Skip to content Skip to navigation

Friday Cyber News, November 4 2016

Cyber technology-related news and links from around the web, for the week of 10/29 - 11/4:

1. This week in the future of work: a backlash against immigration is fueling Singapore's push toward machines in food service. Singapore leads Asia in productivity, and is focusing on replacing human food service labor with machines in part because the sector is one of the less productive sectors within Singapore. Our own Michael Bernstein on understanding long-term crowd worker productivity: worker quality does not decline, despite predictions of remote worker burnout. The consummate sharing economy imagines a smart contract enabling every device to be rented, making you the landlord of all you own. And increased automation in nursing homes improves clinical quality scores. [Bloomberg; Arxiv; Guardian; University of Rochester]

2. George Washington University researchers recommend that the private sector be allowed to counter cyber attacks with "active defense"--distinct from "hacking back", but with a more aggressive sheen than simply managing and mitigating the effects of attacks. [GWU]

3. Shadowbrokers published another dump of NSA hacking tools and configuration files on Monday, along with a list of domains purportedly hacked by NSA. The hacked sites may not have been final targets; some of the tools leaked use compromised sites as staging areas for subsequent attacks. [Ars Technica]

4. Our Crypto Policy project's readout event on Wednesday highlighted the cyclical nature of law enforcement pushback against encryption, as well as the novel techniques developed by researchers to protect information while allowing searches and services to proceed. [Stanford; Govtech]

5. Terms like "machine learning" and "artificial intelligence" assume we're pretty comfortable with the idea that machines think, and that they do so in roughly the same way we do--perhaps with less emotion and better memories. But a look at the history of the application of theory of mind to computing shows that we may be thinking about thinking computers wrong: data itself could be the medium of their minds. [E-Flux]

6. A profile of Cellebrite, the firm suspected of assisting the FBI crack iPhones, and how they got into the reverse-engineering and cyber forensics-for-hire business. [The Intercept]

7. Google AIs evolved their own cryptographic algorithms to protect against a third adversarial AI, and they were able to make changes to the algorithm as the adversary's successes and failures were reported. [Ars Technica]

8. Turkey blocked WhatsApp, Facebook, Twitter, Instagram, and YouTube this week, in part to quash discussion on the platforms of the detention of members of the Kurdish Peoples' Democratic Party. [ZDNet]

9. Using machine learning on State Department cables helps identify aspects of messages more likely to be classified, either to find improperly secured secrets or to verify the legitimacy of leaks. Countering this: the future of privacy is plausible deniability, seeding decoy emails and searches that can be pointed to as purposefully fake in the event of future leaks. [Arxiv.org; The Atlantic]

10. This week in cyber dystopia: A stealth cell tower that records local cell phone activity, cleverly disguised as an office printer. Stanford researchers show that riding Uber while black (or with a black-sounding name) results in longer wait times and more cancelled ride requests. One of the largest insurers in the UK is using Facebook posts to price car insurance for first-time drivers, with use of exclamation points counting against the insured, and use of moderate language like "maybe" counting positively. A mobile ad platform that targets ads based on your feelings. And Facebook lets ad purveyors exclude users based on race (but you can't exclude white people). But don't despair; faith in technological progress is more strongly correlated with life satisfaction than religious faith. [Julian Oliver; Stanford; Guardian; Cluep; ProPublica; Quartz]

Thanks for reading,

Allison
Stanford Cyber Initiative

(To suggest an item for this list, please email aberke@stanford.edu. This email contains links that may not work if HTML is not supported in your mailbox. You can view news from past weeks, subscribe, and unsubscribe at https://tinyletter.com/CyberNewsBytes)