By on December 15, 2016

Volvo Cars and Uber join forces to develop autonomous driving cars

It’s neither on-message for this site nor terribly interesting to my readers, which is why I rarely mention it, but I have been almost feverishly interested in matters of artificial intelligence, machine consciousness, and advanced language parsing for a very long time.

Thirty-five years ago, I tried to write a very simple sentence parser and response generator for the Atari 800 for my school science fair. The effort failed miserably, in large part because AtariBASIC didn’t really have any usable tools for text handling — and because I was nine years old and had the attention span of a fruit fly.

The night before the science fair, I admitted defeat and decided to do something else: I wrote a quick program that would give pre-programmed responses to certain questions.

The next morning, I demonstrated my program to a couple of nuns. I asked them a couple of leading questions to get them to pick the discussion topic I wanted, then I had them type the questions in. The amazingly intelligent Atari responded in full sentences! Not even the utter pathos of my quickly sketched cardboard sign behind it could keep me from getting an A+. What amuses me, in retrospect, was that the nuns weren’t really all that shocked at the idea that an 8-bit computer could parse language and give reasonable answers. Had I demoed this program to anybody who understood technology, they’d have labeled me a genius or a fraud. But to the nuns, passing the Turing test was about as tough as making an artificial volcano. Blame the movies, I guess.

The problem with every “autonomous” car that has appeared so far is simple: they are all equal to my childhood Atari program. Real autonomous operation is a hugely difficult problem. I’d like to illustrate this for you by listing five rather astounding technical feats that will be easier to accomplish than true vehicular autonomy.

My inspiration for this was simple. Fifteen years or so ago, the comedy Ur-website The Brunching Shuttlecocks did a feature (now disappeared) on “Monkeys, Hamlet, and You.” You’ve probably heard that an infinite number of monkeys banging away on an infinite number of typewriters would almost immediately produce the text of Hamlet, but this article listed a bunch of things that a nearly infinite number of monkeys would produce before they produced Hamlet, including:

  • Several perfect cryptograms of Hamlet.
  • A long Usenet argument over whether Boba Fett is alive, complete with spam-blocked e-mail addresses.
  • The phrase “Jesus Christ my ass is chafed” repeated for the length of two letter-sized pages.
  • The text of Hamlet, except everyone dies of food poisoning in Act II.
  • A brief but accurate write-up of the most embarrassing thing you ever did, with full names, dates, and places.
  • Hop on Pop
  • The Denny’s Kids Menu.
  • A short story entitled “Babysitter’s Passion.”
  • “Iii#jd89 pp98&(*(^9 879j; FF”
  • The text of Hamlet, except that Horatio is named “Elvis.”

Some of you will immediately understand why all of those things would happen first — here’s a hint, “Elvis” is shorter than “Horatio”, and there are twenty-five possible cryptograms of Hamlet — but if you don’t understand, let me try to explain. A nearly infinite number of monkeys that each hit the keys of a typewriter a hundred times would instantly produce virtually all the possible legitimate English sentences of a hundred characters, but they would also produce all sorts of unrecognizable garbage that was 100 characters long. Extend to the length of Hamlet, and there you go.

Now, if you’re trying to create an autonomous car, along the way you’re going to accidentally solve problems that are “shorter” — which is to say, easier — than an autonomous car. I could go on about the thinking behind this statement, but it leads pretty quickly to stuff like “NP-complete” and “isomorphism”, so let’s go directly to the fun stuff. These are Five Amazing Things That Are Easier To Invent Than An Autonomous Car:

Number Zero: A Jumbo Jet Without a Pilot. This one’s cheating, because it effectively already exists. Modern passenger jets are all but autonomous as things sit now. The reason for this is simple: the skies aren’t very crowded and visual interpretation isn’t very important for aircraft. It’s far, far easier to follow a flight plan at 600 mph while dealing with everything from turbulence to incoming bird strikes than it is to, say, read the paint lines on a Chicago freeway during a snowstorm. Long before we can safely let a car drive from Joliet to the Miracle Mile in the winter, we’ll be able to let a 500-passenger jumbo go from jetway to jetway without a human being. Whether we’ll feel good about it is another question.

Number One: The Best Radiation Oncologist in History. I have a lot of respect for radiation oncologists: their job is tough and it involves a lot of life and death decisions. But you could train an expert system to do the job simply by having it “watch” a few thousand oncologists make decisions. Interpreting X-rays is far easier than gauging the available traction of the road ahead. And while you might not trust a robot to tell you if you have cancer, it’s going to be as good as its job as the best available human doctor — better, actually.

