Thursday, December 17, 2015

Mobileye vs George Hotz

Bloomberg publishes an article about famous 24-year old hacker George Hotz making a self-driving car in his garage based on 6 cameras and a LIDAR. Hotz says that his technology is much better than one of Mobileye used in Tesla Model S: “It’s absurd,” Hotz says of Mobileye. “They’re a company that’s behind the times, and they have not caught up.

Mobileye spokesman Yonah Lloyd denies that the company’s technology is outdated, “Our code is based on the latest and modern AI techniques using end-to-end deep network algorithms for sensing and control.

SeekingAlpha reports that Tesla comes to Mobileye's rescue. "We think it is extremely unlikely that a single person or even a small company that lacks extensive engineering validation capability will be able to produce an autonomous driving system that can be deployed to production vehicles," says Tesla. "[Such a system] may work as a limited demo on a known stretch of road -- Tesla had such a system two years ago -- but then requires enormous resources to debug over millions of miles of widely differing roads. This is the true problem of autonomy: getting a machine learning system to be 99% correct is relatively easy, but getting it to be 99.9999% correct, which is where it ultimately needs to be, is vastly more difficult. Going forward, we will continue to use the most advanced component technologies, such as Mobileye’s vision chip, in our vehicles. Their part is the best in the world at what it does and that is why we use it."


  1. I do believe talents make things outstanding, while perfect is another thing.

  2. New technology is always started with small group of individuals even a single man. Mobileye has been like this and this great guy will be another starting point of a new tech company which will shadow Mobileye. Maybe this will make Mobileye more humble?

  3. I'm eager to see how things will turn out. Hotz seems to be an incredibly bright fellow, uninterested in petty challenges, who could well give Mobileye a run for their money.

  4. Is Mobileye technology 99.9999% correct with the fatal Tesla Model S crash?

    1. The System 'works' when used correctly. For example, if the Driver though he would be able to leap from the auto-driving car into a chase vehicle, go about his day, then return to the vehicle later then he would discover that does not work.

      It is also not designed for LTAP ( ).

      You need the 'Plus System' for side detection, it's being tested for pedestrians: .

      Reminds me of when "Cruise Control" came out. If you have Cruise Control installed in a Motorhome it doesn't 'work' (let you 'cruise') if you walk to the rear of the Motorhome to use the Washroom or sleep on the Bed.

      What Cruise Control WILL permit you to do (if you have it installed in a Motorhome) is to wander back to the Fridge to get a Beer or mix a Drink. Almost all the time that will be 'OK' (you will live) but it's the occasions where Cruise Control is insufficient that the shortcomings are fully realized.

      Auto-driving Systems (should) constantly determine that they are operating correctly and that they are 'satisfied' with what they 'believe' is occurring - the Driver MUST ALWAYS do the same, be satisfied that they agree with the 'decisions' made by the Computer.

      That's how it works, things were never different (except in the Movie, The Jetsons). Wishful thinking on the drivers part caused that crash. Distance Sensors could have determined that the crossing vehicle was closer than the sky, the color of objects does NOT decide IF you will crash into them, it's _never_ worked that way.

  5. probably, the 0.0001% error rate killed the driver Mr Brown.


All comments are moderated to avoid spam and personal attacks.