\nThe tech behind Artifact, the newly launched news aggregator from Instagram\u2019s co-founders<\/br>
\n2023-03-08 23:00:35<\/br><\/p>\n
Late last month, Artifact<\/a>, a personalized news reader <\/a>built by Instagram\u2019s co-founders, opened<\/a> to the public. The launch was a surprise to many consumers, who wondered why the team behind one of the world\u2019s most iconic social apps would return to startups to focus on one of the toughest areas instead: news. It\u2019s an ecosystem where publishers are failing left and right and misinformation is rampant, as the founders surely saw themselves while working at Facebook.<\/p>\n In an interview, Instagram co-founder Kevin Systrom explains what drew him to this space and how his new app\u2019s underlying technology will serve to differentiate it from the competition.<\/p>\n The startup\u2019s existence, in some ways, can be credited to the COVID pandemic, as that\u2019s what brought the Instagram co-founders, Systrom and Mike Krieger, back together. During the early days of the pandemic in 2020, the co-founders teamed up for the first time since leaving Facebook to create a COVID-19 dashboard called Rt.live<\/a> that tracked the spread of the virus across the U.S.<\/p>\n But in later months, other COVID trackers emerged and people were no longer as interested in tracking the virus\u2019s spread on a state-by-state basis. Invigorated by the experience, the founders began then to think about returning to work on a \u201creal company\u201d instead of a public good side project.<\/p>\n Image Credits:<\/strong> Usis \/ Getty Images<\/p>\n<\/div>\n Deciding on their next act took time. The founders hacked around on stuff for a year and a half or so after their initial collaboration, Systrom says. The founder, who describes himself as a \u201cvery frameworks-driven person,\u201d knew he wanted to do something that involved machine learning, having seen its power at Instagram.<\/p>\n \u201cIt was super fascinating to me to work on it at Instagram and watch it go from a company without machine learning to a company driven by machine learning,\u201d he says. \u201cI also witnessed the rise of TikTok and realized how important machine learning would be for the future of social.\u201d<\/p>\n He himself observed how social networking\u2019s underlying graph had changed a lot over the years, watching as Facebook invented what\u2019s now known as the \u201cfriend graph\u201d \u2014 a user\u2019s personal social network of real-life connections. Later, he saw Twitter pioneer the \u201cfollow graph,\u201d or a graph of connections based on the user\u2019s explicit choices of who they want to follow on a service. Then, at Instagram, Systrom saw firsthand the shift from the \u201cfollow graph\u201d to the \u201cinferred graph\u201d or, rather, the \u201cinterest graph.\u201d<\/p>\n This, he explains, was basically a \u201cfollow graph\u201d powered by machine learning, instead of by users clicking a button.<\/p>\n The potential to leverage machine learning and an interest graph within a new product appealed to him, he says.<\/p>\n \u201cWe looked for an area that was social in nature, but where we could apply 20% new techniques \u2014 and that would be the machine learning side of what we\u2019re doing,\u201d Systrom says, describing how the founders narrowed their focus.<\/p>\n Image Credits:<\/strong> jayk7 (opens in a new window)<\/span><\/a> \/ Getty Images<\/p>\n<\/div>\n The other consideration that prompted the founders\u2019 interest in news, more specifically, was the potential impact if the app succeeded. They wanted to work on a startup they felt the world needed, Systrom says. No doubt, some of that interest may have been fueled by working at Facebook (now Meta), which had changed consumers\u2019 news consumption behavior, impacting publishers as well as the spread of misinformation.<\/p>\n \u201cIt felt like our consumption of information \u2014 both factual, balanced, entertaining, etc. \u2014 had an existential crisis,\u201d Systrom tells TechCrunch. Plus, he adds, \u201cmany of the people producing this content are going out of business.\u201d<\/p>\n Meanwhile, on the consumer side of the news reading experience, there\u2019s so much information swirling around that people don\u2019t know what they can trust or which item to read. People are asking themselves if a link shared by a friend is actually legit and they\u2019re wondering why they\u2019re reading one article over the many others published on the same topic.<\/p>\n \u201cIt might be cheesy to say, because I\u2019ve now said it a bunch of times, but I feel like the worst part about social media is that it\u2019s social,\u201d Systrom says. \u201cI think the \u2018social\u2019 part of social media \u2014 for a long time, in terms of information consumption \u2014 has been a hack to filter for information that would be interesting to you. But we now don\u2019t need that hack, because we can learn what\u2019s interesting to you,\u201d he continues. \u201cWe can quantify it. We can build profiles. And then we can serve you content that is both high-quality, balanced and interesting to you.