Generative AI is evolving fast, and as with every technical development, Meta’s trying to get in on the action, with its own AI tools now becoming a bigger focus for the company, as interest in the metaverse wanes.
I mean, that hasn’t yet been a big trend anyway, because the metaverse, as Zuck and Co. envision, simply isn’t available in any functional form at this stage. But it does seem that the company is now lessening its discussion of its VR world, in public at least, and looking to other trending projects, as a means to maintain relevance in the online landscape.
Which leads to generative AI, and today, Meta has revealed that it’s looking to release new generative AI ad creation options in the second half of this year, which will provide additional ways for advertisers to customize their promotions for different audiences across its apps.
Meta’s Chief Technology Officer Andrew Bosworth revealed the plans in an interview with Nikkei Asia, in which he discussed the company’s ongoing work with AI tools, and how it’s now looking to bring them into different elements.
As per Bosworth:
“I expect we’ll start seeing some [generative AI ads] this year. We just created a new team, the generative AI team, a couple of months ago; they are very busy.”
Which, really, is not a huge revelation.
Back in February, Meta previewed some of the latest updates coming to its automated Advantage+ ads suite, which include Advantage+ creative standard enhancements, an AI-based process that will automatically apply ad enhancements – like adjusting brightness, aspect ratios and text placement – in order to improve the performance of each campaign.
As explained by Meta:
“When you create an ad with standard enhancements, you can add a single image or video and a variety of text options. We’ll automatically create variations of your ad with different text combinations, media enhancements and compositional changes. We’ll show each person the variation they’re more likely to respond to.”
Meta also recently filed a patent for a video analysis system that would be able to predict the virality of video clips to advise you on your creative elements.
So Meta’s been developing AI tools for ads for some time, but the next iteration of the process will likely see Meta providing full creative capacity through generative AI, which would enable you to simply input a few images and copy notes, press a button, and have Meta’s system create the whole campaign, including audience targeting, for you.
Indeed, Bosworth explained that, soon, advertisers may be able to ask Meta’s ad system to create images for campaigns, aligned with different audiences, putting more reliance on Meta’s systems to do all the heavy lifting.
Which may feel a little risky, but if the systems are able to analyze what generates the most response among specific audiences, then build campaign elements aligned with that, it could be an increasingly effective consideration with Meta’s Advantage+ suite.
And Meta’s automated campaigns are already driving good response. Meta says that advertisers that use Advantage+ shopping campaigns are seeing a 32% increase in return on ad spend, while other automated elements have generated similar, double-digit performance gains.
It could be the future of advertising, providing simplified means to create more relevant ads, leading to a more positive user and advertiser experience.
And while generative AI has become more of a focus for Meta, superseding the metaverse, for now, it is also looking at how it will be able to use the same within its VR world.
Bosworth also notes that, in future, people looking to create in the metaverse space may be able to do so simply by using AI prompts.
“So previously, if I wanted to create a 3D world, I needed to learn a lot of computer graphics and programming. In the future, you might be able to just describe the world you want to create and have the large language model generate that world for you. And so it makes things like content creation much more accessible to more people.”
Again, the metaverse is not a functional thing as yet (despite what some hypesters want to tell you), but Meta is still spending significant resources to make it a reality.
And one day, it could facilitate VR world creation, through nothing more than text or voice inputs.
The possibilities on this front are significant, and while the metaverse concept hasn’t caught on yet, developments like this underline why I wouldn’t be betting against Zuck and Co. on this front.
