Everything in society, from standardized tests to AI programs, can feel like it’s trying to be the best it can be. It’s taught that you should know what you want and then figure out how to get there.
Ken Stanley, a former researcher at OpenAI and co-founder of Maven, a new social media platform, has been saying for years that this way of thinking is not only useless but also damaging. Stanley says that instead of putting goals first, we should put chance first.
Stanley told TechCrunch in a video interview, “Sometimes we have to get off the path of the objective and onto the path of the interesting in order to find those stepping stones that will lead to the things we care about.” “Serendipity is the opposite of setting goals to find something.”
The idea of wanting new things just for the sake of it comes from Stanley’s research into open-endedness, a branch of AI that looks into systems that “just keep making interesting stuff forever.”
Stanley said that open-ended systems are like systems that are artificially creative. He went on to say that people, evolution, and society are all examples of open-ended systems that keep building on themselves in strange ways.
Stanley turned this algorithmic understanding into a way of life. He even wrote a book called Why Greatness Cannot Be Planned about it with Joel Lehman, who used to be his PhD student. Stanley became famous around the world for the bold idea that you can do things just because they’re fun, not because you have to meet a set goal.
But Stanley said he was “boiling over with discontent” in 2022, when he was in charge of an open-endedness team at OpenAI. He “had this epiphany” where he chose to stop talking about how to make open-endedness more accessible and start doing something about it.
He thought to himself, “What if I made a “serendipity network” for other people to enjoy? A “serendipity network” is a method that is set up to make serendipity more likely.
He quit his job and started making Maven, a social network based on an AI program that can change over time to look for new things. Users choose a list of topics to follow when they sign up, such as parenting or neuroscience. The algorithm then shows them posts that are related to those subjects. There are no follows, likes, upvotes, or shares, and there’s no way to get a lot of people to see something.
Instead, when a user posts something, the algorithm reads it and tags it with interests that are important to that page, so it shows up there. As users move the serendipity slider, they can find things that are connected to their stated interests. The algorithm that runs the platform then pairs users with similar interests. Like, if you’re watching conversations about city planning, Maven might also suggest conversations about public transportation.
You can’t follow other people on the site, but you can find and talk to people who are interested in the same things you are.
Stanley says are the unintended results of setting high goals and using fame as a measure of quality.
There are many ways that Maven is different from today’s social media, where the “objective paradox is on full display” as people rush to post sensational content that will get more likes and attention.
Stanley said, “The toxicity, the echo chambers, the narcissism have all gotten out of hand, and personal branding has become so popular that people are losing their souls and becoming brands.”
It is well known that social media can be addicting, hurt teens’ and adults’ mental health, and cause divisions between countries. Stanley says these are the unintended results of setting high goals and using fame as a measure of quality.
He added, “And then you get all these other things because once you have popularity, you have strange incentives.”
Stanley said that Maven users can report hate speech or false information when they see it, and the software’s artificial intelligence is constantly looking for content that is very insulting, harmful, or both. He said that Maven can’t change how mean people are, but Stanley hopes that by taking away the reasons for sharing such material, it will change the “overall aggregate dynamic of how people are behaving.”
In the past, some social media companies have tried to stop these kinds of rewards. StumbleUpon, a browser plugin and app made by entrepreneur Garrett Camp years before he co-founded Uber, was the first site to share random content. Then, in 2019, Instagram tried hiding “likes” to stop people from comparing posts and getting upset feelings when they see how popular something is. X, which used to be Twitter, is also planning to make likes private, but for less good reasons. X wants to get more people to interact with his app by letting them privately like “edgy” material that they wouldn’t normally do to protect their public image. This is a very Elon Musk-like way of thinking.
It’s not as important for Maven to connect users with viewers as it is to connect them with interesting things.
The Issue Of Making Money
Stanley and Blas Moros and Jimmy Secretan, who helped him start Maven, gave it a soft launch at the end of January. The platform went live in May with a Wired feature that Stanley says made Maven a top popular item on Product Hunt and led to thousands of sign-ups.
When you look at other newcomers to social media, those numbers are still pretty small. Since it began in 2021, Bluesky has had 5.6 million sign ups. Mastodon had 1.8 million daily users as of January 2024. About 350,000 people have signed up for Farcaster, a new social network based on cryptocurrency that just raised $150 million. All of these new networks will need to get a lot bigger before they can be called successful.
There is still no clear answer to the question of whether Maven will even be able to gain more users without the very things we hate about social media that make us want to use it again.
Stanley told TechCrunch that Twitter co-founder Ev Williams led a round of funding for Maven in 2023 that raised $2 million. Sam Altman, CEO of OpenAI, also took part in the round. Williams and Altman put money into Maven because, like many of us who like Maven’s almost too-sweet-for-this-world attitude, they think the world and the internet need something like this.
In fact, Maven’s lofty goal of introducing people to interesting ideas makes me think of the early 2000s, when the internet was a place to meet new people and learn about new things. Early users’ feelings about the platform are mostly positive and hopeful. This is because many of them came to it looking for real, chance exchanges and the promise of being free from toxic people.
Will idealism, on the other hand, be enough to get more institutional backers when Maven wants to grow?
Stanley said, “I think the challenge we face is that going forward, that becomes a harder and harder way to raise money.” He added that investors won’t put down millions of dollars unless they can see a clear way to get their money back.
He went on, “I just need to find the right investors going forward and quickly get to a sustainable business model.” He was thinking about the idea of a subscription model that would let Maven stay true to its values.
Maven can make money in other ways as well, of course. Advertising is one way, but Stanley is less interested in it because it is linked to going viral and making news.
Over time, Maven might also be able to sell its data to businesses like OpenAI that use huge amounts of data to train their algorithms. Reddit and OpenAI made a deal earlier this month so that OpenAI can teach its AI on Reddit’s data. From an AI point of view, Maven’s value isn’t just the material on the platform; it’s also the open-ended algorithm that runs it.
Stanley told TechCrunch that he thinks open-endedness is important for artificial general intelligence (AGI), which is a type of AI that aims to be as smart as humans or smarter in many thinking tasks. Stanley said that being open-ended is “such an important part of being smart.” “It’s kind of like this creative and curious part of being human.”
Stanley said, “The data is interesting from an AI point of view because it’s data about what is interesting.” He added that current AI models don’t have a natural sense of what is interesting and what isn’t, or how that can change over time. In spite of the fact that AI might find the data useful, Stanley said Maven does not have a deal with any company to share that data.
He also said that he hasn’t ruled out the chance of doing that in the future, but he would give it a lot of thought before doing it.
He said, “That’s not the point of this for me.” He also said that he wasn’t sure if it would be good for neural networks to be completely open-ended, since that could make all creative work by people pointless.
He said, “I really wanted to make this worldwide serendipitous community.” “It’s not like I have a secret plan for how we’re going to use Maven to make open-source AI or something.” I just wanted to make something for people because I thought that as time goes on, more and more people would talk to robots and less and less people would connect with each other. And as a specialist into AI, I helped with that.
Also Read: A Pew Poll Found That 78% of Americans Think Social Media Companies Have Too Much Political Power
“The thought of a serendipity network made me feel better about my actions; it made me feel like I could help connect people more, not less.”
What do you say about this story? Visit Parhlo World For more.
