petercousins47 has added a photo to the pool:
ROTTERDAM (ANP) - Bij een ongeval tussen een auto en een fietser op het Oostmolenerf in Rotterdam is in de nacht van vrijdag op zaterdag een 40-jarige Rotterdammer overleden. Dat meldt de Rotterdamse politie.
De bestuurder van de auto, een 30-jarige Rotterdammer, is aangehouden. De politie doet momenteel onderzoek naar het ongeluk.
With 40 dead and a disaster declared across all 50 city districts, experts warn extreme flooding will only worsen across south-east Asia
Bangkok is used to absorbing the sudden afternoon downpours that inundate the city during the rainy season. Even streets in many central high-rise districts can briefly turn into rivers, with rats and cockroaches racing out of sewers and taking temporary refuge on trees and buildings as the subterranean world is flushed out.
But every city has a breaking point, and this year rain fell so fast and so hard that whole neighbourhoods of the Thai capital have been transformed, not into rivers, but stagnant lakes. In parts of the city, roughly 300mm of rain – the equivalent of half a year of average rainfall in London – fell in about two days.
Continue reading...Key link between two sides of capital hit again as airstrikes force traffic suspensions on three bridges spanning Dnipro River. What we know on day 1,683
Russia struck a bridge linking two sides of Kyiv for the second successive day, officials said on Friday, part of an intensifying campaign of air attacks on infrastructure in the Ukrainian capital. An evening drone strike on the closed Southern Bridge was the latest in a series of strikes since Thursday that forced the full or partial suspension of traffic across it and two other bridges spanning the Dnipro River. Long traffic jams snaked through the city in the morning and evening. Before the latest attack on the bridge, Kyiv’s mayor, Vitali Klitschko, held a city security council meeting, describing the situation in Kyiv as “very dramatic”.
The strikes on the bridges and other sensitive infrastructure appeared to be linked to an order Vladimir Putin made in September to scrap constraints on Russia’s air campaign, a senior Ukrainian official said. About 500,000 people commuted from the Dnipro’s left bank to the right bank daily, an economic analyst said. Earlier on Friday, one person was killed and two injured in a Russian drone strike that caused a fire on the upper floors of an apartment building, officials said. In the north-eastern city of Kharkiv, Ukraine’s second-biggest, two people were killed and 32 wounded in a guided bomb strike, the regional governor said.
Kyiv is entering its most difficult period since the early days of the invasion, as intensifying Russian attacks reshape everyday life in a capital that Moscow is seeking to plunge into darkness. For weeks, the main worry in the Ukrainian capital had been the coming winter, Pjotr Sauer reports from Kyiv. But a sharp escalation in Russian attacks has left many residents feeling that some of the hardship they were bracing for is already here.
The joint US-Ukrainian reconstruction investment fund has sealed its first critical minerals deal alongside four other new projects, one of its top US officials told Reuters. The fund, established as part of the minerals pact the two countries signed in April 2025 after a push by Donald Trump, is focused on five sectors, from minerals and defence to energy and infrastructure. The fund’s projects now totalled about $70m, said the head of investments at the US International Development Finance Corporation, Conor Coleman. The long-anticipated first minerals deal would inject roughly $30m into a “joint investment platform” with BGV Group, a firm founded by Hennadii Butkevych, the billionaire co-owner of Ukraine’s largest supermarket chain, ATB.
Russia must provide humanitarian assistance to civilians in the Russian-occupied Ukrainian town of Oleshky as the situation there is “appalling”, a European commissioner said. “Civilians are trapped under siege by Russian forces without access to food, medicine, clean water and basic care. This is unacceptable,” said Hadja Lahbib, whose portfolio includes humanitarian aid. A former resident quoted by AFP said there was no flour left to make bread and in the early days people ate pigeons but now even those were difficult to find.
Latvians vote in a parliamentary election on Saturday that is likely to keep power in the hands of its pro-Ukrainian incumbent prime minister, Andris Kulbergs, although deep political fragmentation may complicate the formation of a government. Opinion polls put Kulbergs’s centrist United List ahead, with about 14% of votes, but followed closely by the overtly pro-Russian Sovereign Power.
Continue reading...
:format(webp)/s3/static.nrc.nl/images/gn4/stripped/data154354683-77529f.jpg)
Researchers from Carnegie Mellon, MIT, NYU, and Stanford have built an AI called Ataraxos that finally cracked Stratego, beating four-time world champion Pim Niemeijer 15 games to one with four draws. Its key advantage was a second neural network that estimates the identities of hidden enemy pieces, letting the system search plausible game states rather than brute-force a massive hidden-information space. Ars Technica reports: Just like [DeepMind's DeepNash, introduced in 2022], Ataraxos learned by playing against itself -- 163 million games in total. In these self-play sessions, moves that led to wins were reinforced and played more often in future matches, while moves that led to losses were played less, which was the same simple training idea. The difference was in how much Ataraxos adjusted after each game, because hidden information tends to send self-play learning algorithms around in circles. The team addressed this by making big, bold changes in strategy early in training and small, careful ones later.
The even bigger innovation was something DeepNash never had: thinking ahead before each move. AIs like AlphaGo refine their general strategy with a search just before acting. DeepMind couldn't make that work in Stratego because the search space was too large, leaving it an open question whether it was worth trying. "This is one of the things that we did figure out how to do," Farina said. The solution was a second neural network, a belief model, trained to guess the opponent's hidden pieces based on how they had been moving. This way, instead of iterating through every possible arrangement, Ataraxos samples plausible ones, plays out candidate moves in each, and picks based on how they turned out. And it shows in its playstyle.
[...] DeepNash was trained for two to three months on 1,024 of Google's specialized chips, a run the Ataraxos team estimates would cost $3 million to $4.5 million at 2025 prices. Ataraxos, in contrast, needed 16 GPUs for a week, plus an additional four GPUs for four days to train the belief model. Farina and lead author Samuel Sokota achieved this efficiency by writing a simulator that runs millions of moves per second on graphics cards. "At the scale that we are in academia, we don't really have access to an entire field of GPUs," Farina said. The algorithm also learned far faster -- it played about 34 times fewer games than DeepNash, and still ended up much stronger. The findings have been published in the journal Nature.
The even bigger innovation was something DeepNash never had: thinking ahead before each move. AIs like AlphaGo refine their general strategy with a search just before acting. DeepMind couldn't make that work in Stratego because the search space was too large, leaving it an open question whether it was worth trying. "This is one of the things that we did figure out how to do," Farina said. The solution was a second neural network, a belief model, trained to guess the opponent's hidden pieces based on how they had been moving. This way, instead of iterating through every possible arrangement, Ataraxos samples plausible ones, plays out candidate moves in each, and picks based on how they turned out. And it shows in its playstyle.
[...] DeepNash was trained for two to three months on 1,024 of Google's specialized chips, a run the Ataraxos team estimates would cost $3 million to $4.5 million at 2025 prices. Ataraxos, in contrast, needed 16 GPUs for a week, plus an additional four GPUs for four days to train the belief model. Farina and lead author Samuel Sokota achieved this efficiency by writing a simulator that runs millions of moves per second on graphics cards. "At the scale that we are in academia, we don't really have access to an entire field of GPUs," Farina said. The algorithm also learned far faster -- it played about 34 times fewer games than DeepNash, and still ended up much stronger. The findings have been published in the journal Nature.
Read more of this story at Slashdot.