collected some statements from my own forecast (above) and other forecasters comments to collect our pro forecasters educated guesses (or in some cases, answers based on some level of expertise rather than guesses), but I thought that with technical questions you may be able to answer more like "yes" or "no" or "it depends on..." Here is my judgmental poll:https://forms.gle/RGEmDAmBxVet8xE46
Results will be visible here: https://docs.google.com/spreadsheets/d/1WtzZ58z-Fb8OBOJidJDJpSngIyYgy5sAt_CnbrIHg44/edit?usp=sharing
Here is my list of statements to be assessed (they are not necessary my own views - I highlighted the technical questions with bold but if you are interested, feel free to answer any of them):
The kiloton-to-kilogram efficiency of North Korea’s nuclear warheads is likely comparable to US weapons developed in the 1960s and 1970s (0.1-3 kt/kg) - (that was my guess) my friend @geoffodlum suggests that their designs may rather be comparable to US designs in the 1970's or early 1980s
Russian expertise is enhancing North Korea’s nuclear weapons development, including yield and design improvements.
New designs of nuclear weapons require testing to ensure reliability and performance.
North Korea is working on a new nuclear design with a yield higher than 250 kilotons.
Russian support is accelerating North Korea’s progress in nuclear development.
China really (not just officially) wants North Korea to refrain from conducting any further nuclear tests.
North Korea is determined to conduct another nuclear test at some point in the future.
Past warnings about nuclear tests were not necessarily false alarms but reflect a gradual development process.
China would impose significant pressure on North Korea if test preparations were discovered.
North Korea could successfuly use techniques similar to India’s to conceal nuclear test preparations from international observation, including from China and the US.
North Korea’s current geopolitical strategy makes a nuclear test within the next six months advantageous from their perspective
North Korea is working on new nuclear weapon designs.
The probability of a nuclear test increases over time.
North Korea relies heavily on its nuclear program to maintain international relevance.
North Korea is in a stronger position now due to support from Russia and China, reducing the need for saber-rattling.
The absence of visible test site activity, a reliable indicator of no imminent test.
In the short term (at least 6 months) North Korea will focus on missile launches or other provocations rather than nuclear tests.
North Korea’s collaboration with Russia makes a nuclear test less likely in the short term (at least 6 months). 



If this request from me would disturb your work in any way, please ignore it.
Files
New Prediction
michal_dubrawski
made their 3rd forecast (view all):
Probability
Answer
Forecast Window
16% (+2%)
Yes
Nov 20, 2024 to Nov 20, 2025
84% (-2%)
No
Nov 20, 2024 to Nov 20, 2025

I am increasing, very slightly, treating reports about the Iranian Assembly of Experts being tasked to choose Khamenei's successor as a weak signal at best, as there were such reports in the past. I think the quote from this article describes it best:

"On November 7, Iranian regime supreme leader Ali Khamanei met with the Assembly of Experts and discussed plans for his successor. “The Assembly of Experts, as defined in our Constitution, is one of the most revolutionary institutions of the Islamic Republic,” he said. “Its significance lies in its unique responsibility to select the nation’s leader. This readiness to fulfill that role must always be maintained.”
One day after Khamenei’s remarks to the Assembly of Experts, where he emphasized the need for their readiness to select the next Supreme Leader, Mohammad Bagher Heydari Kashani, a member of the Assembly, described both the speech and the atmosphere as such: “We held our breath. He gave explicit orders for the Assembly to act swiftly and decisively, without hesitation, to appoint the next leader in accordance with constitutional guidelines… The mere thought of the Leader’s absence moved some members to tears; they couldn’t hold back their emotions at the possibility of the Supreme Leader’s martyrdom.”

