|
Patch reliability is unclear. Unless you have an immediate, pressing need to install a specific patch, don't do it. |
Our next topic of paranoia is one that there is more paranoia than there is reality: Being concerned about automobiles being hacked. Sure there are
[See the full post at: Patch Lady – 31 days of Paranoia – Day 26]
Susan Bradley Patch Lady/Prudent patcher
I can appreciate any new technology on the road or at a desk. But for now I’m in the driver’s seat behind the wheel and at the keyboard. That’s my personal preference. Either way, I’m living just a notch above a horse, saddle and compass. The comfort comes when I’m at the controls, yet I realize one day I’ll have to trade that horse and buggy in for a Rastro and a Jetson Starship E model. It’s worth it for the time being to put paranoia in the back seat while I can be sure footed and breathe easy. Taking one mode of transportation at a time here. Win7 is running great thanks to Woody and crew, the Jeep starts when I turn the key and manually shift to first gear. Rough terrain ahead but 4WD takes care of that.
Win7, GroupA/B, Mac OSX, Jeep circa ’90s, everything runs with one human at the controls, a little help from my friends here at Woody’s place and at the local car shop.
MacOS iPadOS and sometimes SOS
This links to an F.B.I. Public Service Announcement on automotive cybersecurity issues, that refers to July 2015 concerns:
FBI warns on risks of car hacking
March 18, 2016
The FBI and the US National Highway Traffic Safety Administration have added their voices to growing concerns about the risk of cars being hacked.
In an advisory note it warns the public to be aware of “cybersecurity threats” related to connected vehicles.
…
The public service announcement laid out the issues and dangers of car hacking.
Read the full article here
I am reminded of these lines of dialog from Walter Bernstein’s screenplay for the 1964 film “Fail-Safe” (faithfully adapted from the 1962 novel of the same name by Eugene Burdick and Harvey Wheeler):
KNAPP: “The more complex an electronic system gets, the more accident-prone it is. Sooner or later, it breaks down… A transistor blows, a condenser burns out. Sometimes they just get tired, like people…”
GROETESCHELE: “But Mr. Knapp overlooks one thing. The machines are supervised by humans. Even if the machine fails, the human being can always correct the mistake.”
KNAPP: “I wish you were right. The fact is the machines work so fast; they are so intricate; the mistakes they make are so subtle that very often a human being can’t know if a machine is lying or telling the truth.”
And that was more than a half-century ago… .
Already have had hacks on some cars and car jackers have certainly grown more technical in recent years. We don’t have many self driving cars on the road and those that have are not providing enough data to sustain what happens when tens of thousands of them are on the road. Nor do we know the attraction all those types of vehicles will be to hackers, or our enemies. Oversight seems way behind the advancements, and just like the FAA and drones we may be moving too fast in pushing this stuff into the public without knowing how a large population of them will affect us. I also think of the Spock logic vs human logic interacting with self driving vehicles. Human’s sometimes act illogical while computers may act differently or logically programmed. This could be a significant road block in introducing self driving vehicles to the road in large numbers. To coexist with illogical human’s may be the biggest hurdle.
In a world of driverless cars so dependent on software, it only requires one well-timed (say rush hour) release of a hack to cause hundreds, thousands, or millions of fatal accidents all at once.
They’ll be prying the steering wheel out of my cold, dead hands.
Fortran, C++, R, Python, Java, Matlab, HTML, CSS, etc.... coding is fun!
A weatherman that can code
Slight edit needed in the article, hook > hood?
We are getting to the point where the internet is becoming an essential service. It is one of many utilities that business, government and the consumer considers a necessity. Internet connectivity is in all new private and commercial vehicles, to provide some sort of IOT convenience. We are being told that AI is the future for all vehicles, so like it or not, it’s coming. Hackers are attracted to anything that society considers essential. There is no better place to sow chaos or extort a ransom.
Serious hacking is mostly done by state actors. They are more likely to hack the utility itself. Organised crime is the next on the list when it comes to hacking on a large scale. They are more likely to hit businesses or a city government. I would hazard a guess that they need only threaten to hack in order to get what they want. Eventually, but sooner rather than later, school buses, city buses, commercial vehicles and even big rigs will have some sophisticated AI presence. No human at the wheel.
