There is a lot happening in tech this week. Australia has a proposed Digital Duty of Care regime on the table, Apple has unveiled its latest iPhone announcements and Instagram is asking users to look a little more closely at the invisible force deciding which videos appear as soon as they open the app.
On Thursday, PEDESTRIAN.TV joined an Australian media roundtable with Instagram head Adam Mosseri as the platform launched the local edition of its Algo Confessions series. The rollout features Josh Heuston, Kayla Itsines, Bridey Drake and Ruel talking through the niche interests that have made their way into their feeds.
It is also a push for Instagram’s Your Algorithm feature. The tool shows users a summary of the topics Instagram thinks shape their recommendations, then lets them add interests, remove them, or ask to see more or less of particular things. Those changes apply across Feed, Reels and Explore.
If the app has decided you are suddenly devoted to footy, vintage fashion or cursed memes, you can look at that assumption and correct it.
Mosseri began by trying to demystify the thing most of us call “the algorithm”.
“There isn’t one algorithm,” he said. Instagram uses different ranking systems for Feed, Reels, Stories, suggested message recipients and even the list of people who have liked a post.
Across the major feeds, the platform looks at what people like, comment on, share, watch, watch to the end and tap through to. It then predicts what someone might do with another post, giving content a relevance score and ordering the feed accordingly.
“It’s a guess,” Mosseri said. “It’s an educated guess, it’s an expensive guess, but it’s a guess.”
The problem with an educated guess is that it can turn one moment of curiosity, concern or grief into a recurring theme. Mosseri gave the example of a person who experiences pregnancy loss and then finds their feed full of baby or nesting content, something he said can become “incredibly triggering”.
In the past, he said, they may have had to “do a hard reset and reset the whole thing”. Your Algorithm is intended to let them remove that one topic instead. “This allows you to get in there with something closer to a scalpel,” Mosseri said.
But a more finely adjustable feed also brings up the obvious echo chamber question. Could putting more preference controls in users’ hands make it easier to settle into a feed that reflects only what they already believe?
Mosseri said there were “two forces in opposite directions”. People tend to follow and interact with like-minded people, publishers, brands and companies, he said, while the scale of content on Instagram can push people towards a more diverse set of material. “What we’ve seen is, and when we have looked in the past, that those roughly cancel each other out,” he said.
He argued that these echo chambers do not begin and end online. “We tend to have the same politics as our parents, we tend to be friends with the people with the same politics that we do, we tend to read content from newspapers that we agree with,” Mosseri said.
Referring to the US, he said the historical data he had seen showed an “inverse correlation between age — and therefore technology use — and polarisation”, with older people more polarised despite using less social media. His theory was that many were “just watching cable news”, including overtly partisan networks. “The media diet in the US is something like five hours a day on TV, so it’s big,” he said.
Instagram’s responsibility, in Mosseri’s telling, is not to intentionally reinforce or challenge someone’s worldview. “We don’t feel like it’s our job to try and show you stuff you agree with or you disagree with,” he said. “We think it’s our job to try to show you stuff that you’re interested in that you would say is worth your time.”
That is a fairly restrained description of Instagram’s responsibility. It also puts a lot of weight on a loaded phrase: “worth your time”. Worth it according to whom? The user? A model trained on their past behaviour? Or a platform whose business depends, in part, on keeping people engaged enough to see advertising?
Mosseri said Instagram does not optimise solely for time spent, arguing that a focus on long videos can harm other parts of the service. Instead, he said the company looks at measures including sharing, engagement, user sentiment, app preference and revenue. “We want to be what people say is their favourite app to use,” he said.
Those comments sit uncomfortably alongside the wider legal scrutiny facing Meta, which owns Instagram, Facebook and WhatsApp. In March, a Los Angeles jury found Meta and YouTube owner Google liable in a landmark case involving a young woman known as Kaley, who said her childhood use of the platforms harmed her mental health. The jury awarded $6 million in damages and found the companies had intentionally built addictive platforms; Meta and Google said they disagreed with the verdict and intended to appeal.
Mosseri had given evidence in that case. When asked about the plaintiff spending 16 hours on Instagram in one day, he called it “problematic use”, but rejected the characterisation of addiction. “It’s important to differentiate between clinical addiction and problematic use,” he told the court, per the BBC.
He gave a similar vibe during Thursday’s roundtable. “I don’t believe Instagram is addictive,” Mosseri said, arguing that clinical addiction involves a specific pattern of behaviours and physical withdrawal symptoms.
“None of the science I have seen suggests that social media, or any media for that matter, is actually clinically addictive.”
Still, when asked what Instagram is doing in its ranking systems for people who feel they use the platform more than they want to, he said there was no specific intervention. “I don’t think we know,” he said. The company’s response, he added, is to “try to show you things that you would say are worth your time”.
Your Algorithm is a useful correction tool. It can give a new or low-engagement user a way to tell Instagram what it has missed, and it can offer a more precise response when a topic becomes unwanted or distressing. But Mosseri acknowledged that for people who have used the app for years, manually selecting interests may be “much less likely” to add much to what Instagram already thinks it knows.
The feature gives users more of a say, but it certainly does not give them the whole system. As Australia considers laws that would let people opt out of algorithmic recommendations altogether, that distinction feels harder to ignore.
The post Instagram Says It Wants To Give You More Control Of Your Algorithm. Is It Enough? appeared first on PEDESTRIAN.TV .