Number Two: A Mind-Reading Robot, as Far as You Can Tell. Facial recognition and parsing is tough work, but it’s not as tough as interpreting a Manhattan surface street at rush hour. The day is coming quickly when a robot will be able to watch your face for a short while and learn all the unconscious tics and expressions that betray your inner state, your truthfulness, and your near-term plans for action. And a laser microphone can pick up your subvocalizations pretty easily. The technology to “read” what you unconsciously subvocalize is well underway. Before long, you won’t have any secrets from any machine that can see your face — even through a window, or at a distance.

Number Three: A Completely Automated Fast-Food Restaurant. Imagine walking up to a set of man-height Golden Arches sitting pretty much anywhere in a city, having a pleasant conversation with a very real-looking person on a screen, then receiving your food from a sliding tray beneath that screen. All the technology is already present; it just needs a bit of refining. An underground module about the size of a 10′ x 10′ room could automatically prepare, cook, and package your meal without human intervention. Believe it or not, the biggest obstacle to doing it is customer stupidity. When I was in Barcelona last year, I visited a McDonald’s with no human order-takers — and I saw people repeatedly fail to operate the ordering system properly, even though it was simple enough for a five-year-old. What’s required is natural-language interpretation, which is progressing faster than autonomous-car visual recognition. We’re pretty close to the Idiocracy self-serve restaurant.

Number Four: Almost All the Nightmare Stuff From the Terminator Movies. No, we won’t have “liquid metal” any time soon, at least not like in the movies. But war fighting is a considerably simpler exercise than driving in a city. Consider how much simpler a tank’s job is than a taxi’s job would be: drive over anything that won’t blow up and shoot anything that looks like the enemy. The same is true for drones or bombers: fly past the front lines and kill everything that matches a particular profile. We’re already at the stage of aerial combat where most kills take place past the limits of human vision. So why have the humans? In the war of the future, everybody and everything will carry an IFF (Identification Friend or Foe) device that works on the same principle as the RSA SecureID devices given to cubicle drones. Except the codes will change every second, not every ten minutes, and if you aren’t broadcasting the right code you’ll be marked for death, pronto. Again, I cannot adequately express the difference in complexity between, say, a bomber mission where the plane stays within 100 feet of the terrain at supersonic speed and your daily commute. The latter is much, much more computationally complex.

And that’s our list. But I’d be remiss if I didn’t point out that I deliberately omitted one important thing: long before it will be safe to let computers drive people around, it will be considered safe to let computers drive themselves around. You’ll have everything from pizzas to bank deposits transmitted via autonomous vehicles that are biased towards self-sacrifice. It’s a much simpler problem from a computing perspective if you always assume that your existence is less valuable than the existence of the vehicles and people around you.

The Amazon self-driving motorcycle courier will be programmed to swerve into a wall and not come into contact with a pedestrian or a wayward human-driven automobile. The hardest computing problem of all isn’t visual recognition or traction mapping; it’s ethics. Which leads to the question: Should we be teaching computers to make ethical decisions? And if we should, who among us is qualified to teach?

[Image: Uber]

Get the latest TTAC e-Newsletter!

107 Comments on “No Fixed Abode: Five Amazing Things That Are Easier to Invent than an Autonomous Car...”


  • avatar
    FreedMike

    I’m sorry, Jack. I’m afraid I can’t do that.

  • avatar
    NoDoors

    I’ve seen things you people wouldn’t believe. Attack ships on fire off the shoulder of Orion. I watched C-beams glitter in the dark near the Tannhäuser Gate. All those moments will be lost in time, like tears in rain. Time to die.

  • avatar
    Driver8

    Radiation Oncologist =/= Radiologist

    • 0 avatar
      Lou_BC

      Driver8 – it has been postulated that Radiologists will be replaced by AE. A Radiologist being a Doctor that reads and interprets x-rays, CAT scans and other imaging tests. A Radiation Oncologist might be able to be replaced by AI but I suspect that feat would be as tough as driving a car.

  • avatar
    philadlj

    I avoided 738 accidents on my commute today by not being a robot.

  • avatar
    ajla

    I’m disappointed lovebots didn’t make the list because once those are out I won’t really need a car or 50-hour week job anymore.

  • avatar
    Arthur Dailey

    I have been told but have no confirmation that the human mind can compute the best path to take and speed required to catch a flyball or line drive faster than a computer.

