A Chinese state news organization recently decided to employ an artificial intelligence in the chair (or at least on the screen) of the anchor desk. This idea has been met with a lot of derision, much of which has been very dismissive without a lot of nuanced objection.
China is a large country with complex social issues. It is not this writer’s intention to decide what is or isn’t best for the Chinese people. But there are good reasons to believe a move like this would not work for a country like the US, and perhaps other parts of the world. Here are a few of those reasons.
News Is About Trust
The U.S. has a rich history of news anchors that became the face of news for a generation. That face was of a person we could trust and relate to. We don’t necessarily trust government officials or business executives. So it is that much more important that we have at least one official voice we trust.
“It’s quite difficult to watch for more than a few minutes. It’s very flat, very single-paced, it’s not got rhythm, pace or emphasis,” Michael Wooldridge from the University of Oxford told the BBC. And compared to a trusted human news anchor, he says that “if you’re just looking at animation you’ve completely lost that connection to an anchor.”
It could be that in China, there is more implicit trust in the state and state officials. Since the press is run by the state, it may not need the same kind of human connection to establish trust as we do. We want to look into a person’s eyes and know they are telling the truth. Trust does not come easy for us.
In the United States, the media is the fourth estate. It is not a part of the government. It is a part of what makes the government work. An adversarial relationship with the government is almost a requirement. It is sometimes said that when you report what they want you to report, it is propaganda. When you report what they don’t want you to report, it is news. For that to happen, the news must be completely independent from those in power.
An AI reporter would be seen as being completely under the control of some powerful organization. Even a human newscaster can go independent, or go to another agency, or tell their truth in some other way. An AI reporter only reports exactly what the controllers want, every single time.
We Value the Flaws
Finally, newscasters are only human. That means they have to work hard and sometimes make mistakes. That’s the way we like it. We value the humanity of people we trust. We relate to flawed humans who make mistakes. We wouldn’t connect at all to artificial perfection.
Some critics focused on the inflectionless, emotionless, and wooden delivery of the news as the main problems for the AI. But none of that will be the fatal flaw. We already nave better text to speech than that. We can get to the point where the AI is almost human. But it would matter. What matters is that it is not a source we can trust, is under the control of the government, and assumes an air of perfection that is not relatable.