• Their electronic - Bluetooth and hosted - security was absolute garbage too.

     
  • It’s a blockchain thing - not entirely stupid, like oh so many are - but it still seems like a very complicated way to do something simple: exchange money for a service. Why would you be less afraid of an algorithm than the company that makes LoRA going out of business?

     
  • Hmmm, really? That’s crypto killing, no?


    An interview with Intel’s director of quantum hardware Jim Clarke reveals that the current technology being used in small scale production could eventually scale to beyond 1000 qubits. Limitations due to expansion and shrinking as a result of temperature fluctuations prevent engineers from simply expanding the number of qubits on a chip.

     
  • Interesting at $USD 130 but Amazon’s Alexa Strategy seems to be a hodgepodge of devices with voice control.

     
  • My multi-thousand MacBook Pro keyboard is utter garbage, constantly failing less than 6 months after buying. Avoid.

     
  • Notes: 1

    Tags: LoRACanadaeleven-x

    Hmmm, a Canada-wide LoRA network.

     
  • Post-Jobs Apple is like a teenage kid who picks up a guitar, and after a week of furious playing finds out it’s hard to be good at, and then stuffs it in the corner of the room.

     

  • Waymo’s shift from technology development to commercialization began in earnest in April 2017, when Waymo announced its early rider program, giving handpicked families a chance to take rides in Waymo’s prototype Chrysler Pacifica minivans. The move allowed Waymo to start thinking about customer-facing issues, like the user interface for self-driving cars and how the car’s driving style affects passenger comfort.

    Then in June, Waymo signed a deal with Avis to manage its self-driving vehicle fleet in the Phoenix area. In October, Waymo announced a public education partnership with a number of Phoenix-area non-profits designed to highlight the benefits of self-driving cars for blind people and the elderly, as well as the potential to save lives by eliminating drunk driving and other driver errors

     
  • Notes: 3

    Tags: ML


    image

    The latest three points on that graph, interestingly show reinforcement learning related projects, applied to games by Deepmind and OpenAI. Particularly AlphaGo Zero and slightly more general AlphaZero take ridiculous amount of compute, but are not applicable in the real world applications because much of that compute is needed to simulate and generate the data these data hungry models need. OK, so we can now train AlexNet in minutes rather than days, but can we train a 1000x bigger AlexNet in days and get qualitatively better results? Apparently not…

     
  • Tags: ML

    There’s some degree of “well, AI is stuff we can’t do” to this:


    But as Pearl sees it, the field of AI got mired in probabilistic associations. These days, headlines tout the latest breakthroughs in machine learning and neural networks. We read about computers that can master ancient games and drive cars. Pearl is underwhelmed. As he sees it, the state of the art in artificial intelligence today is merely a souped-up version of what machines could already do a generation ago: find hidden regularities in a large set of data. “All the impressive achievements of deep learning amount to just curve fitting,” he said recently.

     
  • Notes: 1

    Tags: Z80History


    The Z80 was a contemporary of MOS Technology’s 6502, and like that chip, it stood out not only for its elegant design but also for being dirt cheap (about US $25).

    The Z80 ended up in thousands of products, including the Osborne I (the first portable, or “luggable,” computer), the KayPro II, the Radio Shack TRS-80, and MSX home computers, as well as printers, fax machines, photocopiers, modems, and satellites. Zilog still makes the Z80, which is used in some embedded systems.

     
  • Interesting, but someone marginal. I suspect a lot of WiFi and BLE devices are more vulnerable to attacks at pairing / setup time.

     
  • Notes: 1

    Tags: UberCarsML

    Note that I don’t know if Techcrunch is characterizing this correctly:

    Uber told the NTSB that “emergency braking maneuvers are not enabled while the vehicle is under computer control, to reduce the potential for erratic vehicle behavior,” in other words, to ensure a smooth ride.

    It sounds more to me like they have it disabled because they didn’t want two AIs fighting for control of the car, with potentially weird feedback loops etc.. And then the problem is that the Uber AI didn’t see the pedestrian at all?

    In either case, I kinda like the idea of a dumber AI acting as a backseat driver for pure safety considerations, such as not running the car into a lake or over a pedestrian. 

     


  • IBM has laid off approximately 50 and 70 per cent of staff this week in its Watson Health division, according to inside sources.

    The axe, we’re told, is largely falling on IBMers within companies the IT goliath has taken over in the past few years to augment Watson’s credentials in the health industry. These include medical data biz Truven, which was acquired in 2016 for $2.6bn, medical imaging firm Merge, bought in 2015 for $1bn, and healthcare management business Phytel, also snapped up in 2015.

    Yesterday and today, staff were let go at IBM’s offices in Dallas, Texas, as well as in Ann Arbor, Michigan, Cleveland, Ohio, and Denver, Colorado, in the US, and elsewhere, it is claimed. A spokesperson for Big Blue was not available for comment.

     
  • “Skydio R1 review: It could change drones forever“. 7 minute video of drone that chases you around and dodges obstacles.