Do We Really Want Computers That Act Like People?
Computers have talents that far exceed ours. Why do we put so much effort into making them more like us?
Entrepreneur authors’ views are entirely their own.
I consume a lot of information on technical advancements. It is both my professional and personal interest. It’s unusual that a podcast captivates and challenges me to the point that I start scribbling notes for future thought experiments and article themes.
Tim Ferriss’ talk with Eric Schmidt was unquestionably one of those occasions. Schmidt promoted his new book, The Age of AI: And Our Human Future, co-authored with Daniel P. Huttenlocher and Henry Kissinger, on radio shows and podcasts in late 2021. Schmidt delves into the benefits and drawbacks of artificial intelligence.
- Machine vs. man?
One of the most fundamental problems Schmidt and his book pose, in my opinion, is why we even compare computers to people. One of Schmidt’s arguments in the Ferriss interview, for example, is that computers can see better than humans, and he means it literally. He even claims that computers should be able to drive cars and give medical checks. My queries are perhaps more basic, but I’m curious if a machine can truly see. Is a sentient seer required for seeing? And why is human vision used to measure machine vision?
While Schmidt’s book predicts AI’s eventual advancements in computer perception and processing, I disagree.
The humanization of artificial intelligence
In terms of how we envision the evolution and refinement of AI in respect to human capabilities, Eric Schmidt’s description of AGI is instructive. “AGI stands for artificial general intelligence,” Schmidt explains. AGI refers to machines with human-like strategies and capabilities. Today’s computers and algorithms are extremely good at providing content, misinformation, guidance, and scientific assistance, among other things. They are not, however, autonomous. They don’t have the concept of “Who am I, and what should I do next?” Surprisingly, the aim we’re pursuing with AI is the very thing that distinguishes us from the machines we build.
And yes, this is both a frightening and exciting idea.
Schmidt cites two instances.
Schmidt’s book, as well as my uneasiness about AI’s future, is around the ethical question of whether it’s wise to create AI and AGI with these powers. We must proceed with caution. However, there is a chance that we will be able to avoid building our own destroyer.
Schmidt, I believe, touches on one of AI’s ethical centers when he discusses video surveillance. “If you want to eradicate the vast bulk of crime in our culture, place cameras with face recognition in every public location,” he says. It’s also against the law to do so. However, if you are truly concerned about crime at the expense of human liberty and the right to privacy,