Abstract

Forming and maintaining close connections is fundamental for most people’s mental health—and hence overall well-being. However, currently available meeting platforms, lacking transparency and searchability, are deeply failing to bring together thoughtful people. This article lays the path for a platform designed to foster close friendships and relationships for people who prioritize learning, curiosity, and critical thinking. The directory of users will be fully transparent and each profile will contain extensive information, allowing searches over all users through powerful filtering and sorting methods. To prevent any value drift from this pro-social mission, the platform will always be free, ad-free, not for profit, donation-supported, open source, and democratically governed. The goal of this article is to better understand the community needs, as well as to gather feedback and collaboration for the suggested implementation.

How to Help

Give Suggestions or Contribute

Give suggestions or show your inclination to contribute through this form!

Join Chats

You can join the community on Slack or Discord to shape and test the product—or just to chat with like-minded people.

Share

Share the idea and article with people who identify with the community values and may benefit from the product.

Donate

You can already donate to support the initial infrastructure via PayPal or GitHub (GitHub has increased transparency, but requires an account).

GitHub repo

It will be created soon once we agree on the name.

Introduction

I’ll explain below my rationale for suggesting the implementation of a bonding platform; since it would be run by volunteers, it goes without saying that those reasons are pro-social.

The starting point is purely personal; I’m currently looking for a few more close connections and, like many, struggling to find like-minded people—despite trying different approaches like dating apps, forums, and real-life communities. Of course, I’ve made wonderful connections along the way, but they require a lot of effort to find them out of hundreds of other ones, and they may not lead to the emotional and intellectual closeness that is so fulfilling in close relationships.

Most of us know that relationships have a major positive impact on overall well-being [1], and most of the happiest people have great close relationships [2]. Additionally, although groups are useful for meeting people, the main value of close connections arises from 1-to-1 conversations, as they foster emotional closeness and vulnerability. With that in mind, I spelled out the type of connections that would bring a lot of value to my life: a close friendship or relationship, with an emotionally stable person around my age, who feels connected to rationality, intellectualism, minimalism, nature, and animal welfare. Of course, this description is much too specific to fit anyone—let alone some of the readers. My values may also slightly vary as my views change; but the description paves the way for attracting and finding people who are much more likely to connect with me.

The more values you require in others, the less likely they are to fit them. So, a good approach is to enumerate as many values as possible as long as the whole set still fits a few people out there. I personally love evidence-based learning; this framework has had the most transformative impact on my life so far. Understanding the world through the lenses of rational thinking has deeply stabilizing me, helping me reach some form of internal peace—or robustness to life events—much better than any other practice. It never fails to keep me excited everyday. My goal isn’t to change anyone’s core values, but rather to make the case that there are many people in the world who find a lot of meaning and happiness in rational intellectualism.

With that insight, I narrowed down my most important values to rationality and intellectualism—which should still apply to 1-2% of the population. Those are the primary values I’d like to see in the people close to me and, as a consequence, in the platform that I’ll detail below. I imagine a community where people prioritize truth over comfort, depth over instant gratification, and humility over (over)confidence. Such is the essence of epistemological connections—deep bonds rooted in mutual curiosity, love of learning, and evidence-based thinking.

Core Values

By rationality and intellectualism, I certainly don't allude to cold, elitist traits devoid of any emotional awareness. I don't mean that one should have already polished those traits either; one should simply value them enough to have the potential to approach them in the future.

All along this article, “the community” should be understood as the set of people who identify with the core values below—whether or not there exists a platform to connect them. Let's carefully explore and define the 4 core values: rationality, intellectualism, relational fulfillment, and interpersonal maturity.

Rationality

Rationality should be understood—very generally—as the systematic process of forming accurate, evidence-based beliefs to achieve some stated goals.

Instead of retreating into relativism—everyone’s opinion is equally valid—when disagreements arise, one engages in critical thinking by questioning each other down to the core of their beliefs. This truth-seeking process is a collaboration of the minds where everyone wins, and people find it emotionally grounding—not threatening.

Curiosity and depth should rise above comfort; truth requires nuance and thoughtful engagement. Evidence is worth more than status or authority; upon presentation of stronger evidence, one has the intellectual humility to change one’s beliefs without feelings of personal attacks. Awareness of one’s cognitive biases, intending to reduce them, and saying “I don’t know” are fundamental strengths. Likewire, one applauds others when they acknowledge their mistakes, limitations, or change their mind.

Rationality may be seen as the exercise of reason at the expense of the impulses—but certainly not of the emotions as a whole.

Of course, those are very high epistemic standards which, perhaps, no one wholly holds; but, again, intention and care for them are more than enough.

Intellectualism

There is some overlap with rationality, but intellectualism should principally be understood as deriving profound satisfaction from learning, thinking and sharing ideas.

The intellectual enjoys gaining knowledge for its own sake—which may or may not prove useful in the far future through higher-order effects—or to fulfill more direct goals. Endowed this epistemic drive, they engage in debate, discussions, and long-form content (e.g., books, podcasts and forums) covering diverse topics that require the exchange of ideas: philosophy, science, history, art, etc.

Of course, even “truth-seekers” want to be silly and playful from time to time; one just needs to be able to engage in rational intellectualism for important decisions, not all day.

