Generative AI and the future of information

Recorded by:

Written by:

Senior Director and Principal Analyst

Last week, I was putting together an update on 5G materials and applications. For this kind of trends update, I typically draw on our conversations with startups, press releases from major corporations I know are players in this space, industry associations like the Global Mobile Suppliers Assocation, and of course, the news. I was searching for news on 5G materials when I clicked on this article “Revolutionizing Connectivity: How Global 5G Substrate Materials are Shaping the Future of Internet” on Fagen Wasanni Technologies, a news site. The article was about midway up the first page of the Google results. I hadn’t encountered the site before, but something about the article immediately made me uneasy, beyond the fact that it didn’t tell me anything meaningful. It wasn’t more than a minute before I figured out why: It was written by AI. And it wasn’t just the story: The image attached to the story, every story on the website, and everything else — probably down to the names of the “authors” and the website itself — were AI generated. The editorial team appears to consist of five people, who, if real, are the hardest working people on Earth: They have published, as best I can tell, 27,000 articles from mid-June until now, in early August. 

The conventional wisdom is that AI will either unlock a new era of productivity gains by automating busywork or usher in an era of global turmoil by putting everyone out of work. These scenarios are often presented as opposing viewpoints, but the reality is that they agree on the most important points: That AI will work as it claims and that it will have a major impact on society. Fagan Wasanni points to a different outcome: That the information economy will become polluted with AI-produced dreck. In that case, AI won’t revolutionize anything, but it will degrade (perhaps to the point of uselessness) a lot of the current tools used to gather information, like Google searches. If more of these AI-generated sites emerge, it’s not clear if it’s good for established news outlets or bad. Will established outfits like TechCrunch benefit from their existing brand presence and trust from readers? Or will brands cease to matter? The existing forms of internet searching, at the very least, will need to be updated to account for a proliferation of AI slop and try to detect it more accurately — Fagen Wasanni’s material is pretty obvious, but as the tech improves, it’ll get harder to tell. This shift will likely further erode trust in news sources, driving an even deeper divide on important public issues.

That all applies across a wide variety of types of news and information, of course — but what does it mean for innovation professionals? Fagen Wasanni covers a lot in those 27,000 articles, but most of it is innovation-related topics: semiconductors, robotics, startup funding, and of course AI itself. The proliferation of an ocean of AI-generated articles is going to make finding news a lot harder for any innovation professional trying to get a real handle on trends and technologies. The value of information curation and intelligent commentary is going to be more important than ever; finding trusted voices and sources of news will require more effort and more resources. One counterintuitive result may even be a return to some old-school approaches: Since the COVID-19 era, innovation has increasingly gone digital and virtual — but handshakes, in-person meetings, and conversations within trusted networks are much harder to spoof with AI (for now!), and our clients should consider ways to use these “retro” approaches as a source of competitive advantage. 

What do you want to research today?