    Is this still correct?

    • 0 avatar
      319583076

      Why would anyone program a computer to catch a flyball or line drive?

      And I don’t believe what you’ve been told has been correct for several decades. Analog fire control computers were developed about a century ago and their modern digital descendants are invariably more accurate than the meat puppets at their controls.

      • 0 avatar
        Arthur Dailey

        Thanks but aren’t fire control programs for hitting targets? The variables in catching a ball include contact point, velocity, trajectory, arc, sun/visibility, wind direction/speed, humidity, proximity of other players, turf conditions and the running ability/speed required by and available to and height of the person having to catch the ball. And these have to be calculated immediately for each ball hit.

        So much more difficult than it immediately looks. And it appears much more difficult than calculating shooting at a moving tank/truck or ship. And there is no heat detection, proximity radar/fuse, etc available to zero in on the approaching ball.

        And the answer to why is generally, because.

        But then I am not a programmer, and relatively technologically challenged so do/did not know the correct answer. The original information was provided by a ‘systems analyst’ with a PhD from a reputable American university but back in the dark ages of the 20th Century.

        • 0 avatar
          Zykotec

          As someone who has played a lot of games against current AI in different war situations (planes, tanks, warships etc.), I can assure you most people cannot beat a decent AI. But some people are still faster and more accurate. Offcourse these situations have been staged in ways that give the AI some advantages when it comes to knowing the environment and physics and it rules out certain random factors that could happen in a real world situation.

          • 0 avatar
            Arthur Dailey

            Got the fire control. I was thinking primarily in terms of artillery and warship fire control firing from a relatively stagnant/slow moving position.

            Yes players do constantly re-calculate. Hence one fielder ‘peeling off’ to allow the other to get to the ball. But they do have to react immediately when the ball comes off the bat, which is perhaps the most difficult of all calculations as they need to take into account where the ball was pitched, the arc of the swing, the length of the swing, the body positioning of the batter and velocity based solely on the sound of the bat striking the ball.

            And many of these issues must be pertinent to what autonomous vehicles have to track/calculate with the exception that they do not have to calculate regarding height/vertical dimension.

        • 0 avatar
          319583076

          Wrong. None of the variables have to be calculated immediately. A baseball player moving on a fly ball is continuously adjusting to the ball. The problem is simply that the player has to put himself within arm’s reach of the ball before it strikes the ground.

          Furthermore, I would argue that several of your variables are superfluous since their effects could be neglected.

          If you haven’t studied differential equations, it probably seems like an intractable problem, but it’s not. If you don’t see the analogy to fire control, I won’t argue that point any further.

          • 0 avatar
            Arthur Dailey

            Not arguing, actually trying to learn. Just like a student asking questions. I know baseball, but do not know computing.

            The variables listed are all relevant as they equate to who will catch the ball (for example 2 fielders have to factor these to decide whose ball it is), how hard they have to run, what path they have to take, etc.

            So, I do appreciate your assistance plus anymore that you can provide to fully explain this.

  • avatar
    rpm1200

    On the internet, nothing ever really disappears.
    https://web.archive.org/web/20150906050400/http://brunching.com/randommonkeys.html

  • avatar
    Moparmann

    Kudoes for the “Idiocracy” reference. The first time I watched it, my head hurt from the thought that people would become so stupid. Since then, I have to come realize that I see instances of gross stupidity, not only in an every day driving enviroment, but in all other facets of daily existence as well!!

  • avatar
    e30gator

    All of the dangerous scenarios involving autonomous cars suddenly go away once you take non-autonomous cars out of the equation. Human error is by far the number one factor in most crashes. That’s why they’re called “accidents.” If non-autonomous cars were relegated to only using certain roads at certain times the “wayward motorist” is no longer a factor and everyone can speedily get to work without traffic congestion and dangerous idiot road hazards.

    • 0 avatar
      rpn453

      “Human error” is such a polite way to say “systematic negligence”!

    • 0 avatar
      cirats

      “All of the dangerous scenarios involving autonomous cars suddenly go away once you take non-autonomous cars out of the equation”????? Really?? What about pedestrians, animals, cyclists, black ice, other road conditions, etc., etc., etc., etc. That’s pretty much the point of Jack’s article – that the multitude of factors that must be considered for autonomous cars to work and be safe is vast.

      • 0 avatar
        e30gator

        With all due respect, Jack is not a developing engineer for an autonomous driving program. And he doesn’t have to be since this is where common sense comes in, which I’m not accusing him of having.