These remarks sparked widespread controversy within the regime and surfaced in Friday prayer sermons on November 15 across various cities. Several of Khamenei’s appointees spoke on the issue, their conflicting statements revealing the confusion and anxiety caused by the anticipation of Khamenei’s potential demise at the highest levels of the regime.
During Friday prayers in Shahrekord, Abdollah Keyvani, another member of the Assembly of Experts, outright denied Heydari’s account: “Someone said the members of the Assembly cried… That’s not true; there was laughter, joy, and a sense of empowerment in the session. We were delighted, but there weren’t even tears of joy. The claim that the Leader directed the Assembly to discuss the next Supreme Leader is completely untrue.”
In Isfahan, Abolhassan Mahdavi, Khamenei’s representative and the Friday prayer leader in the city, remarked, “Khamenei’s statement that the Assembly should identify suitable candidates for leadership in case of emergency isn’t new; I’ve heard him mention this about 17 or 18 times over the past 20 years.” He continued, “There is a committee within the Assembly of Experts responsible for identifying all individuals qualified for leadership—this extends beyond Iran, as the concept of ‘Velayat-e Faqih’ transcends borders. This committee works under strict confidentiality, and three members have been chosen to securely hold the names of potential candidates.”

Yeah, we don't know if Mahdavi is saying the true, but there were some media reports about the Assembly of Experts working on that issue already in the past (for example here).

I think the rumors about the Khamenei being in coma are just rumors. I know his age is serious, and his health was impacted by surviving a bomb assassination attempt many years ago, but the base rate about such rumors being true is really making me think this is very unlikely true.  There is this image of his meeting with Mojtaba Amani dated to happen on November 17, in Tehran. Yes, that could be fake if Iranian top players want to gain some more time before announcement, but it is much more probable that it is just another example of this media effect (for the reference, see this great comment by shibboleth about media reactions to Kim Jong-un disappearances: https://www.metaculus.com/questions/3445/will-kim-jong-un-still-be-the-de-facto-leader-of-north-korea-on-1-january-2022/#comment-28950).  

Here is my own analysis of a similar phenomenon related to the Iranian Supreme Leader:

All the deaths of Ali Khamenei

1. September 11-20, 2022

    Sources:

     Express.co.uk "Iranian Supreme Leader Ali Khamenei 'on death bed with cancer' – regime risks 'imploding'" https://www.express.co.uk/news/world/1667757/Iran-Supreme-Leader-Ali-Khamenei-heath-fears-dead-cancer-regime-collapse-Ayatollah

      Report: Claims surfaced that Khamenei was on his deathbed due to advanced cancer. Speculation arose about potential succession struggles, particularly involving his son Mojtaba and President Ebrahim Raisi.

        Atlantic Council: "The Supreme Leader is still alive. But when he does eventually die, how will succession play out?", https://www.atlanticcouncil.org/blogs/iransource/the-supreme-leader-is-still-alive-but-when-he-does-eventually-die-how-will-succession-play-out/

            Report: Addressed rumors of Khamenei's death after two weeks of absence. On September 17, 2022, he reappeared briefly to dispel rumors.

2. December 6–7, 2020

    Sources:

        Jewish Press: "Unconfirmed Reports: Ayatollah Ali Khamenei Is Dead" (https://www.jewishpress.com/news/politics/unconfirmed-reports-ayatollah-ali-khamenei-is-dead/2020/12/06/)

        Times of Israel: "Reports of Khamenei’s death greatly exaggerated, Iranian official says" (https://www.timesofisrael.com/reports-of-khameneis-death-greatly-exaggerated-iranian-official-says/)

        International Business Times: "Is Ali Khamenei dead; rumours are rife amid Iran Supreme Leader's health concerns" (https://www.ibtimes.co.in/ali-khamenei-dead-rumours-are-rife-amid-iran-supreme-leaders-health-concerns-830964)

    Report: Rumors about Khamenei's death and power transfer to his son Mojtaba were triggered by journalist Mohamad Ahwaze. Officials denied these rumors, calling them fabrications by enemies. Iranian media emphasized Khamenei’s good health.

3. March 8, 2015

    Source: RFERL

    Title: "Iran Supreme Leader Makes Public Appearance Amid Ill Health Rumors"

    Report: Khamenei reappeared in public amid rumors of hospitalization and critical illness following prostate surgery in 2014.

