The harmful mythology of AI, according to the father of VR
Event Date: 11/6/2018
“Fake people are a bad thing, in my opinion, because we all pay a price for them,” said Jaron Lanier, as he explained his thoughts on the dysfunction of the current framing and philosophy surrounding artificial intelligence, or AI.
Lanier made it clear that he is not critical of the codes associated with or referenced as “AI,” such as machine learning. In fact, he was mentored by Marvin Minsky, a pioneer in the field, and is himself internationally recognized for inventions and advances that have fueled the field.
“I’m not criticizing the code, I’m criticizing the philosophy and the framing — the storytelling, which is absolutely a crucial part of any technology because it influences how it enters into the human world of experience,” Lanier told a crowd of more than 280 people Tuesday night at Purdue University.
Named one of the world’s 100 most influential people by Time magazine, Lanier is a scientist, technologist, author, artist and musician known as the father of virtual reality. His talk, titled “The Concept of AI Harms the Technologies Created Under Its Banner,” was the final keynote address for Purdue’s Dawn or Doom annual conference in Stewart Center’s Fowler Hall.
“The whole point [of the AI mythology] is that in theory we’re creating this sort of free-standing entity, this new creature that is our partner and it has to be sort of person-like and it has to itself be the source of value,” he said. “In order to create that feeling we have to actually devalue real people who are actually needed. There is a kind of a conservation of personhood, if you like. … This principal is a general one in information systems. Any time you create fake people, you actually do deprive real people of value.”
As examples, he pointed to the harm of fake accounts trying to create false social perception online, a female robot that was given citizenship and privileges in Saudi Arabia that real women in the country have yet to gain, and a language translation tool that pulls the data it needs from translations shared in individual communications without informing or compensating the individuals on which it depends.
Lanier also discussed how the majority of AI algorithms are being used by massive companies involved in search engines and communication platforms to manipulate people, despite the great potential for use to help agriculture, health and the human experience.
Lanier, however, offered hope and ways to fix the problems currently surrounding the concept of AI.
For those whose jobs are threatened by the technology, he suggested including them in the process and compensating individuals for data used by the systems.
“Spiritually maybe instead of them perceiving the future as something that is their enemy, that is making them obsolete, that doesn’t need them, they would say, ‘Hey, it’s a future that needs me in a new way,’” he said. “It’s not only more positive, but it’s honest.”
He also mentioned the importance of elevating people over things, adding: “If technology is to serve people, we have to believe in people.”
The conference and Lanier’s presentation aligned with Purdue’s Giant Leaps sesquicentennial celebration and its Ideas Festival theme, Giant Leaps in Artificial Intelligence, Algorithms, and Automation: Balancing Humanity and Technology.” The Ideas Festival is the centerpiece of Purdue’s 150th anniversary, and connects world-renowned speakers and the University’s expertise in a conversation about the most critical problems and opportunities facing the world.
Karthik Ramani, Purdue’s Donald W. Feddersen Professor of Mechanical Engineering, has used machine learning and virtual reality in his research projects. Ramani, who attended Lanier’s presentation, is part of a team that received a $2.5 million grant from the National Science Foundation’s Future of Work at the Human-Technology Frontier program to develop realistic simulations to help businesses and people prepare for the factories of the near future.
“In my work, I want to see more productivity in manufacturing by seeing robots and humans work together,” Ramani said. “He made a very important point about trying to amplify the human capacity. That message was embedded in his talk, and I think it is really important. That way, you are sort of retooling people to do better.”
“I also agree with him on the myth part of artificial intelligence, in that he encouraged students to focus on the math and algorithms part, and solving problems with or without AI in the conversation,” Ramani added. “We do create bubbles often to make progress and AI is also one.”
Soumendu Kumar Ghosh, a graduate student studying electrical and computer engineering who works in the AI field, said the talk was eye-opening.
“The fact that these speakers make us understand the social impact of our work and the ethical values we should have, that’s very important,” he said. “If we go into industry and we continue the same processes and continue the same beliefs that influence our industry, then this fantasy will keep on living. Bringing these speakers to the universities is important because they can influence the students, and students are the ones going to the jobs who will influence the next five years or 10 years.”
In addition to his public talk, Lanier met with students from Purdue’s Envision Center and the Purdue Virtual Reality Club.
Jordyn Lukomski, a graduate student studying computer graphics technology who was a part of the student meeting and attended the public talk, said the chance to meet Lanier in a smaller setting allowed a rare opportunity to get direct responses to questions of personal interest.
“The biggest inspiration I took from meeting him was the idea of this realistic optimism,” she said. “Much of what he explains is surrounded by this attitude and approach to future technologies and the current failings of existing technologies. He does not shy away from the hard truths or the harmful ways technology is created or being used. He is very honest and realistic about what is happening. However, he is still optimistic for pushing for a better future, a more ethical future.”