By on November 19, 2018

autonomous hardware

Apple co-founder Steve Wozniak may no longer work for the company in any official capacity, but he has stayed on as a tech advisor and sounding board. When the Woz says something it usually isn’t without merit, which is why it was interesting to learn he thinks self-driving vehicles aren’t going to happen.

Previously, Apple was said to have hundreds of employees working on an electrified, autonomous vehicle as part of Project Titan. Despite having the necessary testing permits, the company shifted toward developing software for self-driving applications in 2016. CEO Tim Cook confirmed that was the firm’s new focus in 2017 but analysts and industry insiders have continued to claim the Apple Car is still quietly in development. Maybe someone should tell that to Wozniak because he seems to think the entire idea is bogus. 

Confessing that he purchased a Tesla because he initially believed in its autonomous vision, Wozniak told CNBC he gradually became disenfranchised with the idea.

“I wanted to be part of this lead in to autonomous driving,” he explained on Fast Money Halftime Report last week. “I wanted to be a part of that crowd and I kept upgrading my Tesla to one that would have a camera and radar. And then one that would have eight cameras and a radar, because the first one would never do it. And then I gave up and I said it’s really not going to happen.”

While Wozniak praised Tesla’s ability to produce effective electric vehicles and complement its vehicles having the foresight to establish a charging infrastructure beyond city centers, he said he was fed up with the industry lying about autonomous vehicles.

However, he did claim to support advancements in “assistive driving” technology that can allow cars to “spot red lights, and stop signs and avoid some of the accidents today.”

This came with the same warning we like to issue, though. Wozniak said motorists shouldn’t presume driving aids are bulletproof and urged individuals “not to lose sight of the fact you’re not going to get a car that drives itself.”

It may only be one man’s hot take on the issue but Steve is a tech icon and likely someone others will listen to, despite his not being an expert on autonomous systems. It’s also further evidence that many are becoming disillusioned with the concept of self-driving and all of the lofty promises being made by automakers and technology firms. Although, the brunt of Wozniak’s ire seemed to be targeted at Tesla’s AV program.

“Tesla makes so many mistakes,” he said. “It really convinces me that auto piloting and auto steering car driving itself is not going to happen.”

[Image: Ford Motor Company]

Get the latest TTAC e-Newsletter!

56 Comments on “Apple Co-Founder Claims Self-driving Isn’t Realistic, Sick of Lies...”

  • avatar

    Woz is saying something that is obvious. Self driving cars are the trend of the moment and will fade away into history. Ford would be better served upgrading the Fusion and Taurus rather than wasting their time with this frivolous technology.

  • avatar

    Who “REALLY” expects self driving cars everywhere?

    Besides scifi nuts an the uninformed futurists, I don’t think ANYONE in their right mind actually expects self driving cars as we know them today to take over in our lifetimes.

    What I think we DO expect is:
    -Advanced drivers safety aids
    -Situational Autonomous Driving (in built/designed cities, highways, etc.)
    -Advanced public transportation (prebuilt networks/areas)

    I feel like this is as if I went on TV and said, “OK guys… its not realistic for people to become vampires”. No one who knows anything about autonomous cars actually believes we will have autonomous cars.

    • 0 avatar
      FWD Donuts

      I live in Silicon Valley. Been here since 1983. There are scores of people around here who think it’s going to happen. From politicians who approve a ridiculous office park (so what if the roads are clogged already — once cars are automated the throughput’s going way up) to dingleberry techies and punks (man, texting while driving is such a hassle, I can’t wait for cars to be automated).

      Of course, this is one strange little petri dish here — but the attitude seems to be nationwide.

    • 0 avatar

      I do. If idiots, teenagers, and the elderly can do it, a computer can do it better. What you call situation autonomous driving is like 99% of the driving that gets done. Keep a steering wheel for the exceptions but there’s yeah, I think it’ll come. This reminds me of people talking about how the internet would never be a serious threat to retail. It isn’t until it is.

