Generative AI will transform digital media

Advertising is geared towards gaining consumer awareness first and interest second. The more eyeballs that are drawn to an advert and the more interesting it is, the more people consider the product.

Digital and social media have reduced the cost of reaching audiences, while content creation tools have made it cheaper to produce content than ever before. At its peak, digital advertising was the most effective way to reach most audiences, and it rapidly took a major share of media spends. That magical time was 2022. 

 However, today the trajectory is moving against digital media. Generative artificial intelligence (AI) is a black swan event in the world of digital media. 

 In social media, engagement rates are dropping by 50% as the quantity of content increases by 25%, according to the Buffer and BuzzSumo apps. Having fewer people engaging less, with more content competing for their attention and interest, doesn’t sound like a great formula for increasing effectiveness. 

 As generative AI comes into play, more content will be created by marketers and advertising agencies at an even faster rate. “Personalisation” will become easier as deep fakes will make us question every message we see and receive. Over time the fake stuff will crowd out the real stuff and drive engagement down even further. This is known as Gresham's law (bad money crowds out good money — in the 16th century — or bad information crowds out good information in today's world).

While it may sound like an extreme model, just ask yourself: what percentage of your content needs to be fake or feel fake before you opt out of social media channels? 30%? 40%? 50%? The same network effects that made people join social networks in the first place work in reverse. Expect the loss of real users to be an avalanche.  

 Not only will content and information be fake, users will be fake too, as criminals will supercharge their bot farms. So expect a lot of your advertising to end up with bots clicking and becoming leads to generate revenue for fake publishers. This will mean that investments in sophisticated analytics and customer relationship management systems will be less valuable as fakes clog up the systems and your data becomes gobbledygook. 

 So online advertising will continue to grow, but become increasingly less effective for advertisers. Platforms will need to reconsider anonymity and will push for authentication of users to ensure advertisers get what they are paying for. This will be a positive development for advertisers, as only real users will be on the platforms, but privacy will become a central concern for everyone.  Facebook’s recent $1.3 billion fine in the EU for contravention of its data protection regulations will make everyone nervous.

The complexity of authenticating users will be immense, because on the other side will be generative AI specialised in spoofing users. In the end, consumers probably don’t want to jump through so many hoops or lose so much privacy, and things will spiral downwards. 

 As fake information becomes more prevalent, what will the  AI models find when looking for answers? How will they be able to distinguish between fake and real information? Will the entire ecosystem collapse under the weight of bad intentions by a greedy few? Will your Google AI results return a series of misinformation or will only a glorified few be the custodians of truth? 

Nevo Hadas is a partner at DYDX Digital. 

The big take-out: Online advertising will continue to grow, but it will become increasingly less effective for advertisers.