The obvious advise I’d give to vendors is to stop installing IOT/AI in any and all vehicles until product wide security standards are agreed to and made mandatory. Standards driven by a private company or consortium of like minded companies is not acceptable. They can evade security measures that their own products can not meet. They will also put profits before security.
Serious hacking is mostly done by state actors. … Organised crime is the next on the list when it comes to hacking on a large scale.
Of course, the only real difference between a state and organized crime is that the state has declared itself legitimate, and most try to at least appear so.
I am convinced that hacking vehicles in operation is already happening on occasion. If you have the resources, an “accident” would be a clean way to get rid of someone who threatens your organization.
Roads, highways, cities and towns were not designed for driver-less vehicles. It is not just a matter of planning a new infrastructure to accommodate the changes, if that is happening at all, but also a matter of training the human to adjust to their new role – a safety switch! If things get out of hand they are supposed to spring into action and take control of the vehicle. Unfortunately, a hacked vehicle in motion could be totally void of all manual control, e.g. brakes, steering, ignition.
A driver-less big rig seems out of the question, but I have seen video of it. Hacking a big rig would be like weaponizing it. These big rigs will have a human in the cab, but I question how alert that person will be having nothing to do for hours on end. Highway scenery is not stimulating.
This AI test gets discussed a lot …
-Given the binary choice of hitting one person or a group of people (and not being able to swerve or stop on time), what will the AI’s programming tell it to do?
There is always the assumption that, when things get suddenly too dangerous and quick maneuvering could avoid them becoming outright accidents, there will be a driver sitting in front of the car controls that will immediately spring into efficacious action the moment that this occasional driver sees the danger, or hears a warning tone, and will do the right maneuvering, starting from a totally relaxed and distracted condition after a long time on the road without any need for taking the slightest action on his or her part.
How likely is that? You just try it (on a stationary car or on a very slowly moving one, somewhere where there is no traffic around…)
Ex-Windows user (Win. 98, XP, 7); since mid-2017 using also macOS. Presently on Monterey 12.15 & sometimes running also Linux (Mint).
MacBook Pro circa mid-2015, 15" display, with 16GB 1600 GHz DDR3 RAM, 1 TB SSD, a Haswell architecture Intel CPU with 4 Cores and 8 Threads model i7-4870HQ @ 2.50GHz.
Intel Iris Pro GPU with Built-in Bus, VRAM 1.5 GB, Display 2880 x 1800 Retina, 24-Bit color.
macOS Monterey; browsers: Waterfox "Current", Vivaldi and (now and then) Chrome; security apps. Intego AV
Therefore the proponents of driver-less vehicles have made a grossly incorrect assumption.
As far as that question on AI programming, I think the insurance companies would insist the vehicle hit one person rather than several, even if that one person is a child or a pregnant woman (cheaper payout – it is all about the money).
FOB relay hack attack
https://m.youtube.com/watch?v=D_3lgxMwrWI%5D
Edit to remove HTML. Please use the “Text” tab in the entry box when you copy/paste.
This is the kind of stuff that really frightens me and I can’t really understand why we’d want to make computers do everything for us. What happens when the tech fails? Will we even know what to do anymore? I don’t think Tech should replace humans, but assist us. Eventually being forced to hand full control of my vehicle over to some computer is unacceptable to me. A road full of AI controlled cars is a disaster waiting to happen. I think there will be way more driving fatalities than there are now especially accidents involving large numbers of cars and fatalities.
Hacking is just one way this would happen and, IMO, makes the people in this country less secure than they’ve ever been because then, you don’t need guns or missiles. All you need is a computer and a bit of hacker know how to kill hundreds of people. Thousands, millions? Just hack in and tell the cars to floor it and then steer hard to one side at 70mph or something. All of a sudden, a regular guy with a computer can be a murderer without even leaving their house or getting dressed. I’d say that puts people at FAR more risk than ever. Not only that, but it makes humans dumber VIA laziness because we can’t be bothered to drive our own cars anymore.
