Google Developer Believes In Self-Thinking AI Model LaMDA

Abdul Ghani

An AI developer wants to prove internally at Google that the technologies are conscious and is later given a leave of absence. Google employee Blake Lemoine is apparently convinced that the company's AI model LaMDA has some kind of consciousness. The Washington Post reports. It goes on to say that Lemoine, who held this view internally at Google, was temporarily put on leave by his employer.

https://img.particlenews.com/image.php?url=2vwhPK_0g9Nmw3Y00
Google AI Model LaMDA.Image By Carlos Luna From Flickr

According to the report, Lemoine refers to the Language Model for Dialogue Applications (LaMDA) in his assessment of the AI, which is said to have developed a kind of life of its own. Google presented this language model for the first time at the I/O in-house exhibition in spring 2021. According to the announcement, LaMDA's main goal is to develop a system that can lead an open conversation without predetermined paths or sudden breaks. The basis of LaMDA is dialogues that were used for the training.

Lemoine was at Google's Responsible AI department, responsible for investigating LaMDA for possible bias and discrimination - a major and recurring problem with chatbots. In the course of this work, the developer repeatedly conducted dialogues with the LaMDA system, which apparently not only convinced Lemoine of the extremely advanced technology but also led him to the conclusion that LaMDA must have independent consciousness.

Google sees no evidence

Together with a colleague, Lemoine collected supposed evidence for his thesis and presented it internally. However, superiors and colleagues were not convinced. A Google spokesperson told The Washington Post: "Our team - including ethicists and technologists - reviewed Blake's concerns in accordance with our AI Principles and advised him that the evidence did not support his claims. He was told that there was no evidence that LaMDA is sentient (and much evidence against it)."

Google has officially suspended the developer and AI researcher for violating its confidentiality clauses. Lemoine is said to have tried to find legal representation for the LaMDA AI or tried to talk to representatives of the House Judiciary Committee about the technology.

Comments / 0

Published by

Member Of Freelancers Union (USA), Freelance Writer!, and Digital Creator. Ghani Mengal is an enthusiast Freelance blogger and digital marketer. His content has been published and featured on many popular blogs, websites, and publications. Including TeelFeed, LifeHack.org, Data-Driven Investor, TextSniper, Scientific Publication The Predict, The Startup, The Ascent, Heart Affairs, Illumination, And The List goes on.

N/A
2370 followers

More from Abdul Ghani

Comments / 0