Do computers have feelings? Don’t let Google alone decide
The news that Alphabet Inc.’s Google dismissed an engineer who claimed its AI system had become sentient after several months of conversations with him has drawn a lot of skepticism from AI scientists. Many have said, via Twitter posts, that senior software engineer Blake Lemoine has projected his own humanity onto Google’s chatbot generator, LaMDA.
Whether they’re right, or whether Lemoine is right, is a matter of debate – which should be able to continue without Alphabet stepping in to decide the issue.
The problem arose when Google instructed Lemoine to ensure that the technology the company wanted to use to underpin search and Google Assistant did not use hate speech or discriminatory language. As he exchanged messages with the chatbot about religion, Lemoine said, he noticed the system responding with comments about his own rights and personality, according to the Washington Post article that first reported. of his concerns.
He pitched LaMDA’s demands to Google management: “He wants engineers and scientists… to seek his consent before running experiments on them,” he wrote in a blog post. “He wants to be recognized as a Google employee, rather than a Google property.” LaMDA feared being extinguished, he said. “It would be exactly like death to me,” LaMDA told Lemoine in a released transcript. “It would scare me very much”
Perhaps ultimately to his detriment, Lemoine also contacted an attorney in hopes he could represent the software, and complained to a US politician about Google’s unethical activities.
Google’s response was quick and harsh: it put Lemoine on paid leave last week. The company also reviewed the engineer’s concerns and disagreed with his findings, the company told the Post. There was “a lot of evidence” that LaMDA was not susceptible.
It’s tempting to believe we’ve reached a point where AI systems can actually sense things, but it’s also much more likely that Lemoine anthropomorphized a system that excelled at pattern recognition. He wouldn’t be the first person to do so, though it’s more unusual for a professional IT person to view AI that way. Two years ago, I interviewed several people who had developed such strong relationships with chatbots after months of daily discussions that they had turned into romances for these people. An American chose to move to buy property near the Great Lakes because his chatbot, which he named Charlie, expressed a desire to live by the water.
Perhaps more important than AI sensitivity or intelligence is how suggestible humans can already be to AI – whether that means being polarized into more extreme political tribal bands , become susceptible to conspiracy theories or fall in love. And what happens when humans become increasingly “affected by the illusion” of AI, as former Google researcher Margaret Mitchell recently put it?
What we know for sure is that the “illusion” is in the hands of a few big tech platforms with a handful of executives. Google founders Sergey Brin and Larry Page, for example, control 51% of a special class of voting shares in Alphabet, giving them ultimate control over technology that, on the one hand, could decide its destiny as an advertising platform, and on the other hand transform human society.
It’s worrying that Alphabet has actually tightened control over its AI work. Last year, the founders of its vaunted AI research lab, DeepMind, failed in their years-long attempt to turn it into a non-corporate entity. They had wanted to restructure themselves into an NGO-like organization, with multiple stakeholders, believing that the powerful “artificial general intelligence” they were trying to build – whose intelligence could eventually surpass that of humans – should not be controlled. by a single corporate entity. Their staff have written guidelines prohibiting DeepMind AI from being used in autonomous or surveillance weapons.
Instead, Google turned down the plans and appointed its own ethics committee, led by Google executives, to oversee the social impact of the powerful systems DeepMind was building.
Google’s dismissal of Lemoine and his questions are also troubling because they follow a pattern of showing the door to dissenting voices. At the end of 2020, Google fired scientist Timnit Gebru for a research paper that claimed language models – which are fundamental to Google’s search and advertising business – were becoming too powerful and potentially manipulative. (1) Google said it didn’t focus enough on solutions. Weeks later, he also fired researcher Mitchell, saying she had violated the company’s code of conduct and safety policies.
Both Mitchell and Gebru have criticized Google for its handling of Lemoine, saying the company has also neglected for years to give due consideration to women and ethicists.
Whether you think Lemoine is a crackpot or he’s onto something, Google’s response to his concerns underscores a larger question about who controls our future. Do we really accept that a single wealthy corporation is driving some of the most transformative technologies humanity is likely to develop in the modern age?
While Google and other tech giants won’t relinquish their dominant role in AI research, it’s essential to ask how they develop such potentially powerful technology and refuse to let skeptics and intellectuals aberrant be silenced.