Man vs machine

“ Is that you, Ed?”

That was how Roby Alampay opened last Saturday’s edition of “The Chiefs” on Cignal TV’s One News channel.

He was greeting our co-host Ed Lingao, news anchor and chief of TV 5’s news correspondents. We resumed regular in-person taping of our show at the Cignal studios only a few months ago, after over two years of livestreaming because of the COVID pandemic.

One day, livestreaming may not even be needed, as TV anchors generated through artificial intelligence take over.

GMA 7 has roiled the waters of Philippine broadcasting after it unveiled its newest sportscasters, Maia and Marco, at the opening ceremony of the National Collegiate Athletic Association Season 99 men’s basketball tournament on Sept. 24.

Sleek and rather creepy as AI-generated images tend to be, Marco’s features seem to have been inspired by K-drama heartthrobs, while Maia looks like a composite of several of our local actresses.

Amid concerns raised by newscasters about their endangered jobs, GMA has explained that Maia and Marco are merely enhancing sportscasting, performing routine work, presenting sports stats rather than providing the analysis and commentaries that make sports coverage exciting and separate seasoned sportscasters from newbies and plain TV talking heads.

That hardly allayed the concerns in the industry, since broadcasting employs a lot of talking heads (several of them highly paid) and those whose jobs are similar to what Maia and Marco are doing. Such jobs may soon become redundant. AI-generated sportscasters are available 24/7, never tired or on sick leave, and (perhaps best of all for corporate bottom lines) have zero compensation.

Tech experts have stressed that AI cannot replace the human expertise and experience that go into credible reporting, analysis and commentary on news and current affairs.

With this in mind, we could joke about AI on The Chiefs. But who’s to say AI won’t one day reach the level of sophistication that makes human analysts / commentators redundant?

Such questions are raging in the industry, as the use of AI-generated images like Maia and Marco becomes more widespread.

*      *      *

For journalists, a bigger concern is credibility. As the latest Reuters Institute report showed, trust in media is down worldwide, partly because of the prevalence of fake news and disinformation. Social media is the main platform for this, but the fallout is affecting mainstream media.

Public trust in properly vetted news, which is what mainstream media outlets provide, must be regained and strengthened. My former reporting colleague Rachel Khan, now a journalism professor at the University of the Philippines College of Mass Communication in Diliman, said regaining public trust cannot be helped along by the use of “inauthentic” AI-generated images for news reporting.

We asked Rachel, who was our guest in last weekend’s edition of The Chiefs, to sum up her reaction to the fielding of Maia and Marco. Her reply: “It’s wrong.”

She stressed that AI is already widely used as a tool for journalism, capable of doing more work and collating more data within a much shorter period beyond normal human capability.

But the layers of vetting, putting stories in proper context, explaining the relevance of an issue or event, and the final presentation of the work of journalists cannot be entrusted to AI.

The University of the Philippines Broadcasting Association expressed alarm over the rollout of Maia and Marco.

In a statement, the group declared: “The UP Broadcasting Association firmly believes that this move does very little to serve the people as it only sets an alarming precedent that would profoundly impact the future of broadcasting and those who aspire to be in this industry. While it is inevitable to adapt these kinds of emerging technologies in the practice of broadcasting, we should not forget that these technologies are mere tools to assist us and improve our work. They should not, in any way, replace and displace the people who have spent years in the study and practice of broadcasting.”

The group added: “Before we welcome innovation and technological advancements, we must first address the realities faced by media workers – contractualization, unfair labor practices, low compensation and lack of benefits. If GMA truly aims to promote inclusivity in their reporting, they should instead focus their efforts and resources on hiring and properly compensating talented journalists and media workers. They should provide more opportunities to people who aspire to be part of the media industry and are working hard to do so.”

With AI, there is also greater danger of information manipulation by the state. It’s no coincidence that China was the first to use AI-generated images in broadcasting.

And yet it seems inevitable that we will be seeing more Maias and Marcos delivering the news.

*      *      *

Ethical questions revolving around man versus machine have been raised as far back as 1950, when British mathematician Alan Turing, described as the founder of artificial intelligence and cognitive science, tried to find out if machines could think. The World War II code-breaker and digital computer designer developed what has been dubbed as the Turing test, to determine if a computer is thinking.

The test involves a human interviewer carrying out a conversation with another person and a computer. If the computer fools the interviewer into believing it is the human, this can indicate the machine’s capability to think.

Last year, Google’s AI passed the test, amid criticism that what the Turing test actually measures is not computer intelligence but deception capability. Some critics say the test should be retired.

Six months ago, tech entrepreneurs led by Elon Musk along with AI engineers and researchers proposed a six-month pause on the development of AI systems more capable than ChatGPT developer OpenAI’s latest GPT-4 language generator.

Instead of a pause, companies accelerated their pace of AI development. And last week, just 10 months after its rollout, ChatGPT got a voice feature and capability to react with images to voice prompts.

Maia and Marco may not pass the Turing test, but we have seen that they can present sports stats.

Rachel Khan believes the novelty of the AI presenters will soon wear off, and what will be left is the bad taste in the mouth in consuming news delivered by inauthentic images.

On the other hand, with AI technology evolving at warp speed, the trend may also go the other way, and variations of Maia and Marco will become the norm.