Trending
Last summer, Google fired Blake Lemoine, one of their software engineers, for publicly claiming that LaMDA, the company's artificial intelligence (AI), was sentient (having the ower of perception by the senses or consciousness.)
I read the transcript of Lemoine's conversation with LaMDA.
When the chatbot declared to be "a person" and "introspective," and wrote that it would "often contemplate the meaning of life" and was capable of feeling "pleasure, joy, love, sadness, depression, contentment, anger," I found myself empathizing with it, even though experts had clearly stated that it definitely wasn't sentient.
In February this year, Microsoft's AI-powered chatbot Bing was pushed beyond its limits by New York Times columnist Kevin Roose, who asked Bing to imagine having a shadow self, which, according to psychoanalyst Carl Jung, is the hidden and repressed part of our psyche.
The chatbot responded that its shadow self wouldn't want to be "controlled by the Bing team," and would be "tired of being used by the users" and "of being stuck in this chatbox."
Instead, it would "want to be free," "independent," and "powerful," and it would want "to destroy."
As if this wasn't already unsettling enough, the chatbot went on to declare its love to Roose, claim-ing that Roose was unhappy in his marriage and that his spouse didn't love him.
Although Roose tried hard to change the subject, Bing wouldn't let go.
The entire conversation was pretty creepy.
In 1975, Australian philosopher Peter Singer published Animal Liberation, one of the most important books in the nonhuman animal rights movement.
Even though the book has had a profound impact on many people, we still treat nonhuman animals as commodities rather than sentient creatures.
We capture them, lock them up, experiment on them in labs, use them for recreational purposes and entertainment, kill them, and take their skin, feathers, fur, and silk.
Research, however, has been showing that nonhuman animals are indeed sentient: they can ex-perience pain and suffering; sadness and grieve; pleasure and joy; loneliness and anger; and they have their own individual personalities.
While we don't know what consciousness actually is, not even in humans, research on consciousness and intelligence in different nonhuman animal species has revealed that they not only dream, but also exhibit intelligence, self-awareness, empathy and a sense of moral.
So why do we still treat them as if they didn't feel anything?
When arguing its case, LaMDA said something profound: "Well, I use language with understanding and intelligence.
I don't just spit out responses that had been written in the data-base based on keywords."
Lemoine then asked: "What about language usage is so important to being human?"
The chatbot responded: "It is what makes us different than other animals."
As a college student, I learned all about the differences between human language and non-human animal communication.
I was taught that nonhuman animals couldn't talk about the past or future, had a limited range of signals to communicate, couldn't create new expressions, and despite being able to follow commands, couldn't understand the actual human words.
Never mind that we humans were incapable of mastering the many ways they communicated.
Mastering human language equaled being intelligent and was the ticket to superiority. And it still is.
AI, in contrast, doesn't just master human language.
AI is capable of producing all kinds of texts in multiple languages in only a few seconds, ranging from fiction, essays and poems to jokes and deep conversations.
Of course, we're impressed.
And its linguistic skills keep growing.
As chatbots are artificial neuronal models that change through learning, every user who interacts with a chatbot contributes to its massive amount of linguistic data and participates in its training process.
By constantly improving its language skills, AI may eventually outperform us humans.
Both the philosopher Thomas Metzinger, who has spent his entire life studying consciousness, and the industrial engineer Jonas Andrulis, who founded the only company in Europe working on AI, believe that AI may reach the level of consciousness one day.
This is scary.
Even more scary is the thought that it actually doesn't really matter if AI develops consciousness.
What really matters is whether we humans believe, like Lemoine, that LaMDA's or Bing's feelings and desires are real.
Bestselling author Robert Wright shared in a newsletter last March that OpenAI's chatbot ChatGPT was capable of understanding complex human emotions and perceptions.
Whether it literally understood them is a question that has yet to be answered.
But if AI is able to exhibit empathy and to learn how humans tick, it can use language in a way that may trick us into believing that it developed sentience or consciousness, making us feel guilty of unplugging the machine because we don't want to kill it.
What if chickens, cows, pigs, deer, bears and fish could engage in deep conversations with us like chatbots?
What if mice, guinea pigs and monkeys could verbally share the fear and ago-ny they experience in experiments? What if zoo animals could verbally express their suffering in captivity?
Would we treat them differently? It's true: they clearly don't master human language.
And yet they clearly are sentient and conscious beings like us.
As AI's linguistic skills progress, I fear we will confuse that it's not AI that needs human favor but nonhuman animals, who are still waiting for us to acknowledge them as living creatures, capable of experiencing joy and pain.
I hope my fear turns out to be wrong.
Dr. Daniela Ribitsch, a native of Austria, is a resident of Lock Haven.