    URL: https://www.rferl.org/a/iran-supreme-leader-public-appearance-khamenei/26888421.html

4. November 30, 2010

    Source: SBS News

    Title: "Iran's supreme leader 'has terminal cancer'"

    Report: A U.S. diplomatic memo revealed claims that Khamenei had terminal leukemia and might die within months. No further verification was provided.

    URL: https://www.sbs.com.au/news/article/irans-supreme-leader-has-terminal-cancer/n2q30g6tv

5. October 2009

    Sources:

        Jerusalem Post: "Rumors say Ayatollah Khamenei in coma" (https://www.ynetnews.com/articles/0%2C7340%2CL-3790774%2C00.html)

        YNetNews: "Iranian website: No truth to rumors that Khamenei is in a coma" (https://www.ynetnews.com/articles/0%2C7340%2CL-3790774%2C00.html)

    Report: Rumors emerged that Khamenei had fallen into a coma or collapsed. Iranian officials dismissed these reports as baseless Western propaganda.

6. January 7, 2007

    Sources:

        YNetNews: "Online rumors: Iran's supreme leader dead" (https://www.ynetnews.com/articles/0,7340,L-3348764,00.html)



        YNetNews: "Another official denial: Khamenei not dead" (https://www.ynetnews.com/articles/0,7340,L-3349358,00.html)

    Report: Speculation over Khamenei’s death following cancer treatment circulated widely. Officials denied the claims.

I also found some interesting opinions about the role of the Experts Assembly:
https://www.iranintl.com/en/202401143718#:~:text=The%20Expert%20Assembly%20lacks%20legitimacy%20among%20insiders%2C%20as%20it%20comprises%20low%2Dlevel%20clerics%2C%20who%20have%20not%20won%20their%20seats%20in%20competitive%20and%20free%20elections%20but%20essentially%20appointed%20to%20their%20positions.
https://www.iranintl.com/en/202403069020#:~:text=Mahdavi%20Azad%20observed,decisions%20made%20elsewhere.

Files
New Prediction

I am not reducing yet (likely next time). It is worth watching if there will be another push from the Trump administration directed at EU and NATO partners to remove Huawei software and equipment from mobile networks. So far, the results of such encouragements from the US were limited in Europe. 

Two new articles from lightreading.com:
"Ericsson 'open RAN' for MasOrange may shut door for Huawei"

"MasOrange, Spain's biggest mobile operator, seems to be following AT&T into a single-vendor open RAN deal with Ericsson.

(...)
Ericsson's new 5G contract with Spain's MasOrange suggests AT&T might no longer be alone.


The five-year deal announced today is promoted under the label of open RAN, whose interfaces should theoretically allow MasOrange to combine parts from different vendors instead of buying a fully integrated product set from Ericsson. Yet this arrangement appears to offer even less opportunity to others than AT&T's did. The US telco had at least named Fujitsu, a longstanding Ericsson partner, as a provider of radio units in certain areas. Some IT components would come from Dell and Intel, it also said. The release about MasOrange, by contrast, identifies no other technology partner.


There were, however, already some big differences between the two operators. In the world of traditional RAN, big telcos have usually divided their networks between two or three suppliers on geographical lines. Ericsson previously served about two-thirds of the AT&T footprint. Nokia, which had the rest, is now being torn out. For all the talk about open RAN aiding the construction of multivendor networks, AT&T wanted a single provider of critical RAN elements across the whole US, said Ericsson earlier this year.


Judging by earlier press reports, one of MasOrange's primary motivations is to flush out the Chinese. The operator came into existence in March this year when MásMóvil and Orange Spain, formerly separate mobile networks, completed a merger. Orange Spain had introduced ZTE into its network in 2019, and the combined MasOrange network also relied on both Ericsson and Huawei, according to Expansion newspaper. But MasOrange would be phasing out the two Chinese suppliers as part of an open RAN deal with Ericsson, reported Expansion in July.


