4 ms·
The Turing test was passed by dr Eliza in the 70es. Revealing not how to model intelligence rather how keen humans are on fooling themselves.
by throwaway4good 3y ago
The Turing test was passed by dr Eliza in the 70es. Revealing not how to model intelligence rather how keen humans are on fooling themselves.
- throwaway4good 3y agoHere is a beautiful 2 minute clip demonstrating this point and Eliza itself: Before Siri and Alexa, there was ELIZA https://www.youtube.com/watch?v=RMK9AphfLco https://www.youtube.com/watch?v=RMK9AphfLco ELIZA is an early natural language processing computer program created from 1964 to 1966 at the MIT Artificial Intelligence Laboratory by Joseph Weizenbaum. Created to demonstrate the superficiality of communication between man and machine, Eliza simulated conversation by using a 'pattern matching' and substitution methodology that gave users an illusion of understanding on the part of the program, but had no built in framework for contextualising events.
- maweaver 3y agoAnyone using Eliza for more than a couple of minutes will quickly pick up on its tricks. Even in their cherry picked example it barely makes it a few messages in before the illusion starts to shatter with awkward lines like "Do you think coming here will help you to not be unhappy?". It might pass the turing test once or twice by sheer luck but it's clearly not competitive with a human conversing in general.
- throwaway4good 3y agoSure. With today’s eyes it is obviously a primitive computer program. But back then it was convincing to point where people suggested replacing real person psychologists with this. Likewise I think 50 years from now looking at present day ChatGPT it would terrible primitive and very unconvincing.
- emtel 3y agoNo it wasn’t! The Turing test is not “can at least one be fooled”, it’s “can an expert judge distinguish between a human and an AI”. Please read Turing’s paper, it’s very readable and is quite clear on this point