Meta’s Developing ‘Business AIs’ to Enhance Transactional Engagement

I remain unconvinced that people are going to care about Meta’s AI bots that seek to simulate engagement with real humans in its apps.

It tried this back in 2016, with Messenger Bots, which nobody cared about, then it tried it again last year, with celebrity-styled, AI-powered bots, which also nobody cared about, leading to their eventual demise.

But now, it’s trying out bots built by creators and influencers that can replicate interactions, so that they can spend less time replying to rubbish conversations with fans, and more time creating.

Problem is, these people only have the status that they do because of those fans, and as such, diverting them to an AI chatbot, even one that replies in the style of conversation that they choose, feels like short-changing them. And that seems like a backwards step, because no matter how realistic and convincing AI chatbots get, they won’t be real. So you’re selling a fantasy, in apps that have their foundations in actual human interaction and engagement.

It seems, in Meta’s eyes at least, that authenticity is out, in favor of simulated response, but I don’t think that Meta’s users are going to see things that way.

Though there may actually be a more valuable use of chat AI coming, that could offer real value, not just simulated engagement.

In today’s Q2 earnings call, Meta CEO Mark Zuckerberg discussed a future application of AI chatbots built by individual businesses, that would facilitate faster ordering and customized engagement, powered by increasingly adaptive and intelligent AI tools.

As per Zuckerberg:

“Our goal is to make it easy for every business to pull their content and catalog into an AI agent that drives sales and saves them money. When this is working at scale, I expect it to dramatically accelerate our business messaging revenue.”

So, conceptually at least, Meta’s aiming to develop “Business AIs” which will have access to your entire product catalog and info, and will be able to provide personalized customer service within its apps.

Which, essentially, is what Meta wanted to do with Messenger bots, and which, as noted, nobody actually used.

So, based on precedent, the indicators here are not great, but maybe, with smarter AI technology, and with more people growing accustomed to interacting with AI tools, this could be a good bet, providing Meta with another opportunity to integrate AI into its business offerings.

It’s a better use of AI than simulating human engagement either way, and with the broader adoption of tools like ChatGPT, it does stand to reason that there will increasingly be demand for options like this.

Meta’s also, of course, integrating AI into its ad targeting and creation tools, and there’s clearly benefit there, in building smarter systems based on user responses in its apps.

Meta’s vision is that advertisers will eventually only need to select an ad objective and set a budget, with its systems then able do everything else, including creating images from your catalog, populating copy, initiating ad variations, targeting, etc.

That’s evolving, and it’s already driving results. And again, as this is away from the user view, the benefits here are more tangible, but for the most part, I don’t think that people actually want bots replacing human interactions in the app.

Transactional interactions are different, and that could be an area of opportunity. But it feels like there’s a clear divide between these two use case, and that’s being reflected in user responses.

Because overall, social media apps are exactly that, social, and Meta should heed the lessons of chatbots past to avoid the same missteps.