      Woz is cool but Woz is almost 70.

    • 0 avatar

      So, Waymo having driven over 10,000,000 automated miles and preparing to start an uber-like service in metro Phoenix is nothing?
      In 5 years, they will have taken over large and midsized cities and long distance trucking throughout the advanced economies of the world.

      • 0 avatar

        Waymo, you mean the car that gets into regular accidents because it suddenly stops or does weird things so it gets re-eneded. Oh Waymo, you mean the cars that can’t merge into aggressive traffic?

        There is a reason they started in the desert. It’s flat, and there is almost never precipitation. It’s still a hoax. It’s snowing right now where I live, a self driving car couldn’t have made it out of my neighborhood, let alone taken me to work today. I had no issues.

      • 0 avatar

        1/3 of America’s roads- 1.4 million miles worth- aren’t even PAVED.

        I agree they will be in the cities and highways, which is what I said in my post…

        but to have fully autonomous cars, you have to handle the unpaved roads and rural communities that make up a huge proportion of the US.

        until then you’ll have to have cars with self driving FEATURES, but they won’t be autonomous cars. Autonomous cars would have to be able to handle the entire nation.

        I believe we will have “autonomous vehicles” in cities handling taxi/uber type duties in a few years, and that we will have “autonomous” modes to handle highway driving and such… but I think we are an incredibly long way from having actual autonomous vehicles that take you wherever you want to go.

        I live in the suburbs and still 1/4 of houses have gravel or dirt driveways… Are we close to handling those?

        And what about snow, ice, etc? For the 1/3 of the year that 1/3 of the nation is covered in snow and cars cannot see traffic lines or signs, no one knows how to handle that except through V2I… but who is going to pay for 4.2 million miles of V2I?

        and we navigate dirt roads all the time- are we close to handling those?

        GPS units don’t even include half the neighborhoods in the suburbs

        4.2 Million miles of roads in the US, and most people can only perceive a future where about 1/40th of that is transversed autonomously- In the cities and highways where V2I and standardization are reasonable.

      • 0 avatar

        I want to see that Waymo P.O.S. successfully complete just one month’s worth of a Cleveland winter…

  • avatar

    Yyyyyyesss! This also means job security for pilots!

  • avatar

    Woz is likely much more familiar with the technology used in self driving cars than most of this blog’s readers and probably all of its writers. Computer vision and artificial intelligence has been the main focus of the last 6-8 years in computer science. I’m pretty sure he is well aware of the state-of-the-art and the problems to be solved.

    • 0 avatar

      … why do you think that?

      I mean, I’m a professional software developer.

      “Computer science” has no “main focus”, as such – AI and Vision have been *the new hotness* lately, but that’s not the same thing.

      And I see no particular reason to believe Wozniak has any special expertise in either; certainly the work he’s been doing in the past decade doesn’t seem related to either.

      “Famous guy who did a lot of computer stuff and programming” does not translate into “expertise on every computer or software topic”.

      Woz likes to run his mouth. I don’t take him real seriously.

      • 0 avatar

        Agree with Siglvald. I’ve been professionally involved with Computer Science (have an actual scientific degree and everything) as long as Woz. I even studied AI back before most people knew what it really was (most people still don’t understand).

        Does that make either of us an expert on self driving technology? No.

      • 0 avatar

        He’s old now. And perhaps a bit out of touch. But he’s also a genuinely smart guy. In a Valley where genuine smartness have been watered down to almost comical, or tragic, levels; by the past 30 years flood of Federal Reserve welfare.

        In reality, AI today is, at best, no different than it was in the 50s at IBM. Or in the 90s in Japan: Something that “will” happen in a few years….. Smart, even brilliant, young guys plugging away at it as best they can, convinced they can solve the problems this time. Until they too wake up no longer young and smart; but rather old and wise. Like Woz.