        Luckily, I haven’t witnessed many pedestrians trying to jaywalk across Interstate 75, so it shouldn’t be a problem there. And what happens when a dog or person jumps out in front of my car on a 30 mph residential street? I either try to stop or move over (assuming I’m not drunk or texting someone–which computers, unlike people, don’t usually do)….OR I hit hit them. What changes with autonomous cars? And what happens when a road becomes too dangerous to drive on? As far as I know, they close it. Again, what changes? People crash in to each other all the time on black ice. Are we any better off because of it? Are you trying to tell me that people typically make better decisions than computers? Because if so, I can point to all the evidence to the contrary surrounding you.

        • 0 avatar
          Jack Baruth

          “Luckily, I haven’t witnessed many pedestrians trying to jaywalk across Interstate 75, so it shouldn’t be a problem there.”

          You’ve departed from reality. People and animals cross the freeway all the time. Emergency services personnel, police, construction workers, people walking away from broken-down vehicles, hunters crossing, off-road vehicles breaking the law.

          If you don’t think that autonomous vehicles would be held liable for striking any and all of those, then we really can’t have a conversation.

          • 0 avatar

            Yeah we had two walkers in the interstate deaths in the last 6 months here in CT. It happens. Saw two people walking the Maine Turnpike last week, judging by the backpacks I doubt they broke down.

            A number of workers on autonomous vehicles out side of Cali (MIT RIT etc) have indicated they really don’t have a way to deal with snow and very heavy rain yet.

          • 0 avatar
            e30gator

            Or we can just hold the deer responsible. But in all seriousness, is it my fault if I kill something/someone who jumps out in front of my car when I’m doing nothing illegal, assuming I’m not driving something capable of going from 70 to 0 in under 5 feet?

            This is new technology, which 5 years ago was still SciFi stuff. Hell, if my Roomba can be programmed to avoid barriers, is it beyond the realm of possibility that a multi-billion dollar company could improve upon an existing technology that is clearly in demand from a public tired of driver idiocy? I think not.

          • 0 avatar
            mcs

            One tricky scenario is when a sudden rainstorm floods an underpass on a freeway. Washed out bridges are a problem too. And yeah, I’m an engineer working on autonomous navigation systems.

          • 0 avatar
            Exfordtech

            If you have a dog, and the dog has an accident in the house, then your Roomba will quietly and with great efficiency spread that accident everywhere before you get home. I’ve been in driving situations on major interstates where an onslaught of winter weather makes discerning where the boundaries of the road are, never mind the individual lanes, virtually impossible. I would imagine that will be a significant hurdle to any self driving car.

          • 0 avatar
            Arthur Dailey

            This is a rapidly growing problem in the GTA. On December 6th 2016, 22 pedestrians were struck by vehicles within a 5 hour span. On December 17th, 18 similar incidents occurred.

        • 0 avatar
          e30gator

          @cms

          Aren’t flash floods and washed out bridges a problem for good ol’fashioned humans too? Are you working on a system that can levitate over all hazards and create world peace?

    • 0 avatar
      JimZ

      Danny Butterman: Hey, why can’t we say “accident,” again?
      Nicholas Angel: Because “accident” implies there’s nobody to blame.

    • 0 avatar
      Jagboi

      Isn’t “the removal of non autonomous vehicles” something very close to a train?

    • 0 avatar

      Still programmed by humans as such you have not eliminated the error.

  • avatar
    pwrwrench

    Forget not that the primary motivator in this is money: Making more of it or spending less.
    In the airliner scenario there once was a Flight Engineer. That position was deleted when aircraft systems became more reliable and controls semi or totally automated. Airlines did not have to pay the engineer who was not there.
    There has been discussion to make the Co-Pilot redundant as well.
    They will be replaced with a dog who will work much cheaper.
    As it was mentioned the Pilot’s job is now to watch the computers and make sure nothing goes wrong.
    The dog’s job will be,
    Wait for it.

    To bite the Pilot if he/she tries to touch anything.

  • avatar
    orenwolf

    Great article, Jack.

  • avatar
    Flipper35

    Funny this comes up after having just finished Westworld.

    Jack, those first autonomous war fighting machines will be “Loyal Wingmen” attached to a manned fighter or bomber and given targets. They will be flying in high and stealthy though and not low and supersonic.

    I bet we have self driving cars before “EDI”. (Yes, I threw up a little thinking of that movie). It may not be on technical grounds, but the military would both love to have autonomous fighters and bombers and is loath to not have a man in the middle.

