“The reality is that "pink for girls, blue for boys" has existed continuously since at least the 1820s, while "blue for girls, pink for boys" is only recorded between 1889 and 1941. 

A Collection of Anecdotes from the Evolutionary Computation and Artificial Life Research Communities

“We recommend four widely applicable high-impact (i.e. low emissions) actions with the potential to contribute to systemic change and substantially reduce annual personal emissions: having one fewer child (an average for developed countries of 58.6 tonnes CO₂-equivalent (tCO₂e) emission reductions per year), living car-free (2.4 tCO₂e saved per year), avoiding airplane travel (1.6 tCO₂e saved per roundtrip transatlantic flight) and eating a plant-based diet (0.8 tCO₂e saved per year). These actions have much greater potential to reduce emissions than commonly promoted strategies like comprehensive recycling (four times less effective than a plant-based diet) or changing household lightbulbs (eight times less). 

Click the ordinal numbers for the links to the online versions. For a list of board games that you can play against AI opponents (as opposed to other online players), look here.

Some of the biggest online board gaming sites are Boîte à Jeux, Board Game Arena and Tabletopia.

Amazon's cashier-free, Pikachu-tested convenience store is opening to the public.

I've been playing a lot of board games lately. Ars Technica has a nice overview of good, modern ones.

“The newly declassified document, written little more than 24 hours after the massacre by then-British ambassador Sir Alan Donald, gives a much higher death toll than the most commonly used estimates which only go up to about 3,000.

It also provides horrific detail of the massacre, alleging that wounded female students were bayoneted as they begged for their lives, human remains were “hosed down the drains”, and a mother was shot as she tried to go to the aid of her injured three-year-old daughter.

There has never been an undisputed figure for the death toll, but the Chinese Red Cross estimated that 2,700 people had been killed. In 2014, however, it was reported that the Communist regime’s own internal assessment believed 10,454 people had been killed – a figure that would fit Sir Alan’s estimate.

In a one-on-one tournament against Stockfish 8, the reigning computer chess champion, the DeepMind-built system didn't lose a single game, winning or drawing all of the 100 matches played, and it did so without any prior knowledge of the game and after just four hours of self-training.

What’s also remarkable, DeepMind CEO Hassabis explained, is that AlphaZero sometimes makes seemingly crazy sacrifices, like offering up a bishop and queen to exploit a positional advantage that led to victory. Such sacrifices of high-value pieces are normally rare. In another case the program moved its queen to the corner of the board, a very bizarre trick with a surprising positional value. “It’s like chess from another dimension,” Hassabis said.

(It should be noted that Stockfish played without its opening book.)

“If this isn't what you want, please take action at http://autonomousweapons.org/

Some downsides to having super-intelligent robots take over our jobs, even if they don't go haywire and harvest the solar system for raw materials.

“Neural network based classifiers reach near-human performance in many tasks, and they’re used in high risk, real world systems. Yet, these same neural networks are particularly vulnerable to adversarial examples, carefully perturbed inputs that cause targeted misclassification.
(...)
Our work demonstrates that adversarial examples are a significantly larger problem in real world systems than previously thought. Here is a 3D-printed turtle that is classified at every viewpoint as a “rifle” by Google’s InceptionV3 image classifier.

Paul Fenwick comments:

“Imagine hacking electronic billboards so they look like stop signs to cars, or speed signs with *much* faster speeds.

Content-aware fill, now with added context awareness through neural networks.

Remember AlphaGo, the first artificial intelligence to defeat a grandmaster at Go? A new version, AlphaGo Zero, taught itself to play Go from scratch, playing against itself and discovering which strategies worked, without any supervision or use of human data. After just three days, it became good enough to beat the original AlphaGo 100-0.

“The toxic soup of general voice chat is self-selecting. It’s the default state of gamer chat at the most accessible level: Only those willing to share or accept a base level of hateful language remain. Everyone else elects to leave. Can you blame us?

The problem isn’t even the people saying these things; not entirely. The bigger problem is that the systems and general audience for these games are taught to tolerate it. And soon they’re not just tolerating it, they find it normal. And once it’s normal, it’s OK to take part, and anyone who is rightfully upset or opposed to it is accused of upsetting the status quo. 

To exist as a minority online is to have your resilience constantly called into question. As if communicating on social media and wanting to play games with friends and strangers should require deep reservoirs of endurance and strength.

I’m told I need to toughen up when I “complain” about the unending hate speech in online gaming. It’s the internet, what did I expect? We are always being asked to accept racial abuse by people who have no intention of changing the status quo. 

LessWrong has an updated site (with quirky rationality-related content).