        • 0 avatar

          “In reality, AI today is, at best, no different than it was in the 50s at IBM Or in the 90s in Japan”

          That’s totally untrue. Maybe true for the mainstream, but the leading edge of AI has moved beyond the AI you’re familiar with. Collaboration with neuroscientists is pushing the technology in a new direction more closely modeled after the human brain. It’s still in the early stages, but it’s happening. The theory is that evolution probably got it right.

          • 0 avatar
            Art Vandelay

            “It’s still in the early stages, but it’s happening.”

            So, something that will happen in a few years?

          • 0 avatar

            It was totally untrue in the 50s as well. And when Asimo was conceived.

            I’m not saying _nothing_ has changed. Just nothing all that meaningful in the big scheme of things. Processors are faster, sensors cheaper and more plentiful, algos in some cases more refined. And, most importantly, available learning sets much bigger.

            Just like other facets of cars, things have improved a bit. Family sedans are faster now than in the 50s. But that doesn’t make them meaningfully closer to breaking the sound barrier than they were then. Nor to achieve relativistic speeds.

    • 0 avatar

      I graduated with my comp sci degree almost 20 years ago. Neural nets were a thing then too.

      There are technologies that are perpetually 10 years out. AI/Self-driving cars is right there with fusion power – just wait another 10 years.

      When technologists tell me something is 10 years out, that’s code-speak for not going to happen in our lifetimes.

  • avatar

    As Apple stock continues to take a dive the leaders at Apple whine about other companies?

    • 0 avatar

      Wozniak hasn’t worked for Apple since 1985. He is not remotely a “leader” there.

    • 0 avatar

      @VW4, I’d hardly call Woz a “leader of Apple”. Nor do I think he’s too worried about the stock price as his comments were made last week.

      Apple being off 20% YTD today looks like a buying opportunity to me. They’ve made a commitment for some serious share buybacks for the next few years. I could go on about ASP, services, but you can do the work yourself or see if your broker has a pulse. Apple has a completely different storyline when it comes to user privacy vs. Facebook or Alphabet.

      He just made some comments that completely burst the bubble of autonomous driving fanboys from all camps. Last Tuesday if you bothered to read the CNBC article, it sounds like he’s bought a Tesla or 3, and while GM and Ford and especially Tesla pimp this self-driving capability “just around the corner”, it’s not. Maybe that’s why Tesla scrubbed “FSD” (Full Self Driving) from their Q3 shareholders letter and removed the option from their website.

      Seems like there are a few companies that build their share price on autonomous driving, just ask Mary Barra.

  • avatar

    Why do people listen to Woz?

    • 0 avatar

      Because he’s a legend. And because he’s right.

      • 0 avatar

        Not much of a pilot, though. He didn’t single-handedly get the Scotts Valley airport closed, but his f-up was the last nail in the coffin:

        He got his pilot’s license in about 50hrs–an absurdly low time in the Bay Area unless you’re Bob Hoover or Chuck Yaeger (his instructor and check pilot were probably awe-struck)–bought a V-tail Bonanza and crashed on takeoff shortly thereafter. I part-owned a ‘Forked Tail Doctor Killer’ for a couple years; I had over 600 hours, a Commercial Pilot’s ticket, instrument rating, high performance and complex aircraft endorsements, and tail-wheel and aerobatic sign-offs, and it was all I could do to stay ahead of that bird. AFAIK, he never flew his own plane again.

  • avatar

    Self-driving cars are the future! They’re right around the corner!

    And they always will be.

    Honestly it’s more realistic to expect artificially-intelligent androids will begin driving traditional automobiles. If we collectively possessed the technical ability to pull off autonomous driving why would we confine those abilities and the cost associated with it to a seldom-used conveyance machine? Build an all-purpose flexible platform that can slide into the driver’s seat, hop out of the car and deliver/retrieve items, and perform other tasks as required.

    Why would we expect cars to drive themselves but scoff at AI androids?