Having a small amount of those cars on the road is okay. I actually think driverless cars would be great for people who are either disabled or too old to drive safely anymore as it allows those people to maintain a degree of independence in their lives without feeling guilty for burdening others. For the average Joe that is perfectly capable of driving themselves? No.. I can’t go along with that.
If people are really concerned about safety with human drivers, then perhaps we should look at our own driver’s education system. It’s a joke. Drive around a parking lot the size of a mini golf course through some cones and answer a handful of multiple choice questions and you get a permit? Then, drive around with an instructor a few times and tolerate him/her pushing the passenger side brake pedal while you’re driving, play along and watch some educational videos and you get your license? Seriously? We wonder why we have so many bad drivers on the road? We taught them nothing and just threw them out there to learn the rest on their own.
I think driving school should last a year and drivers should gain experience driving in all weather conditions that they will encounter where they live. Rain, snow, ice, fog and all that as well as learning counter steering skills and evasive maneuvers to avoid accidents as well as just basic skills. They should be taught how to change a tire, check their oil and do basic maintenance. Driving school in this country is a joke, so let’s try taking that more seriously before we declare that humans can’t do better and turn even more of our lives over to tech.
I honestly don’t trust the tech world enough to give it that much control. I hope it never happens because someone will abuse it and you’ll have enough people who will float that “well it has to be this way, so no use fighting it” line and enough people will buy it. I just don’t think we’re ready for this. I would much rather humans get better at things with the assistance of technology instead of throwing our arms up and saying “I give up!” and turning it all over to AI. That’s what driverless cars are to me.
If you like self-driving cars you’ll love self-piloting aircraft…..especially Airbus pax liners where the computers “protect the aircraft” from the pilot’s “mistakes”.
Here’s just one example where a pilot found himself in a “knife fight with the computer”. He finally won.
When ‘psycho’ automation left this pilot powerless
Having worked on “mission critical” systems in my career, I know that there are ALWAYS limits to what an engineer can anticipate when designing a system. The more complex the system, the more opportunity for blind spots. It’s a fact of life.
At every level one must try to keep in mind, “fail safe”. I’m here to tell you, that’s impossible. We engineers are not gods.
Without putting too much thought into it you might imagine that “keeping an airplane in its flight envelope by pulling the nose down” or “pulling over to the side of the road” might be a sensible design. And mostly it is.
In HiFlyer’s linked video, a sensor (or group of sensors) failed. Clearly in that case the computer system’s designers didn’t anticipate that such a combination of failures could occur. There was no code in place that said, “just milliseconds ago the plane was flying nominally and level, and now the nose is suddenly ‘up 50 degrees’, so disengage and tell the pilots they need to take over. Why not? Because it’s virtually IMPOSSIBLE to anticipate every possibility, especially self-failure of the system itself. It’s a bit like suddenly developing dementia, while in a position to control people’s lives. Ever meet a demented person who knew or was willing to admit they had lost their facilities? All the ones I know (some in the family) think there’s nothing wrong with them.
But let’s say your self-driving car actually DOES have code that says, essentially, “Whoa, suddenly I’m not getting enough input from the real world, so I need to stop!” Would you think it prudent to just stop suddenly in the middle of a freeway? Or pull over to the side of a road with cliffs on both sides? How well tested do you think such code is going to be?
Take it from a career software / quality engineer: We are not NEARLY to where we ought to be considering putting our very lives in the hands of technology. Note that I did not say “Technology is not NEARLY…” – I said “We”. Digital systems can actually achieve perfection, and make good decisions based on good data far faster and more consistently than people. The problem lies with the fact that we humans are building this stuff, and we are tremendously imperfect.
We might be influenced by someone saying things like “Get this done by the end of the week or you’re fired” or thoughts like “I’m starting that new job on Monday, so **** this” or even “I don’t feel good today.” In my experience EVERYONE is influenced to do less than perfect work virtually ALL THE TIME by SOMETHING or SOMEONE.
