World's First AI Chatbot, ELIZA, Resurrected After 60 Years (livescience.com) 37
"Scientists have just resurrected 'ELIZA,' the world's first chatbot, from long-lost computer code," reports LiveScience, "and it still works extremely well." (Click in the vintage black-and-green rectangle for a blinking-cursor prompt...)
Using dusty printouts from MIT archives, these "software archaeologists" discovered defunct code that had been lost for 60 years and brought it back to life. ELIZA was developed in the 1960s by MIT professor Joseph Weizenbaum and named for Eliza Doolittle, the protagonist of the play "Pygmalion," who was taught how to speak like an aristocratic British woman.
As a language model that the user could interact with, ELIZA had a significant impact on today's artificial intelligence (AI), the researchers wrote in a paper posted to the preprint database arXiv Sunday (Jan. 12). The "DOCTOR" script written for ELIZA was programmed to respond to questions as a psychotherapist would. For example, ELIZA would say, "Please tell me your problem." If the user input "Men are all alike," the program would respond, "In what way."
Weizenbaum wrote ELIZA in a now-defunct programming language he invented, called Michigan Algorithm Decoder Symmetric List Processor (MAD-SLIP), but it was almost immediately copied into the language Lisp. With the advent of the early internet, the Lisp version of ELIZA went viral, and the original version became obsolete. Experts thought the original 420-line ELIZA code was lost until 2021, when study co-author Jeff Shrager, a cognitive scientist at Stanford University, and Myles Crowley, an MIT archivist, found it among Weizenbaum's papers. "I have a particular interest in how early AI pioneers thought," Shrager told Live Science in an email. "Having computer scientists' code is as close to having a record of their thoughts, and as ELIZA was — and remains, for better or for worse — a touchstone of early AI, I want to know what was in his mind...."
Even though it was intended to be a research platform for human-computer communication, "ELIZA was such a novelty at the time that its 'chatbotness' overwhelmed its research purposes," Shrager said.
I just remember that time 23 years ago when someone connected a Perl version of ELIZA to "an AOL Instant Messenger account that has a high rate of 'random' people trying to start conversations" to "put ELIZA in touch with the real world..."
Thanks to long-time Slashdot reader MattSparkes for sharing the news.
As a language model that the user could interact with, ELIZA had a significant impact on today's artificial intelligence (AI), the researchers wrote in a paper posted to the preprint database arXiv Sunday (Jan. 12). The "DOCTOR" script written for ELIZA was programmed to respond to questions as a psychotherapist would. For example, ELIZA would say, "Please tell me your problem." If the user input "Men are all alike," the program would respond, "In what way."
Weizenbaum wrote ELIZA in a now-defunct programming language he invented, called Michigan Algorithm Decoder Symmetric List Processor (MAD-SLIP), but it was almost immediately copied into the language Lisp. With the advent of the early internet, the Lisp version of ELIZA went viral, and the original version became obsolete. Experts thought the original 420-line ELIZA code was lost until 2021, when study co-author Jeff Shrager, a cognitive scientist at Stanford University, and Myles Crowley, an MIT archivist, found it among Weizenbaum's papers. "I have a particular interest in how early AI pioneers thought," Shrager told Live Science in an email. "Having computer scientists' code is as close to having a record of their thoughts, and as ELIZA was — and remains, for better or for worse — a touchstone of early AI, I want to know what was in his mind...."
Even though it was intended to be a research platform for human-computer communication, "ELIZA was such a novelty at the time that its 'chatbotness' overwhelmed its research purposes," Shrager said.
I just remember that time 23 years ago when someone connected a Perl version of ELIZA to "an AOL Instant Messenger account that has a high rate of 'random' people trying to start conversations" to "put ELIZA in touch with the real world..."
Thanks to long-time Slashdot reader MattSparkes for sharing the news.
Nice bit of history but (Score:3)
The Model T is closer to the Airbus A300 than ELIZA is to ChatGPT. Still it wowed a lot of people back in the day.
Re:Nice bit of history but (Score:4)
True dat. Calling Eilza "AI" is really stretching language. Weizenbaum just wanted to show that it's easy to fool people with a very simple program.
Re:Nice bit of history but (Score:4, Interesting)
The term "SI" for simulated intelligence never caught on, but it should have.
Re: (Score:2)
Weizenbaum just wanted to show that it's easy to fool people with a very simple program.
And he did and it still works.
Re: (Score:3)
Calling Eilza "AI" is really stretching language.
In what way?
I think you're confusing 'AI' in reality with 'AI' in science fiction.
