In a sense, there was something almost curious about this week's news that Amazon is doubling its smart home aspirations by creating a series of model smart homes in collaboration with client Lennar. calls centers. "The immersive homes are powered by Amazon's artificially-intelligent, voice-activated Alexa assistant, paired with intelligent light bulbs, smart plugs, and all the other features of Black Mirror -inspired living. (Apple already has something Something similar, in collaboration with home builders to build model homes full of HomeKit-enabled devices that can be installed with Siri or from an iPad or iPhone.) Of course, the timing is a little less than ideal, the last few months have been a massive backlash against the tech industry (artfully referred to as "Techlash") after the Facebook Cambridge Analytica scandal.Technical journalists, privacy and legal experts, lawmakers, political experts and other media hub seekers are all on high alert the dystopian possibilities, the AI, intelligent devices and other techn to personalize our personal information. (1
Nicholas Carlini, a fifth year graduate student in computer security at UC Berkeley, who has helped to write the newspaper, says there is no evidence that what he and he are doing his colleagues have learned, was replicated by hackers, but says that it is not a far-fetched possibility. "My guess is that the evil people are already employing people to do what I do," he told the Times. (Amazon says it is taking steps to make sure its voice activation software can not be used by malicious actors.) Echo and Google Assistant both have speech recognition technology that prevents devices from responding to commands as long as they have the voice of the Unrecognize User
Perhaps the most alarming thing is that kidnapping an Amazon echo does not seem to violate existing laws. Our antiquated US Code has not kept pace with devices like smart devices, and it's legal-though "outrageous," according to the Federal Communications Commission-to transmit subliminal messages by machine. Subliminal messages can be considered a violation of privacy, but this concept has not been successfully extended to court machines. At least not yet.
Like other technology companies producing such smart devices, Amazon insists that these language assistants will not be activated unless they hear the "wake-up word" that wakes them from their sleep. Earlier this year, however, users reported that their Amazon Echo devices were plugging themselves in, responding to a disconnected question, or simply silencing them and – most disturbingly – spontaneously breaking out in laughter. "We added the echo points two months ago, the point we got in the master bathroom has now twice randomly played a trail of a woman laughing around 10 am The first time I thought that was the fire television It has been sending audio through the device since I tried to sync it to the TV, but tonight it was completely random, "said a red editor. "No indication in the app that the device heard a command, we had made the point laugh several times, and it was not the laughter Alexa produced, but definitely like a canned laugh, not like someone who laughs live . "
Critics fear that the pace of innovation will advance faster than our predictability of potential problems or safeguards. In fact, voice-based artificial intelligence platforms are still in their infancy. Amazon's Director of Applied Science and Alexa Machine Learning, Ruhi Sarikaya wrote a blog post in April, introducing some of Alexa's new features to make his echoing devices more personalized, including contextual conversations and smarter abilities even memory. "With this ability, Alexa can remember every piece of information for you so you'll never forget," Sarikaya wrote. "Alexa can store any information and retrieve it later." Amazon is also developing a home robot that can act as a mobile Alexa, moving from room to room like a driverless car.
Meanwhile, Google has at C.E.O. Sundar Pichai unveiled a new A.I. Technology on Tuesday, replicating human language, injected "uhs" and "ums" into their responses to approximate the cadence and tics of its users. Some people were not happy. "Google Assistant makes calls that pretend to be human, not just without revealing that it's a bot, but add" uhmm "and" aaah "to deceive the people on the other end of the room. Taunting Tweeted New York Times Writer Zeynep Tufekci. "Silicon Valley is ethically lost, taxless and has learned nothing." Google responded with assurances that the bot would contain disclosure that turns out to be non-human.