-Noel
Without putting too much thought into it you might imagine that “keeping an airplane in its flight envelope by pulling the nose down” or “pulling over to the side of the road” might be a sensible design. And mostly it is.
Aircraft analogies are fitting when it comes to autonomous cars, since aviation is currently the home of the most sophisticated vehicular AI in actual service use.
I’m not a pilot, but I am an aviation enthusiast, and I’ve watched nearly every aviation video I can find, including the full “Mayday/Air Crash Investigation” series. There have been more than a few times that an onboard computer tried to outthink the pilot with disastrous effect. Air Inter Flight 148 on January 20, 1992 springs to mind. The Airbus A320 hit a mountain on approach to Strasbourg, killing most on board.
The pilot had inadvertently selected a descent rate of 3300 ft/min when he intended to select 3.3 degrees, since he’d forgotten to set the mode first, and in both cases the display on the autopilot appeared to read “33” at a glance (the decimal point and the unit description are easy to miss), and that resulted in an exceptionally rapid descent. Even so, simulations showed the aircraft missing the mountain even at 3300 ft/min, which would have put the runway in sight of the aircraft. It turns out that there was a tiny bit of turbulence that caused the aircraft to bounce up a little in altitude right at the moment the 3300 ft/min descent was selected. The computer read this as a sign that there was an emergency in progress, and it took it upon itself to descend at an even faster rate than the pilot had commanded. Had the plane done what it was told, the plane would nearly certainly have missed the mountain, even with the pilot’s mistake. Even if the pilot was aware of this obscure “safety” feature in the plane, he would probably not have been aware it had been triggered by turbulence like he’s felt a thousand times before.
There are two different philosophies in effect at each of the world’s two biggest airliner manufacturers. Boeing believes that automation can assist the pilot, but the pilot is the boss. Airbus believes that since most air crashes can be traced back to pilot error, the aircraft should override the pilot when it thinks he’s making an error. You can probably guess which philosophy I agree with and which I find absolutely terrifying. The computer cannot think or reason; it can only react in deterministic ways according to what it thinks it knows via its sensors. It doesn’t have the intellect to recognize when it is obviously wrong if it hasn’t been programmed for that exact contingency.
I find the push toward driverless cars to be disturbing. It’s an easier sell in some ways than even an Airbus-level of control, given that airline pilots are nearly always consummate professionals while drivers are… well, not, but the complexity of software required to drive a car without having an airliner-style “Oh no, I don’t know how to handle this! Take over, human!” function makes the odds of dangerous bugs far too likely for my comfort.
My car doesn’t get updates. It’s not all mechanical (carb/points/condenser), but changing the program in the ECU (engine control unit) means having to physically attach a ROM chip with the new program in it to the ECU itself. It has no flash capability, and has been running the same old program since the Bush administration (and I don’t mean Dubya).
Even if it is changed with a new chip, all it can change is the idle, the spark advance curve, and the fuel injector duty cycle. It doesn’t control anything else, including the throttle position. I was disturbed by the rise of throttle-by-wire systems, particularly when mated with automatic transmissions. If there is a bug in the system somewhere and the thing decides to “floor it” of its own volition, it can do that! And with “modern” push-button start cars, you can’t easily turn it off in motion; it won’t allow it at speed, because that would be unsafe (lose power steering and such). Most of the time, yes, but if the car has decided that it saw the checkered flag and it’s wide open, I’m turning it off.
During that spate of unintended acceleration episodes with Toyota Priuses, one of the reports was that a panicked driver or passenger in a Prius that was running away had called 911 while it was happening, but could not turn off the car as described above. I would imagine it would have a long-hold force-off mode like a PC, but I’ve never actually had a car like that, so I don’t really know. It’s apparently not something the driver thought to do in the panicked state, and the car went on to crash.
Whatever the cause of those incidents, whether it was truly the floormats or the throttle pedal at fault or something else, the idea of a car that I can’t definitively force off in a split second if I want to would unnerve me. I prefer to also have a positive, always-works way to disconnect engine from transmission at the press of a pedal in addition to that. I don’t inherently trust machines very much– I have too much experience with them for that.
