Five Betting Markets for 2024 + London Meetup Monday
Bet on Manifold, all the cool kids are doing it...
London Meetup This Monday Evening
Here’s the signup link. Not sure where to host it yet — please put some suggestions in the comments! I’ll also be at Oxford on Thursday hosting something, so stay tuned for more details.
Manifold-ChinaTalk Essay Prize
Over the past year, I’ve found myself spending more and more time on Manifold Markets, the world’s largest play-money prediction market. They’ve generously sponsored this post, where I propose and price five markets on their platform.
By the way, Manifold is partnering with ChinaTalk on a $6,000 essay competition to make a bold prediction on the future of China that’s closing next Sunday. You can write essays making predictions either on any of the markets I created below, or come up with own topic entirely! Click here for more information.
Now, to the markets…
Will a Standing Committee or State Council member get purged in 2024?
In 2023, Xi struck hard into the PLA ranks to clean house in the Rocket Force in particular. With the Li Shangfu purging coupled with the unforgettable “Qin Gone!” saga, two State Councilmembers awkwardly left us in 2023. Will more senior leadership in Beijing lose their plum postings this year?
I set the market at 25%. Bet here.
Will OpenAI, Google, or Anthropic Have a Model’s Weights Hacked before 2025?
ChinaTalk has been deep in coverage of Chinese firms’ attempts to catch up and surpass the western state of the art. In recent weeks, we’ve seen stories of firms like Bytedance violating OpenAI’s terms of service to help train its lagging models, while other firms have been impressing us with their progress. But with China’s long history of industrial espionage, what if firms or the MSS really start to play dirty and get access to the weights of the leading closed AI models?
Dario Amodei said in a Dwarkesh interview said the following:
One of our goals is [to make it cost] more to attack Anthropic than it costs to just train your own model. It doesn’t guarantee things because, of course, you need the talent as well, so you might still [attack]. But attacks have risks, diplomatic costs, and they use up the very sparse resources that nation-state actors might have in order to do the attacks.
We’re not there yet, by the way. But I think we are at a very high standard of security compared to the size of company that we are. If you look at security for most 150-person companies, there’s just no comparison. But could we resist if it was a state actor’s top priority to steal our model weights? No. They would succeed.
My guess was 30%. Bet in this market here.
Will a Taiwanese, US, Filipino, Japanese, or Vietnamese servicemember suffer a casualty in a confrontation with China's military before 2025?
Each week it feels as if we’re reading more and more news stories of dangerous activity on the part of Chinese fighter jets and ships across the straits and in the South China Sea. While in the wake of the San Francisco Summit both Xi and Biden appear committed to turning down the temperature on the relationship, it seems unlikely to me that these sorts of “exercises” will stop in the coming year. Thus far, China has succeeded in getting its point across through these micro-confrontations without causing any casualties.
I’ve set the market at 15%. Bet on this market here.
Will China get mentioned more times than AI over the course of the 2024 presidential debates?
It’s an election year! Presuming Trump actually decides to do debates, both China and AI will be major topics of discussion. But which will get more airtime?
My sense is that it will be easier for a GOP candidate to attack Biden over China than over AI, but 2024 may be such a crazy year for artificial intelligence that moderators will end up asking more questions about it even if there isn’t yet a clear partisan valence on the technology.
I set the market at 60% for China getting mentioned more. Bet on this market here.
Will a SMIC 5-nm chip make it into production in a Huawei device in 2024?
2023 saw Washington freak out over the Huawei Mate 60’s 7nm SMIC-fabbed chip. Rumors back in December of a 5nm, SMIC-powered Huawei laptop were proven unfounded, but many industry analysts expect SMIC to achieve 5nm-level performance with DUV equipment.
Will SMIC do it in 2024? It seems moderately likely given their recent success and continued access to fabbing equipment purchased over the past few years. I set the market at 65%. Bet here.