China’s state-run press agency, Xinhua, just launched its first artificial intelligence (AI) anchors at the ongoing fifth World Internet Conference in east China’s Zhejiang province.
Based on the model of Xinhua’s two anchors and launched for English and Chinese platforms, the model has been created using the latest AI technology.
These AI anchors are basically digital composites created from footage of human hosts that read the news using synthesized voices, which means an AI system was used to synthesise presenters voices, lip movements and expressions.
The press agency even launched the first video of the anchor speaking the news. Watch the video here:
According to Xinhua, the AI anchor “has a male image with a voice, facial expressions and actions of a real person. ‘He’ learns from live broadcasting videos by himself and can read texts as naturally as a professional news anchor.”
It will work 24 hours a day on the official website and various social media platforms of Xinhua, which in turn will help reduce news production costs and improve efficiency.
As always with news on AI, jokes and semi serious observations about more kinds of jobs that could be lost to AI have been doing the rounds. Could journalism be the next career that could be lost to AI?
Driverless cars could put drivers out of jobs; automation has already taken many shop floor jobs; natural language processing and better translation options are taking away even language-related jobs; and now that AI has mastered the ancient Chinese game of Go (it is a far more complicated game than chess, having more than 300 times the number of plays as chess), maybe journalism is the next career that could fall to AI!
We’re not especially worried about this one, however. A tireless machine reading text that has been fed into its systems (almost) as well as a human being is most certainly not journalism.
Until AI finds itself a nose for news and, more importantly, the ability to speak truth to power, journalists will not be without jobs. The beauty of machine learning and neural networks that power AI is that they allow it to learn skills such as driving and master the ability to deal with multiple probabilities, but the ability to handle moral ambiguities responsibly, and to weigh individual frailty against collective good remains a very human part of the job.
No doubt AI can make useful companions to journalists, being able to scrape data, find needles in haystacks, and break complicated codes, all the while ensuring that their fridges are never out of milk (using the wonders of the Internet of Things).
But the day when an AI can make trusted contacts, chase endless empty leads based only on a hunch, and fearlessly stand its ground even as authoritarianism rages around it is not about to happen just yet.