  • avatar
    FerrariLaFerrariFace

    I haven’t even finished reading the article. But I felt compelled to thank you for starting your list at number zero, as any programmer or certified geek properly should.

  • avatar
    Carzzi

    Flight-autopilot avails of the third dimension, hugely increasing the number of collision-avoidance paths, and simplifying the algorithm. Compare this to autonomous cars’ constraints of 2D-only avoidance, coupled with the constriction of streets, further restricting degrees of freedom.

    • 0 avatar
      Flipper35

      Flight computers already fly a proper highway so there isn’t much collision avoidance until you get in the vicinity of an airport where it is hand flown again. Then you have TCAS and computerized ATC and things aren’t really very complicated anymore.

  • avatar
    Zykotec

    I think I could teach a computer ethics, but it would take ages.
    If we just found a way to teach humans ethics we probably wouldn’t have had a need for autonomous cars in the first place. As it is today ethics is one of very few obstacles that are in the way of using current tech to let cars drive themselves in lots of environments. They are already as safe as real people, as long as everyone else drives properly (lol, like if that ever happens)

  • avatar
    RHD

    Okay, Jack, come clean!
    How in the world did you know that Iii#jd89pp98&(*(^9 879j;FF was my Amazon password?!
    Now I will have to change it.

  • avatar
    PrincipalDan

    Tis sad that they failed to calculate how many typewriters the monkeys would destroy in the process.

    They’re not exactly peaceful creatures.

    • 0 avatar
      Lou_BC

      “They’re not exactly peaceful creatures.”

      Better than humans.

      • 0 avatar
        raph

        Are monkey politics really that better? While not a monkey proper I’d say chimps are no better and no worse than humans.

        Especially bonobo chimps it seems who actively murder their competitors which until the behavior was observed was considered a human specialty.

  • avatar
    jmo

    Jack,

    Just for you:

    http://www.nytimes.com/2016/12/14/magazine/the-great-ai-awakening.html

    It goes into how orders of magnitude progress are rapidly being made with AI technology.

    • 0 avatar
      Jack Baruth

      That writer is basically the nun at my science fair.

    • 0 avatar
      JimZ

      well, you have to be careful. In the early days of *any* new technology, progress seems to move at a break-neck speed because you’ve got lots of players all trying different things, with varying levels of success. Then things reach a “shake-out” point where one or maybe two solutions become the way to go, then the *real* drudge work begins. whatever axiom or slogan you want to use (“80/20 rule” or similar) that’s how it almost always goes. examples:

      1) to keep it topical, consider the aforementioned Atari 800. Back in the early days of personal computing and game systems, look at the wide variety of systems and platforms out there. Atari VCS/2600, Atari 400/800, Apple II, Commodore PET, Commodore 64, Colecovision/Adam, the list goes on. Back then you could take a 6502 or Z80 and build yourself a unique computing platform by selecting or designing your own support chips. but the shake-out left Intel the winner with x86, and all of the work since then has been grinding iterations towards minimizing x86s warts for 30 years.

      2) Aviation. the early 20th century saw all sorts of novel designs for aircraft, and progress was fast for a while because the goal was simply to *get off the ground.* the grinding iteration has been towards “get off the ground, stay in the air for hours, get back down to the ground, and do it over and over again for 40-50 years of minimal-fuss service.” And due to that I challenge you to tell a Boeing jetliner apart from an Airbus (save for the obvious 747/A380.)

      • 0 avatar
        brandloyalty

        I heard a line once: “if you want a cheap plane get an Airbus, if you want a safe plane get a Boeing”. Probably not fair, but there are differences in design philosophy that show up in the designs. Boeings are designed to be just as capable in automated functions while being more subject to control by the pilots. And have more critical system redundancy.

        This site has crash rates. But it doesn’t evaluate age of the planes; number of flights; operating conditions; quality of maintenance, pilots and operating protocols.
        http://www.airsafe.com/events/models/rate_mod.htm

  • avatar
    yamahog

    We’re one energy (storage) breakthrough from not needing autonomous cars.

    Cars have to deal with all sorts of bologna, 3rd world drivers, stop lights that don’t work, road signs with contradictory information, markings that may or may not exist, ect.

    If we had better energy storage, we could just use drones to move people places. 20 miles commute to work? Lift off in your front lawn and land near work (on the roof?). For longer trips, we could use electric airplanes / hyperloops / rockets (easy if we can make EM drives better, harder if we have to convert energy to rocket fuel).

