Discover more from Wysr by Cameron Armstrong
SayMore Essay Contest Guest Posts
Note: My startup SayMore is running an essay contest, funded by Purple, in conjunction with FarCon (The Beta) - an unconference I’m hosting for Farcasters. Some folks participating in the contest don’t have a place to host their entries so I’m hosting them here as guest essays on Wysr. All credit goes to the original author and leave a like if you enjoy their work!
Social Media is Dead. Long Live Social Media.
Social Media is Dead, Long Live Social Media.
Like a giant headless cockroach, social media shuffles onward, completely unaware that it is already dead. The decapitation process so swift and sudden that although everyone has seen the blade fall and the head leave the body, few are thinking forward to the coming inevitable end.
When new technologies appear on the scene, there is a scramble to figure out who benefit and who are the walking dead. For social media, the coming end will trace it beginnings to AI. The spark of chatGPT showed us all a small flash of the future, a future that has no place for social media.
Currently billions of users check multiple social media platforms every day. To talk with friends, to read the news, to discover content, to be entertained. Some of them right now may be talking about AI: the potential, the excitement, and the fear of what AI means for the future.
But few if any have given thought to a future which is equal parts obvious, simple, interesting, terrifying, complex, and exciting. I am speaking of AIs with a singular purpose, a singular mission: TO ENTERTAIN US. Not us in the collective sense, each of us, individually and directly.
Social media for news and entertainment is dead. Social media for discovery of content is dead. In a few years, your personal AI will be able to sift through the entire internet of content and hand pick for you the most interesting articles while you brush your teeth. Create a custom song for your morning run. Generate a tailored murder mystery for your movie night with exactly the right level of suspense and violence for your tastes. It will write you a personalized fantasy story for you to read before bed.
People will look back and wonder what they did all day before personalized AI content. Spending time on social media with its endless scrolling for the occasional short lived chuckle will seem quaint and outdated. Like opening up an old fashion newspaper.
But even as it takes its last dying breath, social media will rise again, more powerful and important than ever before. For while humans crave entertainment, they also crave connection. The connection of family, friends, acquaintances. As AI personalizes our media for us, it will also isolate us. No more will we be able to share our favorite TV shows, books, and movies with others, they will not have seen the unique creations of our personalized entertainment AIs. A dark void of loneliness will open up that humans will desperately try to fill. It is this void where social media will again find its place, not as a content generating machine, but as a connection generating machine.
Like a phoenix rising from the ashes, social media will rise again reborn. A little less media and a little more social. Its new purpose: to find and connect humans together. New technologies for identity like cryptographic proofs and nfts will be key for this change. The new social media will use them to weed out what is not human and to help us connect with our personal tribe. Social media will help us find those meaningful relationships that cannot be replaced with personalized AI. The companies that align with this new purpose will experience growth and power like never before, while those that chase after the mirage of social media like it once was will fade away. For social media is dead. Long live social media.
Finding people and conversations on Farcaster
Every time I move, I'm beginning the search anew: Where does my tribe meet? I scroll through meetups, local facebook groups, and location channels on Slack and Discord. I'm attending co-working spaces and networking events, annoyed that physical social networks are tied to locations.
The appeal of digital social networks is that, no matter where I am, my digital friends are a couple of tabs away. I can even travel to meet them IRL. But where can I find them? The online world is a vast space of content and people.
Algorithms! Cut through the noise with the help of technology. We do this IRL, using our own algorithm i.e., intuitive and wisdom. We curate the place we want to go to increase the chances of meeting people we want to hang out with. It’s the same Spiel online.
Kairon talks about bringing your own algorithm (BYOA). The idea stuck with me. Sometimes I get really bored by my feed. It’s not just the performative posts, it’s also the echo-chamber. Everything is bland and sounds the same. The salt is missing. At those moments I wished I could create my own feed, surfacing new posts based on my own criteria.
This is possible with decentralized social graphs, where the graph is open to anyone to build on. I’m already accessing Farcaster using 3 different apps, depending on the situation.
Here are some new ways to highlight and curate users’ feeds. They are based on the assumption that people join social networks to learn from others (I’m going to forget that some really love to talk about themselves…).
The metrics are taken from social network analysis (graph theory). An important consequence is that without clear membership criteria, the metrics become useless. How membership is decided depends on the developer or user. It could be based on NFT, location, topic.
Some metrics are calculated using an iterative approach. Thus, the value changes based on the size of the overall network. They are the more interesting metrics, but also harder to compute. Theoretically, they should be updated whenever a caster enters or leaves the network.
I’ll explain the purpose of the algorithm and link a paper or repo to make it more concrete.
Direct neighborhood: Show to user A only her 1st degree connection (“Following”). Of this Following-list, you could boost those people whose followers are also connected to user A (indicated as purple circles in the Figure below). Enjoy the echo-chamber.
Not my preferred algorithm. At least not in this raw format. I’d tweak this a bit to filter the 1st connection by content. I want to be inside an echo-chamber when I need to dive deep into a topic I’m (not) familiar with. Another scenario where this algorithm could be good, is when someone feels lonely or shit. But for that case, the default setting of 1st degree connection should be ok.
Filtering by content can be done through topics or self-reported expertise when creating an account. More interesting would be to determine expertise or topics of interest through semantic analysis on the posts. The longer a cast, the more accurate the results. The advantage of attributing expertise to casters based on their casts is that changes in interests will be picked up automatically. The resulting interest fields can also be used to create groups (sub communities). A sequence of methods is helpful here: Latent Ditricht Allocation (LDA) followed by clustering in large bi-partite networks.