    • 0 avatar

      Completely agree. If a machine can be created that has the genuine self-awareness, instinct, judgement and self learning capabilities that humans use when driving, building it in the form of a car seems horribly restrictive.

  • avatar

    I disagree with Woz about the value of “assistive driving” technology. Despite his warning not to rely on it, unmotivated drivers will come to grief doing just that. For them, it’s driving that’s the distraction. A perfect example is the Tesla Model 3 driver who killed himself trusting to Tesla’s Autopilot on a piece of road where it had failed before prompting him to register a complaint with Tesla.

    • 0 avatar

      If nothing else, this will be a “jobs program” for tort lawyers and expert witness consultants.

      • 0 avatar

        That’s what they said about airbags. There were a few early cases where airbags got blamed and lawyers got paid, but the sheer numbers of lives saved made them inevitable.

        • 0 avatar

          High-strength steel, crumple zones and airbags were iterative improvements in existing fields.

          Artificial intelligence isn’t really the next iteration of automation – it’s something completely different.

          The biggest problem with AI is predictability. AI will be used in low-risk situations (think predictive text assistants for writing email) for the foreseeable future.

          AI as it currently exists is incredibly difficult to train with reliable results. As such, no one will implement current AI in anything critical that does not have a human backstop.

          AI will assist human efforts for a very long time before it is trustworthy enough to perform critical tasks alone.

  • avatar

    I think Toyota’s head of engineering saying fully automated driving would not happen “in his lifetime” to be more significant. Automated driving is hype used to drive stock prices. I think the bottom is going to drop out of it. There have been countless technologies that were going to be the Next Best Thing that failed to be implemented.

    • 0 avatar

      “Automated driving” is fine, and achievable, as long as the bots are segregated from humans.

      In that/those venue(s), automated driving _will_ happen. Already does in some mining truck and pollution cleanup scenarios. IN a broader context, having to mindlessly sit there and micromanage a vehicle for hours on end down an empty freeway, is a waste of actual, as opposed to “artificial,” intelligent beings’ time, and resources, of the highest order.

      But this is primarily an infrastructure problem. Not primarily an AI one. Which is why it is ultimately solvable. Hence will ultimately be “solved.” People just need to get their priorities straight, and stop falling for paper peddling salesmen flush om Fed welfare.

      • 0 avatar
        Tele Vision

        The only way it will happen is if every vehicle on the road – or that particular stretch of road – is both autonomous AND in complete contact with each other. Otherwise, forget about the Google Pipe Dream.

  • avatar

    There is a lot of misunderstanding with the general public regarding autonomous vehicles and what that implies. Autonomous – “acting independently or having the freedom to do so”.

    There is lots of self driving transportation out there. It’s very likely that the tram you took at the airport is self driving. Large airliners are close to self driving today including landing themselves. If you get a rough landing and there is no obvious weather issues it was probably because the pilot was landing the plane to get his hours in. These forms of self driving transportation benefit from an infrastructure designed around it. The tram has a track and know its location at all times. The plane has multiple systems to keep it where it needs to be. Part of that infrastructure helps insure that there are no other trams or planes in it’s same space.

    We could easily have self driving cars if we were willing to provide the necessary infrastructure to guide them on there journey. This would probably require systems that allow vehicles to communicate with one another as well as guidance systems along the road ways. You would see an orderly progression of vehicles down multi lane roads. You could also probably allow human controlled cars as long as they were broadcasting basic information about where they are on the road, rate of speed, and maybe basic ideas of where they plan to go.

    What the public seems to want are autonomous vehicles that can instead navigate the roadways as they are today much as a human does. I’m sure most of you have read articles about the possibility of bullying a self driving vehicle due to its conservative approach to safety. I will end with that definition again and let you think on that. Autonomous – “acting independently or having the freedom to do so”.