Dell XPS 13/9310, i5-1135G7/16GB, KDE Neon 6.2
XPG Xenia 15, i7-9750H/32GB & GTX1660ti, Kubuntu 24.04
Acer Swift Go 14, i5-1335U/16GB, Kubuntu 24.04 (and Win 11)
My 2001 Honda developed a scary problem last year, the cruise control resume/accelerate button got stuck in the on position, but there was no indication other than unstoppable acceleration. Pulling up on that button did nothing, it was already up. Hitting the brakes did nothing, it resumed and accelerated. Pulling up on the gas peddle was also pointless. Topping that off, the cruise control indicator light burned out. The only solution was to repeatedly hit the cruise control cancel button until hearing a quiet thunk noise which released the resume/accelerate.
I’m lucky I lived through the first time, yet stupidly tried it one more time on a long boring drive and didn’t panic quite as much. I don’t use it anymore, not sure if the cancel button will work next time.
Even cruise control is AI of sorts and I no longer trust that limited function 100%, whether it’s my heavily used car developing a mechanical malfunction or a brand new model. Having no control over a floored gas peddle scared [] me. However I do like anti-lock brakes, could have used those on motorcycles more than once.
Sticking to the subject of AI piloting missiles on public roadways, rather than following your non sequitur misdirections.
Accidents will happen and cannot be eliminated by human reflex nor AI. But a human can be held responsible for the consequences of their own actions or lack of actions. AI makers wish to foist that responsibility onto a human that was not even in control at the already passed moment that could have avoided the impending collision. That is scapegoating at the last moment.
That is a failure that can be avoided by keeping control under the properly liable entity on board, that is a living breathing human. No AI can cause an accident if no AI is given control. That is a zero result, and very easy to attain.
Hey, I’d pay good money for a gadget that would hack and temporarily disable/limp a selected nearby car. It would be great for discouraging illegal left-lane driving and other obnoxious and inconsiderate behaviors.
😉
In my observations, the rude or crazy behaviors almost always happen in ways where the last thing in the world I would want to do to the offending vehicle is to slow or stop it. Far more dangerous than just letting the idiot pass.
-- rc primak
First, let me spell this out up front: I think that both self-driving cars and the Internet of Things are very bad ideas being pushed primarily by people with $$$ signs on their eyes (“imagine that potential huge market out there, and us being the first to go into it and start growing our business from the get go!!!”).
Or, to be more precise, they are really bad for the world of today, with today’s technology. Maybe, several decades from now, things will be different, but where we are today is, in my opinion, not too different from what it was back in the 19th Century with trains. Great idea, until it started killing people standing in the tracks and looking the wrong way (as it actually happened in the UK during the inaugural run of one of the earlier steam locomotives, the “Rocket”), to derailments claiming scores of deaths, railway bridges collapsing because no one had yet the expertise needed to design and, or build them sturdy enough for their purpose. And so, catastrophe after catastrophe, the railway technology has evolved to its (still improving through occasional hard knocks, even if much, much better) safety of today.
Ex-Windows user (Win. 98, XP, 7); since mid-2017 using also macOS. Presently on Monterey 12.15 & sometimes running also Linux (Mint).
MacBook Pro circa mid-2015, 15" display, with 16GB 1600 GHz DDR3 RAM, 1 TB SSD, a Haswell architecture Intel CPU with 4 Cores and 8 Threads model i7-4870HQ @ 2.50GHz.
Intel Iris Pro GPU with Built-in Bus, VRAM 1.5 GB, Display 2880 x 1800 Retina, 24-Bit color.
macOS Monterey; browsers: Waterfox "Current", Vivaldi and (now and then) Chrome; security apps. Intego AV
Two words: Overreaching nonsense.
Maybe cars will shut off and restart in the middle of your commute to install the latest and greatest patches. All for your safety, of course, because you can’t be trusted to do it on your own and also it simplifies their engineering, not having to deal with a variety of patch selections among pesky customers.