    • 0 avatar
      chuckrs

      I’ll put hyperloops in the autonomous cars category. I don’t think you have to go much further than calculating the required minimum radius to change direction at 600-700mph to avoid subjecting the passengers to fighter pilot levels of acceleration. But, if you have an anti-gravity machine generalized to cancel all sorts of acceleration, then you’re good to go.

      Rockets, planes, cars – humans are always the weak link. We are soft and squishy and don’t tolerate acceleration well.

  • avatar
    jmo

    “The problem with every “autonomous” car that has appeared so far is simple: they are all equal to my childhood Atari program.”

    I can’t seem to edit my post. Anyway, how AI works and how your Atari program work have nothing to do with each other. Google uses deep neural networks that don’t need to be programmed to drive, they learn how to drive*.

    * It’s a wild oversimplification but close enough for TTAC.

    http://www.nytimes.com/2016/12/14/magazine/the-great-ai-awakening.html

    • 0 avatar
      Jack Baruth

      Like I said.

      Roger Penrose went through the trouble of writing a whole book to disabuse these people of their notions.

      • 0 avatar
        jmo

        The Emperor’s New Mind? That was almost 30 years ago.

      • 0 avatar
        jmo

        The whole point of the article is how deep neural networks work by avoiding the kind of algorithms Penrose criticized. It doesn’t seem like you’ve read the article if you’re bringing up Penrose’s theories.

        • 0 avatar
          Jack Baruth

          I read the article. I could not be any less impressed if it had been written in crayon. All they are proving is that you can write a better expert system in 2016 than you could have in 2011.

          • 0 avatar
            wumpus

            Reality check:

            DARPA Grand challenge 2004: 150 mile course. “least loser” managed 7 miles before failure.
            DARPA Grand challenge 2005: 5 finishers, most went beyond 7 miles.
            DARPA Grand challenge 2007 (urban course): 6 finishers.

            Today: Autobot failure is *news*. It might not be surprising that a Telsa tried to go under a truck (to anybody but the passenger behind the wheel) or a Uber autobot went through a redlight, but it is *news* and they typically work.

            Hard to say what tomorrow will bring. The hardware is pretty much the same (although improvements of GPUs likely did wonders for improving vision circuits. Don’t expect the same improvements next decade). It appeared that getting to where we are now would be impossible. But you can program your car and it will *usually* work.

            The politics is enough to blow your mind:

            You have the “don’t change anything” folks as usual.
            You have the “who can we sue” lawyers aching to sue everybody when a car crashes, and now they smell new deep pockets.
            You have MADD deathly afraid somebody can leave a bar without the danger of being arrested.

            On the other hand you have a bunch of old people who might not be certain of their driving skills but sure as Hell won’t give up their keys.
            You also have the “soccer moms” and anybody else who has to shuttle around the kids.
            I’m going to assume the “I won’t drive” generation has already opted out of politics. Expect them to show up once they spawn.

            This, I suspect will matter more than getting that last bit making driving safety from 99% (bad driver) to 99.999% (good driver). Current cars might not be ideal for soccer moms (queue “how we rolled in the 1970s picture) but for my father (who tends to drive through red lights) they are probably already good enough. I suspect they are also better than the guy blowing into a BAC straw to start the car.

            Will these cars be able to drive these people? Wait and find out.

          • 0 avatar
            WildcatMatt

            “It might not be surprising that a Telsa tried to go under a truck (to anybody but the passenger behind the wheel) or a Uber autobot went through a redlight, but it is *news* and they typically work.”

            Funny, I would peg Uber as more Decepticon than Autobot.

  • avatar
    Nick_515

    Artificial intelligence was dumb from the get go. Welcome to biomimicry.

  • avatar
    Kendahl

    Doesn’t the Navy already have automated carrier landing technology?

    In response to $15/hour minimum wage laws, fast food restaurants plan to reduce human staff by automating more functions.

    Automation that could get you from the inside of your garage at home all the way to a parking place inside a commercial parking facility is fantastically difficult to implement. On the other hand, we already have systems that can guide a car down a properly marked highway. Therefore, I foresee an intermediate stage where the car handles the easy middle of a trip while the driver remains responsible for the very difficult beginning and end.

    • 0 avatar
      brandloyalty

      My Roomba can dock itself.

    • 0 avatar
      raph

      >>>In response to $15/hour minimum wage laws, fast food restaurants plan to reduce human staff by automating more functions.<<<

      That is such bullshit. Rapid cuisine operations have been planning that well before anybody campaigned for a living wage.

