Jeremy Hsu writing for Scientific American
So Pluribus instead deployed its depth-limited search, which considers how opponents might choose among only four general betting strategies: the precomputed blueprint, one biased toward folding, another biased toward calling and a fourth biased toward raising. This modified search helps explain why Pluribus’s success in six-player poker required relatively minimal computing resources and memory in comparison with past superhuman achievements in gaming AIs. Specifically, during live poker play, Pluribus ran on a machine with just two central CPUs and 128 gigabytes of memory. “It’s amazing this can be done at all, and second, that it can be done with no [graphics processing units] and no extreme hardware,” Sandholm says. By comparison, DeepMind’s famous AlphaGo program used 1,920 CPUs and 280 GPUs during its 2016 matches against top professional Go player Lee Sedol.
Sean Hollister writing for The Verge
Even if you say “no” to one app when it asks for permission to see those personally identifying bits of data, it might not be enough: a second app with permissions you have approved can share those bits with the other one or leave them in shared storage where another app — potentially even a malicious one — can read it. The two apps might not seem related, but researchers say that because they’re built using the same software development kits (SDK), they can access that data, and there’s evidence that the SDK owners are receiving it. It’s like a kid asking for dessert who gets told “no” by one parent, so they ask the other parent.
Sarah Krouse and Patience Haggin writing for WSJ
Internet providers could learn that someone in a household is using online dating services, or deduce that a person is engaged by the number of visits to wedding-planning websites. By combining web activity with voter-registration data, a provider could theoretically identify households with a 25-year-old on the cusp of needing health insurance, and show those households ads for coverage, say ad executives and privacy advocates. In general, marketers are willing to pay a premium for such hypertargeted ads.
Martin Regg Cohn writing for The Star:
Second, Ford’s intervention caused catastrophic harm to the utility’s takeover of U.S.-based Avista Corp., because local state regulators concluded that the Ontario government was calling the shots rather than Hydro One management. When they overruled the transaction, it triggered a “kill fee” that cost Hydro One about $139 million, on top of Schmidt’s severance
Raymond Zhong writing for The New York Times
The app’s simple design makes the inspection process easy for border officers to carry out. After Fengcai is installed on a phone, the researchers found, it gathers all stored text messages, call records, contacts and calendar entries, as well as information about the device itself. The app also checks the files on the phone against the list of more than 73,000 items.
China's surveillance spreading to non-Chinese citizens is concerning. How long until they start using their position as the world's factory to spy on consumers of those products around the world.
This quote from the Motherboard article - China Is Forcing Tourists to Install Text-Stealing Malware at its Border on this same subject really stuck out to me:
“There is an increasing trend around the world to treat borders as law-free zones where authorities have the right to carry out whatever outrageous form of surveillance they want,” Edin Omanovic, state surveillance program lead at Privacy International, told Vice. “But they’re not: the whole point of basic rights is that you’re entitled to them wherever you are. Western liberal democracies intent on implementing increasingly similar surveillance regimes at the border should look to what China is doing here and consider if this is really the model of security they want to be pursuing.”