RichardJensen Posted February 26, 2016 Share Posted February 26, 2016 yes it can discriminate between limb, bird and or Traffic signal Both a functioning traffic signal and a non-functioning traffic signal? Quote Link to comment Share on other sites More sharing options...
fordmantpw Posted February 26, 2016 Share Posted February 26, 2016 Both a functioning traffic signal and a non-functioning traffic signal? My guess is that they are using some type of radar and can determine the shape of an object based on the radar signals returned. Then, they have a known set of objects in their database that they can use recognition to determine what the object is based on the shape. I'm sure it can also determine the material the object is made of too (wood, metal, etc.) A functioning traffic signal will emit light. A non-functioning one will not, so they can figure that part out based on that. Quote Link to comment Share on other sites More sharing options...
rmc523 Posted February 26, 2016 Share Posted February 26, 2016 That was exactly the situation I was envisioning: Your autonomous car coming to an inexplicable stop because it identifies a tree limb as a defective stop light--- And, conversely, what about situations where the signals are themselves obscured? Or where there is no overhead signal (as in many small towns) where the pole mounted signal in a non-functioning state would be interpreted simply as a light post or tree. Adding to this - what if a light has malfunctioned (stuck on green one direction, red another) where a police officer is directing traffic. I suppose the simple solution would be to have an "override" mode for those situations, where the driver would take control. I agree it will be a long time before we see complete autonomous control 100% of the time. Quote Link to comment Share on other sites More sharing options...
fordmantpw Posted February 26, 2016 Share Posted February 26, 2016 Why don't we all just telecommute? Wouldn't that be cheaper and easier? 1 Quote Link to comment Share on other sites More sharing options...
papilgee4evaeva Posted February 26, 2016 Share Posted February 26, 2016 Why don't we all just telecommute? Wouldn't that be cheaper and easier? All work and no play? That's just what the auto-autonomy lobby WANTS! Quote Link to comment Share on other sites More sharing options...
rmc523 Posted February 26, 2016 Share Posted February 26, 2016 Why don't we all just telecommute? Wouldn't that be cheaper and easier? We should just skip all of this driverless autonomous talk and go straight to teleportation 3 Quote Link to comment Share on other sites More sharing options...
papilgee4evaeva Posted February 26, 2016 Share Posted February 26, 2016 Will there be a Google Traffic Cop® in case of traffic signal breakdowns? If all this traffic and road information is shared via Google, what happens when Google crashes for a few minutes? (as I have known it to do more often than not, sadly, and in the process taking Gmail, Maps, YouTube, etc. along for the ride) We're all talking about Google, but we know Apple is making noise about developing an autonomous car as well. What happens when, citing "security concerns," the Apple Car doesn't want to communicate with the Google Car? Quote Link to comment Share on other sites More sharing options...
RichardJensen Posted February 26, 2016 Share Posted February 26, 2016 (edited) My guess is that they are using some type of radar and can determine the shape of an object based on the radar signals returned. Then, they have a known set of objects in their database that they can use recognition to determine what the object is based on the shape. I'm sure it can also determine the material the object is made of too (wood, metal, etc.) A functioning traffic signal will emit light. A non-functioning one will not, so they can figure that part out based on that. Yeah, it's a combination of machine learning and sensor input. My larger--and persistent--issue with machine learning is that it's an inevitably poor imitation of both consciousness and abstract concepts. Young children often make wonderful and humorous mistakes in the course of developing abstract conceptual skills (a kid I knew growing up bit the first black person she ever saw because she thought she was made of chocolate). This is why we don't let kids drive cars. The problem is that machine learning often fails in a similar fashion. I could go on at length about 20th century continental philosophers who were competent (therefore not deconstructionists), and who transformed the theory of mind in a manner no less revolutionary than Einstein's theory of relativity, but I won't. Unless one of you gives me the slightest of pretexts. -- I might also add that the scenarios most amenable to autonomous driving are almost by definition the safest possible scenarios, which calls into question the ability of partially autonomous vehicles to drastically reduce accidents/injuries/fatalities. Edited February 26, 2016 by RichardJensen 1 Quote Link to comment Share on other sites More sharing options...
rmc523 Posted February 26, 2016 Share Posted February 26, 2016 Will there be a Google Traffic Cop® in case of traffic signal breakdowns? If all this traffic and road information is shared via Google, what happens when Google crashes for a few minutes? (as I have known it to do more often than not, sadly, and in the process taking Gmail, Maps, YouTube, etc. along for the ride) We're all talking about Google, but we know Apple is making noise about developing an autonomous car as well. What happens when, citing "security concerns," the Apple Car doesn't want to communicate with the Google Car? That was another thing I was going to mention - any sort of autonomous "network" would have to either be one single system used on all cars from all brands, or at the minimum some standard would need to be set up for the different systems to communicate with each other. Quote Link to comment Share on other sites More sharing options...
Biker16 Posted February 26, 2016 Share Posted February 26, 2016 Both a functioning traffic signal and a non-functioning traffic signal? yes, using cameras. Quote Link to comment Share on other sites More sharing options...
fuzzymoomoo Posted February 26, 2016 Share Posted February 26, 2016 (edited) I have a hard time putting blind faith into computer programs working correctly.* I really don't see how fully autonomous vehicles will become the standard without some sort of manual override. *Especially if Microsoft has anything to do with it. Edited February 26, 2016 by fuzzymoomoo Quote Link to comment Share on other sites More sharing options...
fordmantpw Posted February 26, 2016 Share Posted February 26, 2016 I have a hard time putting blind faith into computer programs working correctly.* *Especially if Microsoft has anything to do with it. I resemble that remark! Of course, I'm sure akirby will agree with you on the MS part... Quote Link to comment Share on other sites More sharing options...
