It feels like everything in society is geared towards optimization, whether it's standardized testing or artificial intelligence algorithms: We are taught to know what outcome we want to achieve and then find a path to get there.
Kenneth Stanley, a former OpenAI researcher and co-founder of the new social media platform Maven, has argued for years that this mindset is counterproductive and harmful. Instead of prioritizing goals, he says, we should prioritize serendipity.
“Sometimes we have to get off the path of purpose and down an interesting path to find a stepping stone that leads to what we care about,” Stanley told TechCrunch in a video interview. “Serendipity is the opposite of finding something through purpose.”
The idea of striving for novelty started with a concept in algorithms that Stanley studies called “open-endedness,” a subfield of AI research into systems that “keep making interesting things forever.”
“Open-ended systems are like artificially created systems,” Stanley said, noting that humans, evolution and civilization are all open-ended systems that continue to build themselves up in unexpected ways.
This algorithmic insight turned into Stanley's philosophy of life, which he wrote in 2015 with former PhD student Joel Lehman, Why Greatness Can't Be Planned. The concept caught on, and Stanley became an international voice for the bold idea that we should do things because they're interesting, not because we need to accomplish some goal.
But while leading the Open Ends team at OpenAI in 2022, Stanley said “dissatisfaction began to bubble up” and he had an “epiphany” that made him decide to stop talking about bringing Open Ends to a wider audience and instead start doing something about it.
He asked himself: What if he created a “random luck network” – a system designed to increase the odds of random luck for other people to enjoy?
So he quit his job and set out to create Maven, a social network built on an open-ended, evolving AI algorithm that seeks out new things. When signing up, users choose a few topics to follow (from neuroscience to parenting), and the algorithm shows them posts that align with their interests. Today's social media algorithms also show users what they think they might find interesting, but the difference is that they're optimized to maximize user engagement, often highlighting sensational content to drive more ad impressions and revenue. Maven, on the other hand, doesn't just show users the most popular posts on topics they're interested in; the algorithm shows them posts based on the likelihood that they'll find them appealing.
Perhaps most revolutionary, Maven does away with the current mechanics of social media: there are no likes, upvotes, retweets, follows, or any way to spread content to the masses.
Instead, when a user posts something, an algorithm automatically reads the content and tags it with related interests that will then appear on that page. Users can also turn up the serendipity slider to broaden beyond their stated interests, and the algorithms running the platform will connect users to related interests. So, for example, if you're following conversations about urban planning, Maven might also suggest conversations about public transport.
While you can't follow other users on the platform, you can see and connect with other users who follow topics that interest you.
Maven co-founder and CEO Kenneth Stanley Image courtesy of Kenneth Stanley
In many ways, Maven feels like an antidote to today's social media, where people compete to create sensational content to garner more attention and popularity, and “the objective paradox is on full display.”
“The echo chambers and the toxicity and the amplification of narcissism and personal branding have gotten completely out of control, and people are losing their souls and becoming branded,” Stanley said.
Social media's addictive nature, its harm to the mental health of adolescents and adults, and its power to sow divisions between nations are well documented. Stanley says these are the unintended consequences of ambitious goals and substituting popularity for quality.
“And when it becomes popular, it creates counterproductive incentives, so all sorts of other things happen,” he said.
Stanley said Maven users can report inappropriate content or misinformation when it pops up, and the AI actively monitors for content that is highly inflammatory or offensive “or worse.” He said Maven can't fix the meanness of human nature, but by removing the incentive to share such content, Stanley hopes it can change “the whole collective dynamic of how people behave.”
Some social media companies have tried to counter these incentives in the past. Instagram tested hiding likes in 2019 to limit comparisons and hurt feelings that can come from making content more popular. Twitter predecessor X is also preparing to make likes private, but for less sound reasons. With thinking heavily inspired by Elon Musk, X's goal is to boost engagement by letting people privately like “edgy” content that they normally wouldn't to protect their public image.
Maven isn’t so much interested in connecting users with their audience, but rather with what’s interesting.
Monetization issues
Stanley and co-founders Blas Moros and Jimmy Secreton soft-launched Maven in late January. The platform launched publicly in May with a Wired feature, and Stanley said Maven became a top trend on Product Hunt and garnered thousands of subscribers.
These are still small numbers compared to other new entrants in the social media space: Bluesky, which launched in 2021, has 5.6 million subscribers. Mastodon has 1.8 million active users as of January 2024. Farcaster, a new cryptocurrency-based social protocol that just raised $150 million, has about 350,000 subscribers. These new networks need significant growth to succeed.
The question remains whether Maven can grow its user base without the deeply toxic qualities that we all loathe but still draw us back into the cesspool that is social media.
Stanley told TechCrunch that Maven raised $2 million in a round led by Twitter co-founder Evan Williams in 2023. OpenAI CEO Sam Altman also participated in the round. Stanley said Williams and Altman invested because, like many of us who were drawn to Maven's “out-of-this-world sweet” ethos, they believe the world and the internet need more of this.
And indeed, Maven's idealistic hope of connecting people with interesting ideas is a breath of fresh air, reminiscent of the early 2000s, when the internet was a place for connection and exploration. Testimonials from early users of the platform have been mostly positive and optimistic, as many have come to the platform looking for authentic, serendipitous interactions and freedom from toxicity.
Screenshot of a post by Rebecca Bellan on Maven asking why people came to the platform. Image credit: Rebecca Bellan
But will idealism be enough to attract more institutional investors when Maven wants to grow in the future?
“I think the challenge we'll face going forward is that it's going to be increasingly difficult to raise capital,” Stanley said, noting that investors won't be willing to put millions of dollars into the company unless there's a clear path to a return on their investment.
“Now we need to find the right investors and quickly implement a sustainable business model,” he continued, musing on the idea of a subscription model that would allow Maven to keep its ethos intact.
Of course, there are other ways Maven can make money, including advertising, but that's less appealing to Stanley because it's so tied to virality and sensationalism.
In the future, Maven may also sell the data to companies like OpenAI, which uses large amounts of data to train algorithms. Earlier this month, OpenAI inked a deal to train an AI using data from social media company Reddit. Maven's value proposition from an AI perspective isn't just the content on its platform, but the open-ended algorithms that run on it.
Stanley told TechCrunch that he thinks open-endedness is essential for artificial general intelligence (AGI), a type of AI that aims to match or surpass human capabilities at a wide range of cognitive tasks. Open-endedness is “a very salient aspect of intelligence,” Stanley said. “It's like the creative, curious side of humans.”
“This data is interesting from an AI perspective because it's data about what's interesting,” Stanley said, noting that current AI models lack the intuition to understand what's interesting, what's not, and how that changes over time. But despite the data's potential value for AI, Stanley said Maven has not signed any agreements with any companies to give it access.
The company also said it was not ruling out such a possibility in the future, but would carefully consider the implications of sharing such data.
“For me, that's not the goal,” he said, explaining that he believes it would not be a good thing for neural networks to be completely open-ended, because that could render all human creative endeavor completely meaningless.
“I really wanted to create this worldwide accidental community,” he says, “It's not like I had some side plan to use Maven to create an open-ended AI or anything like that. I wanted to create something for people because I felt like everyone was talking more and more to chatbots and feeling less and less connected to other people, and I was contributing to that as an AI researcher.”
“This idea of a serendipity network made me feel morally good, like I was actually contributing to making people more connected, not less.”