AI hype is done: Time to embrace reality

This year has been one of nonstop AI hype and it’s been difficult to avoid the ubiquitous subject in everyone’s timeline and content feeds.

But it’s not just hype. The Information reported this week that Microsoft-backed artificial intelligence company OpenAI is on track to generate more than $1 billion in revenue over the next 12 months—up from just $28 million in 2022.

However, now the reality of artificial intelligence is setting in, organizations are discovering what generative AI really means for their businesses and operations.

Axios reported that content checker Originality.AI shows one in five of the top 1,000 websites in the world are blocking crawlers that trawl the web for data for AI services.

Remembering how they were burned by Google in the early days of search, which ultimately destroyed great swathes of their businesses, media outlets including The New York Times, CNN, Reuters, Amazon and Axios itself are blocking OpenAI’s web crawlers from accessing its content — although most of them want to keep their commercial and partner content open to the bots.

It’s something of a Catch-22 for media companies. They know the power of Google for powering access to content. They also know how much the search behemoth has damaged their business and exploited their content in the past. They’re in the early stages of trying to plot a safer path through the ChatGPT minefield.

The Associated Press has done a deal with OpenAI that enables the ChatGPT company to license AP’s archive of news stories. Other media companies are also in discussions about similar deals.

Everyone is also realising that generative AI is raising risks of copyright infringement and exploiting public data without the user’s knowledge that some of the content might be proprietary.

Most companies are now aware that they should invest in a closed AI environment to experiment with generative AI to ensure their confidential and proprietary information doesn’t get into the public domain. Our analysis piece this week digs further into that topic.

My colleague Frank Washkuch, executive editor of PRWeek, participated in a fascinating panel about generative AI and the PR profession with Columbia University’s Strategic Communication graduate program last week.

Angles discussed included reducing tasks from hours to minutes, concerns about whose job it is going to take and how few employees are aware how quickly you can give away a company’s IP onto an AI platform and let others exploit it.

AI tools can be incredibly powerful. They can help sell ideas and concepts quickly and more efficiently. But they can also make mistakes and look amateurish. The human touch and input is still super-important.

As Frank says on the latest episode of The PR Week podcast, AI is on every in-house team and every PR firm’s mind, but he’s not sure they have their arms fully wrapped around it yet.

That needs to change.

And quickly.