\u201d<\/p>\n This realization led to the creation of Artifact, a social news app powered by machine learning.<\/p>\n Image Credits:<\/strong> Artifact personalization and stats<\/p>\n<\/div>\n The app in some ways is very much like others that exist today, which have been founded in other countries, including ByteDance\u2019s Toutiao<\/a> in China, Japan\u2019s SmartNews<\/a> and News Break, another personalized news reader with Chinese roots<\/a>. Like its rivals, Artifact learns from user behavior, engagement and other factors in order to personalize which headlines are presented and in which order.<\/p>\n Despite this competitive landscape, Systrom believes U.S. news consumers want an option that\u2019s actually based in their home market.<\/p>\n \u201cThey want a domestic player with a team they trust,\u201d Systrom says, speaking to the news aggregation landscape in the U.S.<\/p>\n His comment is a timely one, given how tensions in the U.S. have been growing over China\u2019s grip on the social networking market with TikTok. The short video app, which is often used as a source for news by Gen Z, is now banned on government phones and bipartisan legislation to further police it is in the works<\/a>.<\/p>\n In addition to its locale, Systrom believes Artifact will be able to differentiate itself based on its unique combination of technology and taste \u2014 a directive that could also be used to describe Instagram\u2019s founding for that matter.<\/p>\n Unlike Facebook \u2014 which became a platform by which any publisher could deliver news, and oftentimes clickbait \u2014 Artifact\u2019s news sources are curated up front, the founder explains.<\/p>\n \u201cWhen I say taste, I mean the top of the funnel in our system \u2014 the publishers we choose to distribute,\u201d notes Systrom. \u201cIt\u2019s not a free-for-all. We don\u2019t crawl the entire web and just let everything go in.\u201d<\/p>\n Instead, Artifact has selected the top publishers across different categories to fuel the content in the app. Customers can add their own paid subscriptions, as well. At this time, Artifact doesn\u2019t sell those for a revenue share or involve itself in publishers\u2019 ad sales, though one day that could change, depending on how the app chooses to monetize.<\/p>\n For now, however, the focus is on gaining traction with consumers and ensuring the app\u2019s news sources are worth reading.<\/p>\n Screenshot of Artifact app. Image Source:<\/strong> Artifact<\/p>\n<\/div>\n \u201cThe line, internally\u2026is we want a balanced ideological corpus, subject to integrity and quality,\u201d Systrom says. \u201cAnd the idea is not that we only choose left-wing, or we only choose right-wing. We drew the line at quality and integrity subject to a bunch of the metrics that a lot of these third-party fact-checking services have. The third-party services basically rate the integrity of different publishers based on their research and based on public events \u2014 like how quickly they correct their stories, whether their funding is transparent, all that kind of stuff,\u201d he notes.<\/p>\n \u201cYou can claim it\u2019s editorial,\u201d he continues. \u201cBut it\u2019s more about making sure that the set that we\u2019re going to distribute and pour into this powerful machine \u2014 that can distribute content widely \u2014 that we\u2019re being responsible by giving it content that isn\u2019t going to be misinformation.\u201d<\/p>\n Beyond the integrity of reporting, Artifact aims to deliver a news experience that\u2019s more fulfilling.<\/p>\n \u201cIf you log on to a lot of these other sources, you get pretty clickbaity-stuff,\u201d Systrom points out. \u201cI\u2019m not trying to throw shade on folks working in this area, but we wouldn\u2019t work on it if we thought that it was solved. We focus a lot in our system on a different objective, which is this idea of value\u2026if we put in the top-quality sources in these categories and we have coverage \u2014 whether it\u2019s parenting or mental wellness, or exercise \u2014 that we can create a differentiated content set that feels much different than the front page of a major newspaper but also feels very different than a lot of the other aggregators,\u201d he says.<\/p>\n Yet, even as the app personalizes its content selection to the end user, it doesn\u2019t leave them in so-called \u201cfilter bubbles,\u201d necessarily, as Facebook did. Instead, when users click on a headline to read a story, they\u2019re shown the entire coverage across sources, allowing them to peruse the story from different vantage points.<\/p>\n Image Credits:<\/strong> Artifact<\/p>\n<\/div>\n Systrom credits Toutiao for driving innovation in recommendation systems, noting that Toutiao essentially helped ByteDance give birth to TikTok. The technology developed for its news discovery was brought to its acquisition of Musical.ly, which became the Chinese app Douyin and its international counterpart TikTok.<\/p>\n But Systrom believes some of the machine learning that Artificat is doing is different.