Meta has come out with its own version of Google’s AI Mode, which it calls…. AI Mode. Carrying on a long tradition of copying features from other tech players – and not bothering to even change the name (e.g., Instagram Stories) – AI Mode will let Facebook users ask natural language questions.
To answer those questions, AI Mode’s active ingredient is Meta AI. And its sources will be all of the conversations that orbit Meta’s social graph, including Facebook Groups and Reels. Meta is hoping that this provides ample knowledge to reliably synthesize answers in a variety of topical areas.
For example, product-based questions, such as research on a new TV, can synthesize answers from A/V enthusiast Groups who have discussed the pros and cons of a given brand or model. The same goes for questions about local bars and restaurants, given the discussion around business Pages and Groups.
Strengthening that library of knowledge, Meta recently launched its “Ask” tab – a Reddit-like forum for posting questions that spark user-generated discussion threads. This is another example of Meta copying successful products across the web and appsphere (but this time it bothered to rename it).
The main point is that this one-two punch of Ask and AI Mode follows a path already validated by Google and other AI search engines. As we recently examined, they increasingly tap into Reddit as a source of authority, given extensive discussions on granular topics. This makes it gold for AI engines.
Traction Isn’t Guaranteed
In the end, it doesn’t matter which elements Meta copied – an accepted practice in the ‘all’s fair’ tech landscape – but rather if AI Mode will be any good. And if so, another question follows closely after that: will people use it? Do people view Facebook as a place to get answers, and will they trust it?
Taking those one at a time, let’s start with whether or not AI Mode will be any good. Meta’s AI engine is fairly robust, so it’s more a question of data and content on Facebook. Here, we’re talking about a corpus of data that has faced scrutiny for its veracity; and has seen content moderation levels pulled back.
Beyond Meta’s trustworthiness with user-generated content, a broader question faces Google, ChatGPT, and others about tapping into sources like Reddit. Positioning blowhards on discussion boards as authoritative sources could backfire, whether it’s Google, OpenAI, Meta, or anyone else doing it.
As for whether or not people will use Meta’s AI Mode, sometimes these types of things are harder to predict. It could strike a chord with users in unexpected ways – such as the ways TikTok surprised us all as a popular local search destination for Gen Z. Or it could suffer from Meta’s trust issues.
That last part looms over almost everything Meta does, given its past misdeeds and the trust issues inherent in its core advertising model. With several AI engines offering to answer questions – including the late arrival of the highly-trusted Apple* – traction isn’t guaranteed. We’ll be watching closely.
*Stay tuned for next week’s feature article on Apple’s potential to alleviate AI’s trust issues, now that it has entered the picture.