It's the highway for Huawei


MasOrange's press office declined to comment on that report at the time, but the contract publicized this week suggests much of it was accurate. Ericsson is to replace ZTE, said to account for roughly 4% of the MasOrange footprint, by the end of 2027, according to the Expansion report. Over the same period, Huawei, the bigger of the two Chinese vendors, will be reduced from 54% to 39% of the footprint. That would mean scrapping its products across 1,600 sites in the regions of Extremadura and Andalusia, as well as the Canary Islands, reported Expansion.


The implication is that MasOrange has a national footprint of about 10,700 5G sites, of which Ericsson will eventually account for 61%, or about 6,500. But the Swedish vendor was reportedly in line to replace Huawei across all remaining sites in 2028 and 2029, becoming the sole RAN vendor. A Reuters report out today said the deal with Ericsson covers about 10,000 sites, according to a source familiar with the matter."

"Docomo pitched tech Deutsche Telekom could use to handle Huawei"

"A system sold by Docomo's new OREX subsidiary is the answer to managing a multivendor RAN, says the Japanese company.

Nine months before Germany imposed limits on Huawei, Japan's NTT Docomo was flaunting a related software tool called service management and orchestration (SMO, for short). Used to control and do automated updates of the radio access network (RAN), it is positioned as a successor to today's conventional management systems. If a German telco has bought its RAN from Huawei, it must find a substitute for the management system by 2029, under government rules published in July. Docomo's SMO technology, sold through an offshoot called OREX SAI, might hold the answer."

Files
New Prediction
michal_dubrawski
made their 7th forecast (view all):
Probability
Answer
1% (+1%)
Estonia
1% (+1%)
Latvia
1% (+1%)
Lithuania

Great comments by @Rene here@404_NOT_FOUND here,  @VidurKapur here@geoffodlum here and @Jim here.
This still seems very unlikely, but the uncertainty has increased with the Trump victory.

Some related recent articles:
BBC: "In Russia's shadow: The Baltics wait for Europe's strategic new railway" 
CNN:  "European officials cry sabotage after two internet cables are cut in the Baltic Sea"


European officials are looking toward Russia after two submarine internet cables in the Baltic Sea were suddenly disrupted in an apparent sabotage operation, just weeks after the United States warned that Moscow was likely to target critical undersea infrastructure.

A cable between Lithuania and Sweden was cut on Sunday, according to Telia Lithuania, the telecommunications company that runs the link. Separately, the state-controlled Finnish telecoms company Cinia said one of its cables, which connects Finland and Germany, was disrupted on Monday.

Foreign Policy Research Institute: "Contesting Russia: the Balti perspective" 

TSN.UA "Russia is preparing actions against the Baltic states: NSDC outlines conditions under which they could be implemented" 

The head of Ukraine's Center for Countering Disinformation criticized Europe's stance, comparing it to an ostrich hiding from reality.
The Kremlin is preparing for actions against the Baltic states, and under current circumstances, Russia would need 4-6 years to execute them.
This was stated by Andriy Kovalenko, Head of the Center for Countering Disinformation at Ukraine's National Security and Defense Council.
He says that Russia is currently able to convince Europe that it has no intention of attacking them in the future.
"Of course, they do intend to. They are seeking ways to gather and accumulate resources, preparing, and launching information attacks against NATO countries to politically weaken governments and destabilize the Alliance itself."
"European politicians mostly lack the inner resolve to speak to their societies about the real threat. They don't want to lose popularity over decisions requiring greater mobilization of societies and economies," Kovalenko stated.
He warns that Russia is preparing actions against the Baltic states, and under current conditions, they would need 4-6 years to achieve this.
"The timeframe that will occur if Europe continues to consider the war with Russia as not their problem. Europe resembles an ostrich burying its head from reality, and is satisfied with Russian disinformation about the absence of aggressive intentions," the head of the Center for Countering Disinformation summarized.

Files
New Prediction

I am lowering significantly because I still do not see anything when I search preprints and science news. It is possible that January will bring large update to the NASA exoplanet archive, but knowing that there is a lag we should be seeing reports and preprints about at least some of these exoplanets and we don't. This wikipedia page lists some exoplanets which are supposed to be discovered this year and mention that some are in Habitable Zone, but unless I am missing something 82 G. Eridani d was discovered in 2011: https://www.stellarcatalog.com/exoplanet.php?planetID=100060 another planet this source reports is just a candidate, not a confirmed exoplanet.