    • 0 avatar

      Not even a system with tracks can necessarily be made both autonomous and safe. The Washington DC Metro has been manually operated ever since a problem with its automatic collision avoidance system led to 10 deaths in June 2009 (on a train line I used to ride routinely) and although many new cars have since been added, no one has said when the system will be restored to automatic operation.

      As for systems that would allow cars to communicate, that would mean retrofitting many millions of conventional cars or waiting until they were all in wrecking yards. So I’m skeptical.

  • avatar

    “It really convinces me that auto piloting and auto steering car driving itself is not going to happen.”

    Someone probably said the same thing about “personal computers” once upon a time. AVs will happen but not tomorrow…or even the next day.

  • avatar

    Cars (or whatever vehicle replaces them) will drive themselves some day. Just not in our litigious society.

  • avatar

    I think self-driving will work brilliantly for shuttle type of downtown buses on semi-dedicated lanes. Otherwise, hmmm not so sure.

  • avatar

    …However, he did claim to support advancements in “assistive driving” technology that can allow cars to “spot red lights, and stop signs and avoid some of the accidents today.”

    This came with the same warning we like to issue, though. Wozniak said motorists shouldn’t presume driving aids are bulletproof and urged individuals “not to lose sight of the fact you’re not going to get a car that drives itself…”


    One thousand times — this.

  • avatar
    Art Vandelay

    I don’t claim to be up on the tech behind the autonomous driving piece of the puzzle, but as one who does work on the security side of vehicle platforms I can say we aren’t even close and that is on vehicles that don’t make driving decisions on their own.

  • avatar

    It doesn’t take a Wozniak to recognize that programming problems of a moral nature are inherent in autonomous vehicles, as exemplified by a study in Science a few years ago:

    • 0 avatar

      Will autonomous cars swerve away from pedestrians, choosing to hit hard, inanimate objects instead of living things? Follow-up question: if they do then can wait in between concrete planters, jump out in front of them at the last second, and cause them to crash into the planters? Mwuahahaha!!

  • avatar

    Yup, control system and firmware guy in the valley too. Just my 2 cents and then I’ll go away:

    1) Woz’s word is likely twisted a bit by others compare to what he meant to say. He likely said something along the line of “perfect” self driving will never happen, but the industry will just roll it out as long as the self driving is better than a new driver just come off of driver’s ed and pass his road test. That I’ll agree.

    2) Woz hasn’t been doing anything new since the 80s. He, like Linus, has the name recognition but hasn’t been hands on in the trench for a long time. I would not listen to anyone who isn’t in the trench for a while on these “what is and isn’t possible” just because he or she was famous.

    3) Like any war time tech advancement, you will get some result if you put a lot of resource (how’s throwing 300-500k pay check to get an engineer from a competitor to join sound) in it, and have an arm race in it, and are willing to let people die to test it in real life (road test). We are almost there, just missing the letting people die part, or have the media campaign that said it is still saving M-N lives compare to drunk driving or typical new driver caused fatalities.

    4) As more people die, we’ll likely found solution to our problems. Road sign may change, transponders will be introduced for human cars, pedestrians, dogs, etc. Road work will have a database so that driverless car will know where are they. Every car will upload road condition to Google and then broadcast to nearby driverless cars to notify something that “may be there”, etc. We’ve been through the horses and buggies to automotive transition, we’ll make it eventually.

    • 0 avatar

      You’ve hit the nail on the head – AI powered cars will probably never be adaptable enough to work on roads and infrastructure designed for humans.

      Hell – many humans can’t seem to navigate roads built specifically for them.

      I think Google finally realized this and came up with “Sidewalk Labs”:

      Now this is just a WAG – Google figured out that AI powered cards in today’s cities is damn near impossible to get right – so instead of beating their heads against this wall – they decided to start changing cities to be more autonomous car friendly.

      It might be a good thing in the long run – but we’ll all be long dead before all cities are built this way.

    • 0 avatar
      Art Vandelay

      Up until recently Linus was still fairly hands on. Moreso than Woz anyway.