nicktom Posted February 26, 2016 Share Posted February 26, 2016 I lurk and lurk and never post here but... there are a couple of things that bug me (pun intended) in this discussion: First off, ford has the largest fleet of autonomous cars in the industry. How are they behind the curve? But my real problem is AI security. All computer systems bug and are vulnerable to virus attacks. Even if all the other variables are working wonderfully, what happens when the system shuts down, or malware screws the whole thing up, or do we really believe that nothing can possibly go wrong with Elon & co's super-systems? Quote Link to comment Share on other sites More sharing options...
nicktom Posted February 26, 2016 Share Posted February 26, 2016 Oh, and Richard, I am a post-deconstructionist. What have you got against us? Quote Link to comment Share on other sites More sharing options...
fuzzymoomoo Posted February 26, 2016 Share Posted February 26, 2016 I resemble that remark! Of course, I'm sure akirby will agree with you on the MS part... I've seen screenshots of some of the source code to Windows. The stuff hidden in <!> tags is crazy Quote Link to comment Share on other sites More sharing options...
fordmantpw Posted February 26, 2016 Share Posted February 26, 2016 In all seriousness, Windows is pretty stable, and I have very few issues with any of the MS products I use. I just rebooted my server at home that runs 8-10 VMs including my web server and SQL server after it ran non-stop without being touched for a year. I only rebooted it because they were switching our our electric meter and I was going to be without power for longer than my UPS could support. Software is much more reliable than humans. 1 Quote Link to comment Share on other sites More sharing options...
papilgee4evaeva Posted February 26, 2016 Share Posted February 26, 2016 I saw "Vista" in the code and became REALLY thankful that none of my machines ever ran that abomination of an OS. Same with ME. Quote Link to comment Share on other sites More sharing options...
fuzzymoomoo Posted February 26, 2016 Share Posted February 26, 2016 I saw "Vista" in the code and became REALLY thankful that none of my machines ever ran that abomination of an OS. Same with ME. Personally, I've had more issues with Win7 than I did with Vista. I didn't want to upgrade at all but ProTools 10 doesn't run on Vista very well (wouldn't even install IIRC) and I needed it for school at the time. Quote Link to comment Share on other sites More sharing options...
papilgee4evaeva Posted February 26, 2016 Share Posted February 26, 2016 Personally, I've had more issues with Win7 than I did with Vista. I didn't want to upgrade at all but ProTools 10 doesn't run on Vista very well (wouldn't even install IIRC) and I needed it for school at the time. Which version of Win7? I've noticed quite a large gap in functionality and stability between Pro and Home Premium (same story with 2000/ME, XP, 8, 8.1, and 10) Quote Link to comment Share on other sites More sharing options...
fordmantpw Posted February 26, 2016 Share Posted February 26, 2016 ME was absolutely horrible! Worse than WIn 98 SE and 95. Vista was a big memory hog, but not bad otherwise. XP wasn't too bad, Win 7 was the first really good desktop OS (IMO) from MS. It's pretty stable, and that's what I run on my dev machine at one of my clients. Personally, Win 10 is my favorite by far, and that's what I run on my personal machine. Quote Link to comment Share on other sites More sharing options...
fordmantpw Posted February 26, 2016 Share Posted February 26, 2016 Speaking of, you know why they skipped 9, right? Because 7 ate 9. Seriously, though, it's because there was so much old code checking for '9' as the first number of the version dating back to Win 95/98 that it was really screwing things up. It was easier to change the new version number to 10 than to change that old code. Quote Link to comment Share on other sites More sharing options...
papilgee4evaeva Posted February 26, 2016 Share Posted February 26, 2016 You'd think someone could just do a find-and-replace to get rid of the legacy code that was causing problems... (but what do I know, I left the programming world over 13 years ago ) Quote Link to comment Share on other sites More sharing options...
fuzzymoomoo Posted February 26, 2016 Share Posted February 26, 2016 Which version of Win7? I've noticed quite a large gap in functionality and stability between Pro and Home Premium (same story with 2000/ME, XP, 8, 8.1, and 10) Home Premium 64-bit. Had to go with that version because of what ProTools required. Quote Link to comment Share on other sites More sharing options...
RichardJensen Posted February 26, 2016 Share Posted February 26, 2016 I lurk and lurk and never post here but... there are a couple of things that bug me (pun intended) in this discussion: First off, ford has the largest fleet of autonomous cars in the industry. How are they behind the curve? But my real problem is AI security. All computer systems bug and are vulnerable to virus attacks. Even if all the other variables are working wonderfully, what happens when the system shuts down, or malware screws the whole thing up, or do we really believe that nothing can possibly go wrong with Elon & co's super-systems? There is that as well. The more complex the code, the more points of failure/security weaknesses. You can write a single function program that is practically bulletproof. And then it all goes downhill from there. 2 Quote Link to comment Share on other sites More sharing options...
RichardJensen Posted February 26, 2016 Share Posted February 26, 2016 (edited) Oh, and Richard, I am a post-deconstructionist. What have you got against us? Well at least you're a post-deconstructionist! I have two main issues with deconstructionism, at least as Derrida explained it: The first is the apparent retreat into a Cartesian theory of mind implicit in its attempt to assign meaning to concepts strictly in terms of rational analytics and text (it ignores or minimizes what was one of Heidegger's greatest contributions to philosophy, IMO, the difference between 'ready-to-hand' and 'present-at-hand'.) The second is the metaphysics of presence which is, I believe, based on a fundamental misunderstanding of Heidegger's concept of thrownness. My belief is that if Derrida truly understood Heidegger, he would not have been as obsessed with texts. (and yes, Heidegger is politically problematic, but so is Werner Heisenberg) Edited February 26, 2016 by RichardJensen Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.