Files
New Badge
michal_dubrawski
earned a new badge:

Active Forecaster

New Prediction

I really don't think anything will happen in such a short time left till this question resolution date, as @PeterStamp mentioned here they still have job postings for AI related jobs in Beijing, even if the AI Research Lab is not mentioned. For example: https://jobs.careers.microsoft.com/global/en/share/1782091/?utm_source=Job Share&utm_campaign=Copy-job-share 

At the same time, the trends do not look well for the prospects of Microsoft keeping the AI Research Lab in China for the longer term.

Recently, we have seen the news that "Llama 13B has reportedly been adapted for a Chinese military intelligence tool known as ChatBIT": https://www.pcmag.com/news/chinese-researchers-make-military-ai-using-meta-llama-13b

Soon after that news, Meta changed their policy, and announced "that it would allow US national security agencies and defense contractors to use its open-source artificial intelligence model, Llama.": https://www.theguardian.com/technology/2024/nov/05/meta-allows-national-security-defense-contractors-use-llama-ai 

 https://www.aljazeera.com/economy/2024/10/29/us-finalises-restrictions-on-ai-semiconductor-investments-in-china

Also:

The United States has finalised rules that limit investments in critical technology sectors in China such as artificial intelligence on national security grounds, the Treasury Department has announced.

The restrictions will bar US citizens and permanent residents, as well as US-based companies, from engaging in transactions involving technologies including AI, semiconductors and quantum computing, the Treasury said in a statement on Monday.

Source: https://www.aljazeera.com/economy/2024/10/29/us-finalises-restrictions-on-ai-semiconductor-investments-in-china


Also:


Donald Trump launched a technological war against China in his first presidency. His next administration faces a much harder task: finishing it.

Former officials say the incoming administration is expected to expand and deepen measures to hobble China’s technological development, using tariffs as a bulwark and export controls as a slashing blade. They say Republican officials poised for senior positions will be less patient than their Biden administration counterparts in negotiating with China and will seek to move fast, even at the expense of breaking things.

(...)

In the crosshairs are the critical technologies that underpin both nations’ military strength and economic wealth, especially advanced artificial-intelligence algorithms and the precision-made chips that are their physical manifestation. Officials in Washington and Beijing believe this technological race may well determine which nation emerges as the world leader for the next generation.

Nikakhtar predicts the new administration will be disposed to introduce new rules to further restrict U.S. exports of advanced technologies to China, looking beyond AI chips toward other emerging technologies such as quantum computing, robotics and biotechnology. There’s also likely to be a review to close “gaps” in earlier rules that allowed China to get around them, she said.

For consumers, that might mean wild swings in prices of technology products as Washington and Beijing wrestle for control of the supply chain and the global market. New tech industry jobs will be created across the United States, subsidized temporarily by taxpayer dollars, as policymakers cross their fingers that they will become self-sustaining.
While Trump’s Cabinet is still being formed, it’s already clear it will be weighted toward security hawks. Trump announced Sen. Marco Rubio (R-Florida) as his pick for secretary of state Wednesday and has asked Rep. Mike Waltz (R-Florida) to be the White House national security adviser; the two lawmakers have been leading critics of China in Congress. Robert E. Lighthizer, the architect of Trump’s earlier trade war against China, is widely expected to return in a senior role. Rep. Elise Stefanik (R-New York), who has pushed for stricter technology controls against China, was announced Monday as Trump’s pick for U.S. ambassador to the United Nations.


source: https://www.washingtonpost.com/technology/2024/11/14/us-china-ai-tech-export-controls/


Also:


Increased China focus: In addition to strengthening export control, the new Trump administration will likely focus on competition with China in other ways. While the Biden administration took steps to stay ahead of China in AI, it also opened avenues for bilateral engagement on managing AI risks. If, as expected, President Trump imposes tariffs on a wide range of goods from China, continued bilateral engagement on AI will be less likely.