      The same tired "observation" will start making the rounds again when Toyota perfects its robot assemblers and builds the worlds first completely automated plant.

      Machines don't ask for pay raises or more benefits (or form unions), they don't celebrate holidays or take much vacation, they don't get pregnant, they rarely need to take breaks, they rarely argue, they rarely get sick and medical is a helluva lot cheaper than a human making them perfect for endeavors that have a thin margin.

      Human obsolescence in the business world has been an ongoing project for a long time. Eventually any job that is formulaic in function will be done by a machine.

  • avatar
    Arthur Dailey

    Something to prevent me from losing my keys would be most appreciated.

  • avatar
    rentonben

    I certainly agree that Jack is correct. The only clever bits would be to reduce the challenge of driving:

    Paint driving lines with RFID chips in them so AI cars could see lanes through snow.
    Tag all humans with an RFID so AI cars could see small children popping out from between cars before they become visible.
    Retro-rocket braking to stop if the AI gets into trouble – no good going over 8g acceleration though. The meat pockets are fragile.
    Roadside flares with RFID chips to mark off danger.
    Low latency data/visual feed to actual humans to give an AI car commands if there was an unsolvable problem. (Think Mars Rover)

    My only point is that we may not have to solve the current problem – by making the playing field better we could *perhaps* get there from here.

  • avatar
    Prado

    A very thought provoking article. Yet I still believe we will see autonomous cars sometime in the near future AND before all of those ‘easier’ things. Why? Because the value added to an individual who may own/purchase an autonomous car is huge compared to those other tasks. That value added comes in 3 forms. real money, time saved, and reduced stress. How much money would it save on the average ticket price to get rid of the pilot of a jumbo jet? Maybe $5 on a $500 ticket? Woop-de-doo!

  • avatar
    Boff

    “Distronic performs marvelously on a lightly traveled highway but takes on the personality of an amphetamine-addicted cabby in denser traffic. The S600’s silicon cerebellum is rabidly protective of its personal space, so the throttle and brake pedals are in constant tango as they work to keep others at the perimeter of the zone. The incessant braking earns you laser-beam looks from fellow drivers while your passengers scramble for the barf bags. To be perfect, the S600 needs a supercomputer powerful enough to predict traffic’s ebbs and flows and intelligent enough to know when to bend its own rules for the sake of comfort. Oh, wait a minute. It already has one–behind the wheel” ~ Aaron Robinson, commenting on Mercedes’s first implementation of radar cruise control in the May 2001 issue of Car & Driver. Autonomous driving technology has obviously progressed immensely since then, but then again, not that much.

  • avatar
    StarAZ

    Amazon just had its first drone delivery a few days ago. The drone was piloted by a human being, of course.

    Also, Some of the items might not be in the way of the research towards autonomous cars. It’ll be much easier to develop an AI specialized in driving than a general purposed AI, and the oncologist and mind reading robots need to take quite a detour from self driving cars.

    I also expect that we ditch physical money for something more portable and secure (imagine the movie In Time, but you don’t die when you go broke).

    • 0 avatar
      NeilM

      StarAZ writes: “Amazon just had its first drone delivery a few days ago. The drone was piloted by a human being, of course.”

      Nope, that Amazon drone was entirely autonomous. In fact one of the reasons this delivery occurred in the UK was that an autonomous drone was allowed there, but not here in the USA.

  • avatar
    VoGo

    Radiation oncology is being automated, but not in the way you expect, Jack.

    Instead of automating the reading of film, we are now starting to digitize the images of tissue samples. We then run a series of algorithms against these images that analyze serious biology stuff, like the proximity of cell nuclei, which gives a far more accurate understanding of the cancer than any human could.

  • avatar
    DirtRoads

    I’m not the droid you’re looking for.

  • avatar
    Not_a_luddite

    I worked for a couple of years on a DoD autonomous vehicle project and can tell you first hand how terrifying they can be. I was sitting behind the wheel when a 10ton vehicle decided to attempt the Scandinavian flick and the promptly nearly put itself into a unrecoverable rollover situation. I was able, and ready, to take over, and put all four tires back down. To be fair, afterwards, I had to extract my undies from my bum. I’m pretty sure my butthole gasped, and my Calvin Kline’s were sucked in.

    The reality of autonomous vehicles has terrified me ever since.

