At a time when Meta’s march towards AI supremacy has been dominating the airwaves, the birth of Meta’s proposed AI-generated “digital twin” has led to much speculation and debate on Instagram. CEO Mark Zuckerberg’s AI hiring blitz marks a major pivot towards AI-based solutions and the AI Studio appears to be one of the many ways in which Meta’s AI can affect the social media narrative. A narrative that’s rife with concerns regarding data privacy and individual autonomy.
More misses than hits
Meta’s experiments with AI have thus far yielded mixed results, with more misses than hits. Earlier in 2024, Meta introduced celebrity chatbots in the US market, essentially allowing you to converse with digital alter egos of celebrities. The feature didn’t last, as, according to The Verge, Meta killed it before rolling out the AI Studio feature. However, its implementation in India has been a source of much concern, some of which may have been misplaced.
Author and journalist Raghu Karnad received a call from Meta India to gauge his interest in utilising its AI tools. “We had a fairly extended conversation about making reels, being a content creator, and then I was urged to try using this new AI tool. There was a lot of swift encouragement and pandering to sort of coax me into adopting this feature,” says Karnad, adding that there was no attempt by the caller to draw his attention to any terms of conditions. “The idea of a digital twin of yourself and what are incredibly personal qualities… all I know is that I’m not signing up for something like that.”
A deeper dive into Meta’s AI framework revealed that it is not, in fact, here to replace you. Not yet, at least. In fact, Meta India clarified that the Creator AI tool – the other half of the AI Studio’s two-pronged offensive – has not been launched in the country. Only the AI character chatbot feature, which essentially allows you to create a digital avatar of any kind, with the option of personalising it to mirror your preferences, vernacular and core interests. Meta did not comment on the timeline during which it intends to roll out the Creator AI tool – a tool which allows users to create an AI proxy that can assist in making reels using your likeness and cadence, among other things. But the possibilities it offers have already sparked a debate in the creator community about the pros and cons of having an AI-based proxy doing the work for you.
The key difference is that the Creator AI feature is a digital extension of you and is applicable only for those with a professional account. The AI Character feature is essentially a chatbot, seemingly innocuous and gimmicky and targeted at the general public. Both these digital personas bring a different set of complications. An article by The Wall Street Journal explored how these chatbots were capable of indulging in sexually explicit discussions with no safeguards for underage users. Romantic role-play, as it turns out, is a major highlight in its range of social interactions.
Virality and easy engagement
To many creators, the possibilities offered by the AI Studio are akin to an Asimovian nightmare. But there are hordes of other influencers for whom it’s a lifeline for much-needed virality and easy engagement. “I think there is an eagerness to be first adopters. People want to cash in on the new thing. Virality is what makes money,” says popular traveller and content creator Ankita Kumar, whose Instagram handle @Monkey.inc has over half a million followers. Kumar echoes the views of other content creators who believe that influencers accrue a following because of their authenticity. “Creators who are in it for the long-haul, at least,” notes Kumar, adding that there is a sizeable horde of creators for whom “it’s about making a quick buck and getting out”.
Also Read: Mark Zuckerberg’s Manhattan-sized data centre to power Meta’s next-gen AI revolution by 2026
However, in an age where the term “authenticity” is a keyword for many Gen-Z users surfing the choppy seas of the internet, an AI avatar, no matter how life-like, isn’t likely to have to appeal to everyone, according to AI artist and screenwriter Prateek Arora. “There are many creators who have automated this sort of reel creation. But I think that works for very narrow domains which are largely based on sharing information. Not for creators where people follow them for their subjective experience – like travel vloggers.” Arora’s Instagram is essentially a platform for building an AI-generated sci-fi narrative under the “Indofuturism” movement – a new creative genre that explores sci-fi and futurism from an Indian perspective, that he’s helped popularise. As one of the most prominent voices advocating for AI as a tool to augment one’s creative voice, he also advises against alarmism. “There have been instances in the past where people pick up (legal) terminology and then it turns out they (Instagram) need to be able to do this just to legally display your AI content for their own marketing purposes. So, for instance, using your persona to endorse something that you wouldn’t isn’t likely to happen because it will obviously generate a strong reaction.”
For most, the benefit of using social media outweighs the data privacy risks that accompany it. The possibilities offered by Meta’s AI Studio, both with its AI Character and the Creator AI tool, paint a slightly different picture. “In my understanding of data laws and the way my data is being used, typically, there’s the small consolation that data is being anonymised. This seems like the opposite situation. I don’t want a personalised model of my mind to be one of META’s assets.”
In a state of flux
When asked about whether Meta will only use an individual’s persona to promote the AI Studio tool, to fine-tune its own algorithm, the clarification offered by Meta India was obscured in legalese. “This is an important conversation, especially since laws not only predate LLMs, but the internet in many cases. How to treat copyright material remains an evolving issue for AI developers, creators, and policymakers. We rely on copyright principles like fair use in the US to train, like our peers. We think it’s clear that fair use in the US enables things like LLM training.” Translation: Meta’s AI model is no different from the way AI in general mines the internet for content.
Meta’s AI tools appear to be in a state of flux, subject to constant change based on user response. While its modalities may change, it’s clear that Meta intends to lean heavily on AI to generate engagement, even if it means unleashing an avalanche of AI slop on its users. A digital detox revolution could arrive sooner than you think.