<\/p>\n \u201cThe machine learning that a lot of what we\u2019re doing is based on was invented in 2017 at Google. It\u2019s called the transformer\u2026without that, GPT 3, 3.5 etc., wouldn\u2019t exist. Without that, you wouldn\u2019t have DALL-E. Without that, you wouldn\u2019t have ChatGPT,\u201d he says. \u201cYou\u2019re inventing a core technology that can then be applied in many different ways.\u201d<\/p>\n The app\u2019s algorithms are focused on more than just tracking clicks and engagement. It weighs other factors, too, like dwell time, read time, shares, stories that get shared in DMs (private messages) and more.<\/p>\n \u201cIf you let your algorithm focus on clicks, it will end up serving clickbait. If you simply optimize for only what people have clicked on, you end up having tunnel vision,\u201d Systrom explains.<\/p>\n The model additionally involves an algorithm called Epsilon-Greedy<\/a>.<\/p>\n \u201cThat\u2019s a technical term for you taking some small portion of your time \u2014 like 10% or 20% \u2014 and you explore. You do something different than you would normally do\u2026you go to the outside of the recommendation spectrum rather than the core of it,\u201d Systrom says. \u201cIt\u2019s shown many, many times, especially in reinforcement learning, that having this \u2018explore\u2019 budget we\u2019ll call it, actually is optimal for users.\u201d<\/p>\n That\u2019s the same reason why TikTok has begun testing tools that let users refresh their feeds<\/a>. Without the added spice of unexpected content, the video app\u2019s suggestions had grown stale for some users.<\/p>\n But because the news changes every day, Artifact\u2019s use of this Epsilon-Greedy algorithm also has to adapt as users\u2019 own interests in topics grow and wane. For example, someone might be very into reading about the upcoming elections up until Election Day has passed. Or a new story may immediately capture their attention when it comes out of nowhere, as the story about the Chinese spy balloon did.<\/p>\n Artifact\u2019s editorial consideration, if you can call it that, will also come into how the app\u2019s algorithm is programmed, not just its publisher selections.<\/p>\n \u201cActually, building the algorithm is enormously editorial,\u201d Systrom says. \u201cBecause what you choose to train your algorithm on \u2014 the objective function, the data you put in, the data you include, the data you don\u2019t include \u2014 is all in editorial judgment. The way you weight different objectives.\u201d<\/p>\n \u201cAll of the edge a company has is in its ability to wrangle the data in such a way that produces an outcome that\u2019s optimal,\u201d Systrom says.<\/p>\n If anything is Artifact\u2019s secret sauce, it\u2019s that.<\/p>\n Whether all this tech and \u201ctaste\u201d is enough for Artifact to succeed remains to be seen. The U.S. news aggregation landscape is not like China\u2019s. And here, Artifact will have to fight against the default ways people access news, including through the news apps that come bundled with their smartphones, as well as the large internet portals like Google and Facebook, and some extent, YouTube and TikTok too.<\/p>\n At launch, Artifact added new functionality, including a new feature that allows users to track how they\u2019ve been engaging with the app and its content in a metrics section, which shows a list of publishers and topics they\u2019ve been reading. Over time, Artifact plans to let users adjust which topics they want to see more and less of, or even block publishers.<\/p>\n They\u2019ll also later be able to socialize through comment threads around the stories themselves, in what could be a stealth competitor to Twitter \u2014 an app that\u2019s seen a small exodus in the wake of Elon Musk\u2019s takeover. (In fact, Artifact just added a feature similar to Twitter\u2019s<\/a> with its ability to show you articles trending in your network, if you sync your contacts to the service.)<\/p>\n As for Artifact\u2019s future, Systrom says he hopes it will eventually become a place where users can go, first and foremost, to discover content around their interests and be able to discuss them with others.<\/p>\nMachine learning is the \u201cfuture of social\u201d<\/h3>\n
![\"AI](\"https:\/\/i0.wp.com\/techcrunch.com\/wp-content\/uploads\/2021\/09\/GettyImages-1136663830.jpg?resize=680%2C453&ssl=1\")
<\/p>\nFixing a broken news ecosystem<\/h3>\n
![\"\"](\"https:\/\/i0.wp.com\/techcrunch.com\/wp-content\/uploads\/2023\/03\/newspaper-stack_GettyImages648942926.jpeg?resize=680%2C453&ssl=1\")
<\/p>\n![\"\"](\"https:\/\/i0.wp.com\/techcrunch.com\/wp-content\/uploads\/2023\/02\/artifact.jpg?resize=1140%2C817&ssl=1\")
<\/p>\nClearing out the clickbait<\/h3>\n
![\"\"](\"https:\/\/i0.wp.com\/techcrunch.com\/wp-content\/uploads\/2023\/03\/IMG_5041.png?resize=314%2C680&ssl=1\")
<\/p>\nArtifact\u2019s recommendation system<\/h3>\n
![\"Artifact](\"https:\/\/i0.wp.com\/techcrunch.com\/wp-content\/uploads\/2023\/01\/Product_Shot.jpeg?resize=680%2C453&ssl=1\")
<\/p>\nDoes Artifact have a future?<\/h3>\n