Not news, but this seemed the best place to put it

While browsing for some information I came across this page: This academic seems to have no grasp whatsoever on the current state of AI research (and robotics for that matter). It is so disconnected from reality that it is just funny, so I wouldn’t keep it from you.

http://www.nutramed.com/Philosophy/digital_limits.htm

The 1950’s called, and they want their vacuum tube computers back.

I wonder if he still thinks we enter programs by entering punched cards?  Or perhaps by re-wiring the ENAIC.

I have some old computer programming textbooks
that I love to read to study computer science history.
So, in my opinion, this man is not uninformed.

I have books like that too, and like to read them. However, I don’t kid myself in thinking that’s still the current state of technology. Mind you, he published this stuff in 2004, by then almost everything he wrote (especially on robotics) was already largely been invalidated (e.g. the last Asimo model was developed in 2000).

8PLA, don’t get me wrong, my wife says my book called “Bit By Bit : the illustrated history of computers” is my bible

Reading about ENAIC & EDVAC, all very cool.  Did you know the clock rate of ENIAC was 100 KHz ?  Hum, 1/10 th of the clock rate of my Apple II back in the day.  The ENAIC had 17,468 vacuum tubes and was entering a program was done by manually plugging different sections together.

I also like Charles Babbage Difference Engine and Analytic Engine.  I mean we still, to this day, use the word “engine” when we refer to the core of our software systems, in his honor.  In CLUES there are 2 types of variables, ones that come from lookup from disk, and others that are algorithmically generated, I call the the “from disk” ones STORE variables, and the computed ones, MILL variables , because he refered to “the store” as memory, and “the mill” (as we call the CPU now).  That’s the respect I have for that great computer pioneer !

One of my favorite old books:

Understanding Computers

Thomas H. Crowley
Bell Telephone Laboratories
Murray Hill, New Jersey

Copyright (c) 1967 McGraw-Hill, Inc.
Library of Congress Catalog Card Number: 67-14891

Something that bothered me lately:  I’m on a mailing list for electronic music.  Somebody mentioned Watson.  I suggested that there are much more interesting developments coming pretty soon.  I got such a backlash from these people, many of them used to be programmers.  For instance: “I was a programmer and these guys were programming AI in video games, and all they were doing was giving the computer more time to make a decision.” “I worked with AI guys, and they never made any headway and always seemed to get funding, and nothing today implements any AI”  “AI will never happen, they were promising it for 50 years.” “Roger Penrose yadda yadda”  “I saw a computer pass the Turing test using Eliza.  Fooled everybody in the room.”  Yes, there are uninformed dumbasses out there.

I have a confession to make… I still have a Windows 3.1 computer.
It’s a Toshiba portable 386, which is sort of like a cross between a
desktop and a laptop. 

With a bright 256 color, 640x480 folding flat screen with solid glass,
it is bright and dazzlingly beautiful to look at.  The keyboard feels like
you are typing on a piano.

It has ISA bus slots, and there is no battery.  Just a standard PC plug
and an ON/OFF switch for power.  It’s dark gray, so it sort of looks like
a suitcase, the military might have used in bygone years to set off a nuke.

Well, Toby, the AI field certainly hasn’t lived up to the promises of 50 years ago. The problem is, all the fascinating new development has happened under different banners than “AI”. Although visual and audio software, along with text sorting algorithms like those used by search engines, do incorporate AI techniques, that part isn’t focused on due to the stigma that has grown up around AI research. It hasn’t helped that video games have branded their NPC’s as “AI” when they use non-AI tricks to simulate intelligence and learning.

Seems that really the only programs that get called AI are the ones where AI is not central to the application. Take chatbots for example. Most of them reference databases and other knowledge bases that were not generated by the AI itself. What’s intelligent about that?