Re:Nice bit of history but (Score:5, Insightful)
The definition of AI is shifting. Whenever a computer did something that people would have considered AI before, they searched for something that is missing for it to be a REAL AI. It's called AI effect.
https://en.wikipedia.org/wiki/... [wikipedia.org]
Back then, ELIZA was AI. And astonishing for its testers. Today people slowly start denying that LLM are AI. Probably if we ever would get AGI, people will find reasons why it is not AI. Maybe the whole term is meaningless. Let's rather talk about what the tools actually do instead of trying to define if they are intelligence, when we do not even have a good definition what is intelligence in biological beings.
Re: (Score:2)
What was astonishing to its testers was that the subjects believed they were talking to a real Rogerian therapist when using Doctor.
Re: (Score:2)
Back then, ELIZA was AI.
I think the question is whether that really a good thing to call it. Artificial intelligence implies something that operates like intelligence, or is intelligent but was manmade. But it wasn't either of those things, it was only designed to mock someone with intelligence and fool people with that mockery. It didn't learn (you could add rules to this original version, but you had to do it explicitly, it wasn't automatic) and it could easily be led into a loop that revealed how primitive it was.
ELIZA was call
Re: (Score:2)
I think it was one of the first "I want to be able to talk to the computer" implementation and kinda was AI at the time. Personally I even struggle to see expert systems as AI. Still they are categorized that way. It may help at bit to see that the I in AI may be Intelligence like in CIA and not Intelligence like in brain.
Trs-80 (Score:3)
I was playing around with Eliza back in the early 80's on my TRS-80 computers. It was simple yet somehow fun to play around with. Feeding her false data could result in very amusing outcomes. I suppose these are called hallucinations today.
Re: (Score:2)
I played with it, too. I was able to "stack it" where its response got longer and longer because I added a question about it's response. I got to maybe five levels before it reverted to a generic response.
Maybe, we could have a stacking challenge: longest response with the dialog that got it.
Re: (Score:2)
"She" was on my university's mainframe (Honeywell 66/80, IIRC) in the 1980s, and I had versions of her appear on computer-magazine cover discs several times in the 1990s (before I got either a phone line, a modem or an internet connection). To suggest that "she" was a 1960's program is bullshit - "she" was being compiled and executed at least until the 1990s, and very probably has never been out of commission.
emacs doctor (Score:2)
Re: (Score:2)
Came here to mention Emacs. What's fun is to do M-x psychoanalyze-pinhead
Re: (Score:2)
Since noone mentioned it, the command is
M-x doctor
Script (Score:5, Informative)
A bit of that code....
(HOW DO YOU DO. PLEASE TELL ME YOUR PROBLEM)
START
(SORRY
((0)
(PLEASE DON'T APOLIGIZE)
(APOLOGIES ARE NOT NECESSARY)
(WHAT FEELINGS DO YOU HAVE WHEN YOU APOLOGIZE)
(I'VE TOLD YOU THAT APOLOGIES ARE NOT REQUIRED)))
(DONT = DON'T)
(CANT = CAN'T)
(WONT = WON'T)
(REMEMBER 5
((0 YOU REMEMBER 0)
(DO YOU OFTEN THINK OF 4)
(DOES THINKING OF 4 BRING ANYTHING ELSE TO MIND)
(WHAT ELSE DO YOU REMEMBER)
(WHY DO YOU REMEMBER 4 JUST NOW)
(WHAT IN THE PRESENT SITUATION REMINDS YOU OF 4)
(WHAT IS THE CONNECTION BETWEEN ME AND 4))
((0 DO I REMEMBER 0)
(DID YOU THINK I WOULD FORGET 5)
(WHY DO YOU THINK I SHOULD RECALL 5 NOW)
(WHAT ABOUT 5)
(=WHAT)
(YOU MENTIONED 5))
((0)
(NEWKEY)))
(IF 3
((0 IF 0)
(DO YOU THINK ITS LIKELY THAT 3)
(DO YOU WISH THAT 3)
(WHAT DO YOU THINK ABOUT 3)
(REALLY, 2 3)))
; duplicate line removed: (WHAT DO YOU THINK ABOUT 3) (REALLY, 2 3)))
(DREAMT 4
((0 YOU DREAMT 0)
(REALLY, 4)
(HAVE YOU EVER FANTASIED 4 WHILE YOU WERE AWAKE)
(HAVE YOU DREAMT 4 BEFORE)
(=DREAM)
(NEWKEY)))
(DREAMED = DREAMT 4 (=DREAMT))
(DREAM 3
((0)
(WHAT DOES THAT DREAM SUGGEST TO YOU)
(DO YOU DREAM OFTEN)
(WHAT PERSONS APPEAR IN YOUR DREAMS)
(DON'T YOU BELIEVE THAT DREAM HAS SOMETHING TO DO WITH YOUR PROBLEM)
(NEWKEY)))
(DREAMS = DREAM 3 (=DREAM))
(HOW (=WHAT))
(WHEN (=WHAT))
(ALIKE 10 (=DIT))
(SAME 10 (=DIT))
(CERTAINLY (=YES))
(FEEL DLIST(/BELIEF))
(THINK DLIST(/BELIEF))
(BELIEVE DLIST(/BELIEF))
(WISH DLIST(/BELIEF))
(MEMORY MY
(0 YOUR 0 = LETS DISCUSS FURTHER WHY YOUR 3)
(0 YOUR 0 = EARLIER YOU SAID YOUR 3)
(0 YOUR 0 = BUT YOUR 3)
(0 YOUR 0 = DOES THAT HAVE ANYTHING TO DO WITH THE FACT THAT YOUR 3))
(NONE
((0)
(I AM NOT SURE I UNDE
Re: (Score:2)
I am SO surprised to hear about ELIZA after all these decades!
