The Chinese AI anchor man looks very much like the average Chinese citizen, a typical Chinese guy with that oddly intellectual look. He looks reassuring, made for his market like most news readers’ images are supposed to be. He’s modelled on a Xinhua regular called Zhao Zang, and his AI makes this reader more effective and efficient than conventional computer graphics interfaces (CGI).
Does this mean the end of news readers?
Yes and no. The extent to which an AI can interact with live casts, etc. isn’t known. You’d need some pretty agile software to hold a conversation, and while it’s not at all unlikely that can be done, getting an AI to formulate and ask questions may be some distance in the future. (Advanced AI can do it, but contextual questions are typically about what the AI itself is doing, not its immediate external environment. Memory function alone would be an obstacle.)
The other issue, less obvious in technological terms, but pretty clear in marketing terms, is the image to the user. A good regular news reader is considered an asset on most news networks, a good draw for viewers, but 24/7 AI news readers. It can’t be the same thing.
Some might say that cookie cutter image news readers are basically all the same. Not true. Human perception is far more subtle than that, and reacts more to a human face is doing than to what is basically an image. Human beings automatically read faces, and even the most stoic, impassive news reader sends different signals.
Why do this at all?
AI newsreaders are not an end in themselves. It’s an interesting idea, but not the whole story. The ability of a focal point like a human image to deliver information is a critical point. It’s fair to say people are motivated to watch news largely through imagery, and that a good reliable image which never gets sick and never makes mistakes is pretty useful.
This technology, however, goes a lot further. This is putting delivery of information in to a very different and potentially very valuable context. Imagine AI lecturers, AI people able to interact with the public in high stress situations, AI emergency services on your phone, etc. If this very basic type of communication works, it can be used in a lot of very important areas, safely, without committing human resources on the ground.
A big future
More responsive, agile AI communications can do a lot of very useful things. AI can respond effectively, and deliver information in real time. It can at least theoretically answer questions, ask questions, and act as a mobile interface. In a terrorism or crime situation, an AI could be invaluable as an observer using CCTV, or other remote information gathering technologies.
Artificial intelligence in business and security is already a huge employment market. Most jobs focus on managing AI, some in development, but the range and scope of AI is spreading daily. You can use the employment market and its ever expanding scope as a good metric for how real world AI is entering the economy and human space. Our Chinese news reader is a good example of how far this technology has already managed to go.
This is where AI is going – Straight to basic communications, and bringing a suite of capacities with it. Ultimately, it boils down to how good the AI is, how many capabilities it has, and its ability to manage its accumulating real time information effectively.
This could well be the first public AI operation, known to be AI, and directly replacing humans in the information industry. How much bigger could it get? We’re about to find out.