Toby Graves - Mar 18, 2011:

Something that bothered me lately:  I’m on a mailing list for electronic music.  Somebody mentioned Watson.  I suggested that there are much more interesting developments coming pretty soon.  I got such a backlash from these people, many of them used to be programmers.  For instance: “I was a programmer and these guys were programming AI in video games, and all they were doing was giving the computer more time to make a decision.” “I worked with AI guys, and they never made any headway and always seemed to get funding, and nothing today implements any AI”  “AI will never happen, they were promising it for 50 years.” “Roger Penrose yadda yadda”  “I saw a computer pass the Turing test using Eliza.  Fooled everybody in the room.”  Yes, there are uninformed dumbasses out there.

Ye, AI has been over-hyped for so long, people stopped believing. In the mean time, when you look around, there’s AI (be it simple) all over the place: roomba’s, smart cars,....

Toby Graves - Mar 18, 2011:

AI will never happen, they were promising it for 50 years.” “Roger Penrose yadda yadda”  “I saw a computer pass the Turing test using Eliza.  Fooled everybody in the room.”  Yes, there are uninformed dumbasses out there.

 

I agree with CR; AI has partially succeeded I think.  Narrow AI at least, has certainly been achieved.

dumbasses…. Yes, I think the masses base a lot of their beliefs on what Hollywood tells them.  They think we’re going to go from Zero-to-Strong-AI in no time flat.  Hell, some people think we have AI right now, I worked in a call center and asked “Ok, sir, did you reboot, and what is your computer doing now?”  ... “Well. . .it’s thinking….”  The guess all we have to do is have a O/S boot with startup scripts and that’s called ‘thinking’.

The reason for this is the follow-the-leader mentality.  Some program back in the mid 80’s, I think it was PrintShop, had a screen that, after you input all your desired fonts, color,etc,  it would put a huge message on the screen “THINKING”..  what it was doing was just rendering of course.  But it catches on.

Another thing is the general public has NO CLUE at all how different the quest for strong AI is compared to any other human endeavor. 

My own 2 cents on the development of true AI is, it won’t be someone thinking up one single complex algorithm that will ‘do it all’, it will come one painful step at a time, each functionality building on another, from the ground up.

It would be similar to expecting in 1946 when they completed ENIAC, and expecting Watson within 10 years.  The general public doesn’t see all the steps and milestones that have been achieved so far, because they don’t have knowledge of how computers work, what an algorithm even is, all they know .. is that they don’t see Mr. Data or Sunny yet.

Victor Shulist - Mar 18, 2011:

Narrow AI at least, has certainly been achieved.

Agreed, but it’s EXTREMELY narrow.

Victor Shulist - Mar 18, 2011:

Another thing is the general public has NO CLUE at all how different the quest for strong AI is compared to any other human endeavor.

Totally agreed. And as the website of the Singularity Institute states; real Strong AI will be the last great invention of mankind.

Victor Shulist - Mar 18, 2011:

My own 2 cents on the development of true AI is, it won’t be someone thinking up one single complex algorithm that will ‘do it all’...

Ahum, well, actually…...

C R Hunt - Mar 18, 2011:

Well, Toby, the AI field certainly hasn’t lived up to the promises of 50 years ago. The problem is, all the fascinating new development has happened under different banners than “AI”. Although visual and audio software, along with text sorting algorithms like those used by search engines, do incorporate AI techniques, that part isn’t focused on due to the stigma that has grown up around AI research. It hasn’t helped that video games have branded their NPC’s as “AI” when they use non-AI tricks to simulate intelligence and learning.

Seems that really the only programs that get called AI are the ones where AI is not central to the application. Take chatbots for example. Most of them reference databases and other knowledge bases that were not generated by the AI itself. What’s intelligent about that?

C R,

What do you mean?

A.I. breakthroughs are happening all the time…  http://79b.org/A.I.breakthrough.php

- Just made this breakthrough today especially for you guys :- )

8PLA • NET - Mar 18, 2011:

- Just made this breakthrough today especially for you guys :- )

I like it