    I’m no idiot, and I realize that technology will march forward. Processor speed has gotten better, and I know for a fact that the engineers on that program learned some indispensable lessons that day about vehicle control. I also know that in spite of all the research and development done since 2010 on robotic vehicles, we’re still likely a long way off.

    Give me three pedals and leave the dynamic control for the amateurs.

  • avatar
    CGHill

    Ah, yes, the Brunching Shuttlecocks. Never to be forgotten.

    L. Fitzgerald Sjöberg — now answering just to “Lore” — would be worth remembering for his quantum-physics love song to Icelandic singer Björk, and his Trent Reznorized Christmas carols called “Nine Inch Noels.”

    Now I wonder what the heck Sjöberg drives.

  • avatar
    brandloyalty

    Driving in snow can be accomplished the same way people do. You slow down and follow tracks. If there are no tracks you use what you can see to figure out where to go. If conditions are bad enough you pull off and stop. Ford is working on mapping EVERYTHING along the roads to improve on that.

    If any car loses control, even just wheelspin for a moment, this is reported to all other cars in the vicinity. They all adjust accordingly. Potential hazards such as animals are similarly reported when detected. Humans alone on the road cannot do any better and multiple motorists have no practical way to communicate quickly with each other.

    • 0 avatar
      brandloyalty

      My thoughts on this were informed by observing the behavior of the Mobileye aftermarket lane departure warning function on gravel or snow-covered roads.

      While it is remarkably effective in rain at night, it does need lane markings. Without lane lines or defined road edges thete is nothing to trigger the warnings. After enough time without lane markings and it goes into an error state. It recovers by itself when it again sees lane lines. Autonomous systems would easily report these conditions and react accordingly. The idea that they would go speeding blindly into oblivion is silly.

      So how do I stay on the road on gravel or in snow? Obviously by using the information available to recognize the configuration of the road. I don’t see why a computer could not do the same thing.

  • avatar

    The human brain can handle amazingly complex things(well complex for computer anyway) as much as human error is a real lets not pretend making a computer do everything we can do is easy.

  • avatar
    James2

    There are a LOT of drivers in my neck of the woods who could stand to be replaced by a computer. The world would be an infinitely better place. Traffic would actually flow, because there’s no longer a need to figure out what a green light means. The computers won’t be scared of a little rain. And then there wouldn’t be as many crashes that compel –I say COMPEL!– our brain-dead police to shut down the freeway for hours at a time as they lazily ‘investigate’.

  • avatar
    Shortest Circuit

    I would invest a modest amount in a company that would invent a car rental place where I don’t have to interact with people, who type my details into computers slower than my mother “the flash” (she never hits the same keys twice). I imagine an online registration process, and a sort of key vending machine, where I only have to insert my credit card (or touch my phone) and *clink* a key pops out with a text to my phone with the position of the car.

    • 0 avatar
      VoGo

      You’re joking right? All the big rental car agencies have clubs that send you a text of which stall your car is in, and you just get in and go. They check your license at the exit and email you your contract. Easy peasy.

  • avatar
    wumpus

    “The effort failed miserably, in large part because AtariBASIC didn’t really have any usable tools for text handling”

    Actually, the Atari BASIC (or any BASIC) probably was the best tool for string handling at the time (unless you had access to MIT machines and LISP or something). But it wasn’t all that good at letting you build up from that. Subroutines simply added spagetti code and overwrote global variables, so good luck turning string handling into text parsing.

    To be honest, had you typed in “Eliza” from BASIC Computer Games* (and fixed things since ATARI strings needed sizes, but MSBASIC didn’t. I can only imagine the frustration of a 9-year old typing something in “correctly” and have it not work) you would have blown the minds of even those with a little more on the ball.

Read all comments

Recent Comments

  • Master Baiter: ^ This. Climate change is the disaster that’s been 10 years away–for 40 years....
  • mcs: @VWGolfGuy: Yeah, mountians are trick for EVs. You use a lot of power going up, but then gain some back....
  • wolfwagen: If those vehicles are Nissan’s future, they might just want to pull the plug now. Ugly is an...
  • SCE to AUX: “more chargers at work” My office has zero chargers for ~700 workers. We have maybe 4 BEVs in...
  • multicam: Oh my god, it’s so hard to care… when’s the shoe going to drop on this climate change #%[email protected]? I...

New Car Research

Get a Free Dealer Quote

Who We Are

  • Adam Tonge
  • Bozi Tatarevic
  • Corey Lewis
  • Mark Baruth
  • Ronnie Schreiber