Alyza Mari Mendoza

Reviving one of the pioneers of artificial intelligence (AI), researchers brought back the world's first AI chatbot, ‘ELIZA,’ as they found long-lost 60-year-old computer code from archived records.


AI significantly developed in the present after almost 60 years, and ELIZA’s resurrection reminds the world of ELIZA — the pioneering milestone that paved the way for present AI development.

ELIZA was developed in the 1960s with only 420 lines of code by Massachusetts Institute of Technology (MIT) professor Joseph Weizenbaum as the first-ever chatbot simulation.

Weizenbaum was the first one to make a human-to-machine program that imitated human-to-human interactions, trailblazing the way for the chatbots of today.

As a language model and chatbot simulation, ELIZA responds to questions by acting like a psychotherapist, using phrases of “Please tell me your problem” and “Why do you feel sad?”

Initially, to write ELIZA, Weizenbaum used Michigan Algorithm Decoder Symmetric List Processor (MAD-SLIP), a programming language he invented which was immediately copied into the language Lisp.

Thought to be “lost” and hidden among the archives at Massachusetts Institute of Technology (MIT), ELIZA’s long-lost computer code was discovered by Stanford cognitive scientist Jeff Shrager and MIT archivist Myles Crowley in 2021.

Not knowing if it would work and despite the challenge, the researchers used the original code to painstakingly debug and recreate the program — bringing ELIZA back to life on December 21.

“It not only worked, but worked extremely well,” said Shrager as the team successfully resurrected ELIZA.

“Bringing ELIZA back, one of the most — if not most — famous chatbots in history, opens people’s eyes up to the history that is being lost,” said digital humanities professor David Berry.

ELIZA being a pioneer chatbot remains a legacy, and its resurrection highlights the history and progress of AI development.