Nothing new under the sun, that's for sure.
MS-Clippy and Eliza eloped and (Score:4, Funny)
...birthed ChatGPT in a manger in Bethlehem.
At least that's what BingGPT tells me.
In the 70's (Score:1)
Re: (Score:2)
any claim that it was in danger of being lost is a downright lie
singularity is here! (Score:2)
Hey, write me a python version of Elisa chatbot.
Sure, here's an example of a simple chatbot implementation in Python using the Natural Language Toolkit (NLTK) library:
import nltk
from nltk.chat.util import Chat, reflections
# Define the chatbot's responses
pairs = [
[
r"my name is (.*)",
["Hello %1, nice to meet you!"]
],
How was this "lost"? (Score:3, Interesting)
I wonder if anyone is trying to measure the loss of knowledge as older people die. Supposedly books and documentation preserve knowledge for the future, but that doesn't actually seem to be true. People stop caring about some stuff, knowledge of it fades, language fades, attention shifts, and then decades later it is unearthed as though it is some great mystery. It would be interesting to see a study of how much knowledge becomes useless and how fast, and how much knowledge is actually useful but lost because either no one cares about it, or because all the people that knew it are no longer here. Technology such as software seems to disappear relatively fast, infrastructure knowledge (e.g., how to repair elevators) disappears more slowly, specialities in science that only have a few researchers can disappear overnight. One of the problems with scientific papers and documentation is that it does not include information about the systems, tools, context, and infrastructure needed to make use of or replicate what is described. Like old computer science papers assume a particular computer is available. There is a push to provide source code with papers currently, but not all code can run on newer machines or compile in newer compilers. Embedded systems can be very specialized and only exist within a certain time frame. Similar things are happening with infrastructure, few people in the US know how to do big long term engineering projects successfully any more.
Same as it ever was, but in a technological society the consequences happen sooner.
Re: (Score:3, Informative)
"Everyone had a copy of the code, it is in Amiga repositories, it came with some Unix operating systems (I played with it on an SGI workstation), it was everywhere, anyone who wanted a copy had a copy. So how is this "lost" and "resurrected"?"
You didn't read the summary:
"Weizenbaum wrote ELIZA in a now-defunct programming language he invented, called Michigan Algorithm Decoder Symmetric List Processor (MAD-SLIP), but it was almost immediately copied into the language Lisp. With the advent of the early inter
Re: (Score:1)
Re: (Score:2)
The Amiga had "Racter" - the big advance it had was it used the Amiga narrator.device to speak.
Re: How was this "lost"? (Score:2)
But the tools erode our ability to analyze and think deeply
Don't forget AZILE (Score:3)
ELIZA's evil twin [macintoshgarden.org].
these AOL conversations were interesting (Score:1)
Re: (Score:3)
It was good enough to pass the Turing test ... and demonstrate conclusively that Turing test is silly nonsense.
Re: (Score:2)
ELIZA: Tell me more about that.
420 lines of code.. (Score:3)
..and it displays more intelligence than current "AI" instances. :-p
Pygmalion? (Score:1)
Nice. 'Course there's an emacs command to do that (Score:1)
To test this kind of software (LISP edition), in the "Help" menu, "Emacs Psychotherapist", or M-x doctor
(1626 lines of LISP in doctor.el)