  • avatar

    Google has Google Lens. It does a pretty good job of defining objects.

    How about installing a camera facing the driver. If the driver has a dog in their lap, the car stops. Over the last 4 days I’ve encountered 2 drivers with their fur babies in their laps that makes texting while driving look tame.

    The goobers dog is looking out the front window, then jumps to look out the side window, then back to the front window. The driver keeps in sync with brakes, gas and turning the steering wheel.

    I know one thing, if they crashed in front of me and no one was around but me and it was because of the dog, I’d just drive on.

  • avatar

    Every so often, we get into one of those AI/Technology bubbles. We just witnessed peak AI bubble several months ago. Now, all that stuff is crashing. AI, Socials, everything tech. Self driving cars were just part of the AI bubble hype.

    Hope everyone was smart enough to dump those homes in Silicon Valley several months ago … otherwise you are going to lose your shirt … no one wants to live in the bay area garbage can. It will be at least 10 years before the next generation gets sucked into the next AI bubble.

  • avatar

    Self-driving cars will not completely displace drivers.

    AI will not completely replace human thought.

    Automated Teller Machines will not completely replace bank tellers.

    Humans are a very flexible and adaptable species. Betting on these technologies to replace humans is a stupid bet.

    All of these technologies will assist humans in performing their activities to a higher standard, but they will not replace humans.

  • avatar

    I read a 1905 article on the possible future of mans flight and questioning who might accomplish it first. This seems similar.

  • avatar

    The self driving car will never fly with the public. Lets just stop reporting on something that won’t happen.

  • avatar
    R Henry

    The first AVs will be heavy trucks…likely trundling east-west along I10. A human driver will take the loaded truck/trailer to a holding yard on one end, it will then be queued onto a dedicated AV lane…and set off on its own. It will the exit into a holding yard some distance later…and then be human piloted the last mile to final destination. There will be no AV to human-piloted vehicle cohabitation at any time.

    This will be step one, and I reckon it is unlikely to occur in my lifetime.

  • avatar

    Problem is the mix with “normal” cars, imagine rush hour in NYC, you stand in a red light, light turn green and the car in front of you is not moving, what a computer programer tell our car to do in this situation?
    Or, I’m standing in front of a self driving car, I’m not moving, I decided to block it from moving, what would it do? and what about snow driving? you can’t see the lane markings, so we all stay home?

  • avatar

    Kind of mute point isn‘t it? I‘m assuming none of the commentators believe in climate change? The math sys that at least 05% of the cars on the road need to be off the road within 15 years. Really doesn‘t matter if its internal combustion or electric. The emissions goal for 1.5°C is 0 tons/person (budget of around 400 Gt CO2), same as the goal for 2°C (budget around 690Gt CO2). Have yet to see a bicycle made with 0 tons emissions, let alone a car.

  • avatar

    Well, I quite agree with Steve, though. These cars won’t come to real life, not until maybe 10 years from now, until these automakers could make the cars 99% safe for anyone inside and outside the cars. I have just read an article that discussed about this at I think this might be useful for anyone who seek for information about these cars, especially the legal aspects of them.

  • avatar

    The mass produced autonomous car will never happen. So lets stop talking about it.

    Case closed.

Read all comments

Back to TopLeave a Reply

You must be logged in to post a comment.

Recent Comments

  • EBFlex: “Another thing is that I sneak up on deer all the time on my mountain bike.” No….you...
  • EBFlex: ORV is just off road vehicle. A more broad term than ATV or UTV. And again, those are not analogous. Those...
  • Kenn: When I walked by the open door of the GM’s office at a SoCal Toyota dealer, the day I took delivery of my...
  • slavuta: Before traveling to space he could take care of public transport. You should like this...
  • ToolGuy: I spend that $169/year on washer fluid and oil filters instead.

New Car Research

Get a Free Dealer Quote

Who We Are

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