On Thu, 4 Dec 2003, [EMAIL PROTECTED] wrote: >Eventually machine intelligence will replace human intelligence >throughout the economy. Wonder if the final outcome will be "good" or >"bad" Productivity will have increased but human interaction (at least in >these traditional areas such as education and probably health care) will >have decreased. > >arthur
I guess this is a good place to relate an experience I had today. I'm currently at CERN, helping to install some pieces of hardware we've cobbled up into the next great accelerator - big science at its most impressive. Anyway, we had this huge piece of hardware held up on supports in the middle of a large workroom, when a couple of girls came in, one with a camera, and one with a laptop under her arm. I thought, perhaps the CERN Courier is going to do another little article on the progress of our project. But instead, these two take out a bunch of little black squares about the size of postit notes, and start climbing up and sticking them all over the construction. I'm not sure if they were adhesive, or like fridge magnets, or both. Each square has a one cm white spot in the centre, but each has a differently segmented white circle around the central dot, at about 3cm diameter. Then they take out a pair of telescoping rods and extend them to about a metre and a half, and clip them to our construction, one horizontally, the other vertically. Each rod also has one of the black patches with white coding, mounted at each end. Then one sets up the laptop, while the other starts taking pictures, walking around the device. While the picture taking is still proceeding, the one with the laptop says, "Would you like to see?" and shows a diagram already appearing on the laptop screen. You see, these girls are the survey team, and they are generating a full 3D map of the device. The camera has a wireless connection to the laptop and is uploading images. The laptop identifies the little targets in the photos and does a brutal quantity of computation in real time among the photographs to deduce the position of the targets based solely on the multiple images and the two reference rods. As the surveyor operating the laptop explained to me (she is now a CERN employee, but used to work with the company which developed the technology) by taking a sufficient number of photographs, with a sufficient number of targets (I'm guessing they used a binary multiple, 32 or 64) it is not even necessary to have a pre-calibrated distortion free lens on the camera. The software can deduce and correct for any aberration in the lens as part of the overall calculation. The accuracy of the process is somewhat limited by the image quality of the digital camera, though it does much better than simple resolution of the camera image - for our gadget, about 6x6x3 metres, they get down to about 1/2 mm. So much for theodolites, and a day's computations, to generate a survey. Well, that's my whizzbang techno story for today... -Pete -----Original Message----- From: Franklin Wayne Poley [mailto:[EMAIL PROTECTED] Sent: Wednesday, December 3, 2003 8:42 PM To: [EMAIL PROTECTED]; [EMAIL PROTECTED] Subject: [Futurework] Future Teaching Have a look at the robotic teacher I'd like to hire from King's College, London: <http://www.geocities.com/machine_psychology/IMP_Cover_Page> ---------- Forwarded message ---------- Date: Wed, 3 Dec 2003 17:35:46 -0800 (PST) From: Franklin Wayne Poley <[EMAIL PROTECTED]> Reply-To: [EMAIL PROTECTED] To: [EMAIL PROTECTED] Cc: [EMAIL PROTECTED], [EMAIL PROTECTED], [EMAIL PROTECTED] Subject: [IMP] Final Lesson 36 There are typically 36 hours of class time for a one semester, 3-credit course. Lessons 31-35 are more in the nature of an assignment: draft out a set of menus and prompts for the SEE-to-C program or even go further and turn that into C code if you are so inclined. How much of my notes on SEE-to-C I will post eventually on the expert system program for C code writing, I do not know. If I am correct about this (and you can find out by trying to write SEE-to-C for yourself) then future students can forget about texts like Aitken and Jones ("Teach Yourself C in 21 Days") or a course like COMP 2425 at BCIT which takes about 144 hours. Gary Livick's C-programmed robot, Etcetera, will be able to teach C in one hour. Final lesson 36 is titled "Godbot" and it is designed to stimulate some creative and metaphysical thinking. If anyone has SPECIFIC criticisms I will welcome them. I certainly don't want to cap off a course which I have spent so much time developing, with any errors. <http://www.geocities.com/machine_psychology/The_Ghost_In_The_Machine> FWP ------------------------ Yahoo! Groups Sponsor ---------------------~--> Buy Ink Cartridges or Refill Kits for your HP, Epson, Canon or Lexmark Printer at MyInks.com. Free s/h on orders $50 or more to the US & Canada. http://www.c1tracking.com/l.asp?cid=5511 http://us.click.yahoo.com/mOAaAA/3exGAA/qnsNAA/pU_rlB/TM ---------------------------------------------------------------------~-> Machine Psychology has to do with machine substitution for the phenomena which are the traditional subject matter of psychology. IMP is being taught according to the slogan of automated teaching: THE STUDENT IS ALWAYS RIGHT! "Courseware" is being designed so that IMP will "stand alone" as a machine teacher without needing any further input from human teachers. <http://www.geocities.com/machine_psychology/Table_of_Lessons> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/ _______________________________________________ Futurework mailing list [EMAIL PROTECTED] http://scribe.uwaterloo.ca/mailman/listinfo/futurework _______________________________________________ Futurework mailing list [EMAIL PROTECTED] http://scribe.uwaterloo.ca/mailman/listinfo/futurework _______________________________________________ Futurework mailing list [EMAIL PROTECTED] http://scribe.uwaterloo.ca/mailman/listinfo/futurework
