This post is heavily inspired by a post by Noah Smith^[1] and a lot of the evidence comes from the Sentinel Newsletter, which I highly recommend.

Previously, I've argued that:

The Ukraine war is not only a proxy war between the West and Russia: China is aiding Russia's war effort. And so, we're fighting a hot proxy war with China^[2] (similarly, the West backs Israel, while China backs Iran, which in turn backs Hamas and Hezbollah).^[3]
Russia^[4] and China^[5] are waging shadow wars on the west.
Increasingly, the Chinese defense industrial base is on wartime footing, especially relative to the US. Also, China tries to become independent in terms of food, fuel, and finances.^[6]

And so, the risk of revisionist great powers launching a World War to overturn the global order is now a lot higher than just a few years ago. Few are talking about it. WWIII might start if and when either Russia or China start a war with NATO.

Russia

Forecasts show a 30% chance that Russia will attack a European NATO member before 2035.^[7] The Sentinel Newsletter has the following evidence and analysis to support this: ^[8],^[9]

Russia's 2025 draft budget allocates a Cold War-high one-third of total spending to the military, driving inflation and necessitating cuts and tax hikes to fund the war.^[10]

Russia is adding new nuclear-tipped air-to-air missiles to its nuclear arsenal^[11] and expanding bases and planning to mass large numbers of troops along its borders with Finland and the Baltic states,^[12] in an indication that it may be preparing for war in the Baltics.^[13] A Russian invasion of one or more Baltic states would challenge NATO and would be a test of whether Article 5 still has teeth. Chinese troops have also held military exercises with Belarus on the Polish border.^[14]

NATO deterrence posture and warning signals: NATO allies are conducting nuclear drills and exercises and senior NATO/European officials have publicly warned about Russia's potential to threaten NATO territory; military planners are rehearsing defensive scenarios beyond Ukraine's borders.^[15] US nukes have been deployed to the UK, for the first time since at least 2008.^[16] There's a 22% chance that the US deploys nukes to Japan or the Philippines before 2035.^[17] Germany has sent troops to Lithuania to help protect NATO's eastern flank.^[18]

China

Forecasters believe that:

there's a ~35% chance that China invades Taiwan in the next 10 years,^[19]
and if that happens a ~60% chance that will the US respond with military force,^[20]
and, conditional on that, a 66% chance China will retaliate.^[21]

The US has been supporting Taiwan more, and withdrawing its pretense that Taiwan is a part of China. In response, China suspended talks about nuclear proliferation and started increasing its stockpile, staged large military exercises, practiced cutting undersea cables, built 'invasion barges', increased its calls for reunification. The US and allies have responded with their own exercises and freedom of navigation exercises, but also by trying to on-shore some chips capabilities.^[22]

Other flashpoints are just as scary and probable, but overlooked (e.g. a new North vs. South Korea war^[23] or a war between China vs Philippines^[24] or India^[25], are all forecasted to ~10% probability).

Sentinel writes:

Philippines: China is clashing with the Philippines about territory in the South China Sea, and the Philippines is able to respond because of the US' support.
Japan: China and Japan clash over the Diaoyu/Senkaku islands, and Japan has been surprisingly military active over the last year.
India: The relationship between China and India has been improving.
Korea: China isn't really keeping North Korea in reins.
Australia: China conducted a naval drill near Australia earlier this year.

More generally, forecasters think there's a 15% chance that, before 2035, there'll be a US-China war.^[26] And one forecaster thinks, before 2050, there's a ~30% chance of such a war involving at least one of the US, China, Russia, and India before 2050; but these forecasts are operationalized as >1,000 battle deaths in a year, but odds of an all-out 'World War III' (e.g., widespread nuclear use), is much less likely.^[27]

But how much less likely? I'm neither a forecaster nor expert, but, looking at the state of the world, in terms of rough orders of magnitude, it seems that we cannot be confident that the chance is less than 1%, which is a very high and scary number.