[sarcasm OFF]
Better yet, a digital and mechanical key (signature and shape have to match) only provided to qualified repair shop personnel. Although, I would like for these keys to be made available to independent auto mechanics, not just dealership shops. Not perfect protection, but better than OTA updating, even with digital signing.
-- rc primak
PC motherboards used to have a jumper that had to be installed in order for the flash to be possible. I wish they still had this… if OEMs are concerned that businesses won’t like it because it will be too much work to update the firmware, just leave the jumpers in the default “on” position, but still accessible if the case is open. Those of us who want the extra protection can remove the jumper (on a desktop, I would just move it over so it is only on one pin. If it were a laptop, I would be afraid it would be jarred loose, so I would just remove it completely).
Dell XPS 13/9310, i5-1135G7/16GB, KDE Neon 6.2
XPG Xenia 15, i7-9750H/32GB & GTX1660ti, Kubuntu 24.04
Acer Swift Go 14, i5-1335U/16GB, Kubuntu 24.04 (and Win 11)
Re: @anon #227603
“These big rigs will have a human in the cab, but I question how alert that person will be having nothing to do for hours on end.”
NW airlines overflys destination 150 miles – Did not answer FAA for over an hour
https://www.youtube.com/watch?v=7_jBmo_TIgM
Re: @Ascaris #227592
“And with “modern” push-button start cars, you can’t easily turn it off in motion”
Or in the garage apparently.
Re: @OscarCP #227614
“…a driver sitting in front of the car controls that will immediately spring into efficacious action the moment that this occasional driver sees the danger, or hears a warning tone…”
with a paid driver on board for the test.
Those Prius models are supposed to have a “hold until disengage” function. It does not appear that in any of the crashes the drivers did the procedure correctly. That’s either poor documentation or drivers who never RTFM.
That’s true, but under stress, people don’t tend to think clearly and remember something they’ve read in a manual. They revert to habit, and the habit is to press and release to turn off the engine, and this time, it didn’t work. The good old keyswitches are not prone to that… the same action the driver has done a thousand times to make the engine turn off is the correct one to force it off if it ever becomes necessary. There’s no ifs, ands, or buts about cutting all the power to the ECU, the ignition, the fuel pump, and the fuel injectors– the engine will turn off.
Dell XPS 13/9310, i5-1135G7/16GB, KDE Neon 6.2
XPG Xenia 15, i7-9750H/32GB & GTX1660ti, Kubuntu 24.04
Acer Swift Go 14, i5-1335U/16GB, Kubuntu 24.04 (and Win 11)
” Nothing in the current implementation of driverless cars prevents manual takeover or self-maintenance of things like tire pressure and oil changes. ”
Concerning manual takeover: Nothing can prevent it, except the driver’s condition, attitude and readiness: see my longish post further up. This is precisely what is so unrealistic about current expectations of “driverless but with a human ready to take control” cars, trucks, etc. that are expressed so very optimistically and so repeatedly by so many.
Ex-Windows user (Win. 98, XP, 7); since mid-2017 using also macOS. Presently on Monterey 12.15 & sometimes running also Linux (Mint).
MacBook Pro circa mid-2015, 15" display, with 16GB 1600 GHz DDR3 RAM, 1 TB SSD, a Haswell architecture Intel CPU with 4 Cores and 8 Threads model i7-4870HQ @ 2.50GHz.
Intel Iris Pro GPU with Built-in Bus, VRAM 1.5 GB, Display 2880 x 1800 Retina, 24-Bit color.
macOS Monterey; browsers: Waterfox "Current", Vivaldi and (now and then) Chrome; security apps. Intego AV
I am convinced that hacking vehicles in operation is already happening on occasion.
Exactly what the FBI and other government agencies want us to believe. But where’s the proof?
Here, here, here and many others mostly involving hacking into the car’s ECU. It’s fairly common for cars to be operated remotely VIA keychain remotes especially these days. Some cars can even be operated with smart phones and an app. Hacking cars is possible for someone who knows what they are doing. This will only get more sophisticated as hackers are always one step ahead of the patchers and no software can be hack proof.. at least, not that I am aware of.