LDA is a text analysis method to connect topics (e.g. keywords) to people. This results in a n*m matrix (topics * Casters). That matrix can be used for bipartite clustering or Topic Affinity Propagation.
Show me new stuff
Other world: Create a feed for caster A with casts from unique casters, those casters whose connections are unique to caster A. By using distinctiveness centrality, casters who are different will be highlighted. The metric calculates how unique someone is based on their structural position in the network. The network could be a group (e.g., based on NFT), or all Casters.
Another way to show new content is to create a feed with casts from 2nd degree connections (the followers of your followers). Casters in this list could be boosted using activity, topics, or some other criteria (e.g., location).
Curators: A feed with users who are in a great position to curate content across different corners of Farcaster. Using betweenness centrality, the feed will include Casters who are optimally placed in the network to pass information between people.
Is Farcaster sufficiently decentralized?
This isn’t an algorithm for a user, but a curiosity-driven question if we can measure if Farcaster is sufficiently decentralized. I’m suggesting that we calculate the network's Small World index. The Small World metric is a way to measure if sub-communities have a positive or negative impact on a network. An oversimplification of the Small World metric is (Kevin Bacon’s 6 Degrees of Separation). In a world, where sub-communities are sufficiently connected with bridges, you can quickly “travel” from one sub-community to another. The result: You are connected with any other person through, on average, 6 steps.
The small world metric is an improvement to the Gini metric shown on the Farcaster.network, however, it’s more computational intensive to calculate.
The following essay was inspired by the contest prompt, but as it falls outside the scope of the submission guidelines, it will be showcased here along with the other essays, but is not eligible for any prize money funded by Purple DAO.
The Cosmic Blockchain Theory: A Deep Dive into the Manifestation of Collective Consciousness and Reality
Author: Derek Hall
The Cosmic Blockchain Theory proposes that our universe is an actual blockchain. This blockchain is upheld not by lines of code or digital transactions, but by something far more profound and ubiquitous – the collective consciousness. We are all nodes being powered by our body’s digestive system and communicated with electromagnetic impulses through our brains. This concept encapsulates our shared beliefs, memories, and intentions, offering a unique perspective that our collective psyche might have the power to influence and shape reality. This compelling proposition is seemingly echoed by the enigmatic phenomena of the Mandela Effect, the power of prayer, and the Law of Attraction, each offering intriguing insights into our collective reality.
The Mandela Effect is a captivating psychological phenomenon where large groups of people distinctly remember events differently from their official recorded versions. It poses a baffling question: Why do collective memories deviate from documented history? Consider this phenomenon through the lens of the Cosmic Blockchain Theory. The parallels become evident. In a blockchain, when more than 50% of the nodes in the network agree on a transaction, it is validated and is therefore true. Similarly, the Mandela Effect could be perceived as a form of 'consensus' within our collective consciousness. When enough individuals 'agree' upon a certain memory or belief, does it alter the 'block' of reality within our collective memory? This would imply that our collective memory and belief have the potential to revise the reality we inhabit. Such an understanding offers compelling support to our cosmic blockchain theory, bringing us closer to deciphering the enigma of the Mandela Effect.
Echoing this perspective is the principle found in Matthew 18:19-20 of the Bible: "Again, truly I tell you that if two of you on earth agree about anything they ask for, it will be done for them by my Father in heaven. For where two or three gather in my name, there am I with them." This verse subtly implies the power of agreement and consensus, a cornerstone of the blockchain technology, reinforcing the significance of collective agreement in shaping our reality.
The power of prayer, particularly collective prayer, finds resonance with this perspective. Collective prayer is the shared intention expressed by a community. It's a potent force that, throughout history, has been believed to bring about change, often in times of crisis or despair. Now, consider this collective prayer as a consensus mechanism in our cosmic blockchain. If the intent and belief behind a collective prayer are hold the majority, could it generate a consensus that amends the cosmic blockchain of our reality? This hypothesis is in line with Acts 2:44-46 of the Bible: "All the believers were together and had everything in common. They sold property and possessions to give to anyone who had need. Every day they continued to meet together in the temple courts. They broke bread in their homes and ate together with glad and sincere hearts..." This passage attests to the power of a unified belief system shaping a shared reality, reinforcing the postulates of our cosmic blockchain theory.
Lastly, consider the Law of Attraction, a philosophical concept suggesting that individuals can influence their personal reality through focused thoughts and intentions. If each person can be seen as influencing their individual 'block' within the cosmic blockchain, then the Law of Attraction gains a new layer of significance. Numerous anecdotes suggest that people have successfully manifested their desires using this law. Could these be instances where individuals have managed to add a new 'block' to their personal chain within the cosmic blockchain? If so, this lends credence to the theory that the cosmos functions like a blockchain, where individual and collective consciousness can manifest into reality.
As we delve deeper into understanding these phenomena, we inch closer to a unified theory of consciousness and reality, one 'block' at a time. Yet, several intriguing questions remain: How much influence do we truly have over our individual and collective realities? To what extent does consensus in belief impact the fabric of our lived experience? Does each species, in fact, represent a unique blockchain, contributing to the richness of our shared cosmic ledger? These are the conundrums for us to ponder and, possibly, to explore in our continued quest for understanding our universe.
The Contest is still ongoing until June 8th!
I hope this added value to your day.
Please share this with someone who might find this interesting!
If you have any thoughts or questions about this essay - Let’s Chat
and, of course, please subscribe to Wysr