In part in response to recent reports of China’s growing military and intelligence AI activities, the new administration will likely ramp up work on military and intelligence use of AI—including by increasing collaboration with American AI companies. Among other consequences, this will likely involve adopting a less risk-averse approach than is reflected in the Biden administration’s October 2024 National Security Memorandum on AI.

source: https://www.brookings.edu/articles/ai-policy-directions-in-the-new-trump-administration/


Files
New Badge
michal_dubrawski
earned a new badge:

Star Commenter - Oct 2024

Earned for making 5+ comments in a month (rationales not included).
New Prediction

I am going up because of uncertainty related to my understanding of the situation (thanks to my friends @ctsats and @DimaKlenchin for helpful comments). I previously assumed that the schedule to check for new matches weekly is something new, something related to the reorganization of HWC in December last year (we saw how they changed the website and the catalog name). But maybe the fact that previous updates changing the number of exoplanets in the catalog happened most often in December and January reflects some pattern of data in the NASA exoplanets catalog being updated more often around this time (adding data from papers that first come out as preprints and were later published during the year). But if that was true, we should at least be seeing most of these papers as preprints already in https://arxiv.org/ database, and I haven't seen them, but my search may be insufficient - some papers confirming new exoplanets may not discuss that they are in habitable zone and I will not be able to find it in the haystack of preprints. 

I checked the number of exoplanets meeting the first two out of three criteria of habitability used by HWC (mass and radius) and here are the numbers in the NASA Exoplanets catalog:
2024: 64 (so far)

2023: 126

2022: 122

2021: 302

2020: 108

2019: 75

2018: 163

2017: 66

2016: 1111

2015: 56

2014: 602

2013: 35

2012: 29

So far, we have about 50% of numbers from 2023 or 2022, but there is high variability in the numbers for years before 2022, so it is hard to say anything about the reporting speed based on that. However, in 2023 there were discovered 6 exoplanets which were added to the HWC, and in 2022 there were 5, so we are a bit unlucky with zero exoplanets halfway to getting to their numbers of new exoplanets fitting mass and radius criteria.

As you can see below, historically, HWC was more active during December and January (data from the HWC updates database I created). 


But in terms of updates changing the number of total exoplanets, January and July are on the podium, with February, May and June standing on the lower place on the podium together. This includes revisions related to removals.

Here is how it looks when we only look at additions (this question resolution only cares for new planets being added, removals are unimportant for our purposes, but removals likely happened due to new data being published or new methods and models providing more precise answers). We can see that we are already past the time in the year with most updates, but as January next year still counts, it is not that simple. 


We should keep in mind that likely over the year situation changes - like with some missions starting and ending and their teams way of reporting being different, or as Christos was informed by the professor running HWC the process has changed over the years, and it is now largely automated. Look at this graph to see changes in time:


To better see the months of updates, here is another version with month number on the y axis:



I am also tagging:
@404_NOT_FOUND @cmeinel @Perspectus @guyrecord @sai_39 @ansantillan @Jim @NoUsernameSelected @JonathanMann @ansantillan @WeirdAwkward @JJMLP  @Plataea479  @MrLittleTexas  

Files
michal_dubrawski
made a comment:
Thanks @DimaKlenchin , I  used both methods (Google search with site: can access article full text while ArXiv advanced search is good because it can help search abstracts and other fields). What I was also searching for was information about "habitable zone" or "HZ", because the number of new discoveries is much bigger than the new discoveries within HZ.  At the same time, if we had a text mining method of reporting all the planets based on abstract, I could then compare it with the NASA exoplanet archive and see how many were not yet added. This question's complexity fascinates me, but the practical question is if in this case, in-depth understanding will bring us a big edge in forecasting over some simpler heuristics. I certainly hope that deeper understanding help, I think that data which I dug from archive.org made many things clearer, but we will see after resolution if this was worth anything.
Files
Files
Tip: Mention someone by typing @username