And yet, in the real world, no crashes on the highway which can be traced to hacked cars. Still nothing but Proof of Concept and news story stunts. All your examples were staged stunts.
A successful hack attack wouldn’t be obvious; you’ll just have a dead driver in a “typical” accident.
If I help you cross the median and meet an oncoming truck, will anyone ever consider that you might not have been attempting suicide?
That danger doesn’t exist with any other IoT device.
Internet of Things also includes remote monitoring of natural gas supply networks in some areas. (Re. the Merrimack Valley/ Columbia Gas, which was doing their servicing under remote control from a center in Columbus, Ohio with no way to shut off excessive pressure from the control center:) Tell us in Mass. that remote control can’t fail in a way which can kill people or leave communities without heat or hot water or cooking gas for months.
You don’t have to be going 70 mph on a highway to be killed by poor implementation of IoT technologies.
-- rc primak
I can foresee a time when the police will have the ability to remotely shut down your car if they are chasing you, in order to avoid a high-speed chase. They would sell the public on the benefits of this sort of technology, because it would be for “public safety”. The problem is, once you have it, you will never be able to get rid of it. And it is possible that this technology will be abused at a future time.
I would even go so far as to say it’s a near certainty that it will be abused and sooner rather than later if/when this gets implemented on a larger scale. Isn’t that how it works in today’s world? After all, the public is on a need-to-know basis and most of the time, we don’t need to know. The public gets told the side of the story that gets everyone on board, but we don’t get told the other side of it. Wasn’t telemetry supposed to be used for improving software and discovering bugs to be fixed? Things aren’t as they seem to be more times than not. I don’t think we are even close to being mature enough as a species for this kind of tech.
Also, bugs on an OS can be fixed or we have backups to restore from to mitigate those issues. If such things happen to your car, it may cost you your life and the lives of others. It’s a huge raise of the stakes. I just don’t have the degree of trust necessary to be anything other than dead set against it.
As another poster above eluded to, if hacking (and crashing) of cars was already happening, we wouldn’t know about it. It most certainly can be done though and it certainly wouldn’t stop at PoC concepts. I’m sure the NSA or someone will develop exploits and it will get leaked to the public or a handful of gifted hackers will develop something. I would not be surprised if they already exist TBH. I do not trust people to do it right and not abuse it.
We’re already living in a surveillance state. It is, IMO, a foregone conclusion that anything that can be used to expand that agenda will be used to do so and having control of anyone’s car remotely (satellites) anywhere at any time would surely be a valuable tool in such an agenda as well as anyone wishing to act maliciously with malware or ransomware. I don’t think it would make us safer, but less safe as people’s lives are directly at stake and our personal freedoms (the ones we have left anyway..) would also be at risk due to the loss of control over one’s own vehicle at any time for better or worse.
Maybe a little overboard, but this just seems to be the way of the world anymore. I would love it if we were a mature enough species to have such powerful technology without abusing it, but we are nowhere near that point right now. Not even in the ballpark, in the country or on the planet. Technology is already used to take advantage of and abuse people. It’s quite frightening to think of upping those stakes. We’re just not ready.. not even close. We use tech for exploitation, making money and war. We already abuse pretty much everything.
FYI
“….It’s estimated that there are approximately 21 million Internet-connected cars on the roads today, and researchers predict that by 2020 this number will grow to 200 million. ”
https://www.automotive-iq.com/autonomous-drive/articles/dangers-connected-cars
You and I are obviously less informed than the vaunted researchers and their crystal balls. Do you believe there will be 9½ times as many currently in use, in only 14 to 26 more months?
I do not. Guess I’ll have to write a note to check back on my prediction.
Donations from Plus members keep this site going. You can identify the people who support AskWoody by the Plus badge on their avatars.
AskWoody Plus members not only get access to all of the contents of this site -- including Susan Bradley's frequently updated Patch Watch listing -- they also receive weekly AskWoody Plus Newsletters (formerly Windows Secrets Newsletter) and AskWoody Plus Alerts, emails when there are important breaking developments.