Relational Fulfillment & Agency

Relational fulfillment should be understood as experiencing satisfaction and completeness within close relationships—agency being the drive to form and maintain them.

Close relationships are relationships endowed with an emotional bond resulting in a profound care (and sometimes love) for each other. Key mechanisms that usually develop close relationships include value alignment, mutual vulnerabilities, proper communication, and the mere exposure effect.

Interpersonal Maturity

Interpersonal maturity is an umbrella term that applies to relationships in general. It involves, among others, self-awareness, accountability, transparency, conflict navigation, and the ability to care simply and sanely.

To maintain close connections among rational intellectuals, interpersonal maturity is as fundamental as the three other core values; it is the glue that bonds them together and allows smooth interactions. Indeed, rationality primarily helps determine how to achieve goals, but not which ultimate goals to pursue. So, to understand others and care for them, logic must be paired with self-awareness, emotional intelligence, and an understanding of their deeper influences.

Audience

Now that the core values are defined, I’ll describe some other traits in people that partially correlate with them. (Note that the community is defined based on the core values only; anyone who fits the core value, regardless of the traits below, is already part of the community.)

So, here are some tendencies that highlight how most potential community members might be.

Asking clarifying questions, citing evidence, responding thoughtfully, inviting dialogue before delivering facts
Valuing evidence-based practices (e.g., medicine, economics)
Attending reading groups, philosophy salons, or academic communities
Lifelong learners (e.g., college educated who continues learning on the side)
Enjoying online connections (at first)
Fostering friendships before romance

Naturally, the people identifying with the core values are quite rare (probably around 1% of the population). Most people have everyday priorities which clashes with the cognitive load required for rational thinking. Also, seeking the truth is usually linked with an initial change of worldview, which is often destabilizing at first, before becoming stabilizing. Social barriers are there as well, especially under social coherentism or revealed foundationalism—the most popular epistemological frameworks.

Revisions

The core values are subject to slight revisions upon receiving feedback from people who mostly align with them. For example, I have no issue making the community more inclusive by adding debate partners (i.e., with no emotional bonding), as long as everyone is transparent about their intentions.

Market Review

I’ll briefly go over some currently available products / methods that connect people and I'll review if they follow the core values.

I’ll start with the traditional dating apps (Tinder, Bumble, and the like), as they are the easiest to reject. Almost every aspect in their business model and functionalities goes against the social mission of bringing together rational intellectuals for close connections; they compos less than 2% of the use base and they can't be searched with the few available filters. Textual profile information fails to bring any valuable insight into their personality and values, which are the most important causes of good connections. Their business model is even less aligned with the pro-social mission than their functionalities. They aim at maximizing profit by maximizing user retention through manipulative neurological and psychological techniques such as instant dopamine gratification and fear of missing out.

A dating app which could be praised for its more transparent user base is Feeld. There is no swipe; you can browse a directory and like whoever. Their focus on non-traditional relationships and poor filters make it clearly unsuitable for our stated needs, however.

The most popular rationalist communities are LessWrong and its derivatives: Effective Altruism (EA), This Part Of Twitter / Post-Rationalists, Astral Codex Ten, etc.. They are mostly online communities (custom forums, Reddit, Discord, etc.), but most of them organize local meetups. The derivatives mostly embrace the two first core values as well (rationality and intellectualism), and usually add a third one unique to their community. To this day, those communities and their platforms are probably the most likely way to create close relationships between people who follow the core values above. I would love to get some feedback about the number of close connections that emerged from those communities and the proportion of members looking for such connections.

It is yet unclear, however, how much they embrace the third core value (actively seeking to form deep, lasting bonds). Those online platforms do permit to write a profile with interests / preferences and contact info, and to search and contact any other member; this is great for general connections. But they don't allow for detailed, standardized profile information or filters in order to find the people most likely to form deep bonds. So, as of now, the best shot for someone looking for close connections is either to travel to local meetups or to find members with shared values / interests (maybe something they posted or on their profile, with some amount of serendipity) and hope that they are also open to closer bonds. In my opinion, what could be improved on those platforms—or created in a new one—is having access to detailed information about plenty of people and being able to quickly filter the most suitable connections.

The EA community is probably the one that made the most effort to circumvent those limitations and form close, fulfilling relationships. Besides a few unsuccessful calls for building an EA dating app, they published an interesting dating spreadsheet that provides dating documents—which contain information about who they are, what they are looking for, and how to be contacted. The concept of dating docs is a promising tool to connect people with highly specific preferences and standards—as is the case for most rational intellectuals. But, at this stage, those non-standardized documents are scattered across the web, so there is no easy way to filter them.

Other EA platforms for connections include Reprocity and the EA community on Facebook Dating. DateMe, perhaps appealing to a broader rationalist audience, is a web app with basic filters (gender, age and location) where users can share a dating doc and contact information. These three platforms, however, suffer (again) from poor filtering methods and scarce profile information, making it immensely hard to find close matches.

More recently, a few apps have tried to run against the industrial business model of traditional apps. Not a Zombie (yes, don’t ask me why) is an app that is currently being developed, with a “focus on compatibility first, photos second”. Users would first see a text-based description of other people, and photos would only be revealed after some engagement. Interestingly, they allow any user to message any other user, as they “believe that the act of sending a well-crafted first message can prompt interest even if there wasn’t already interest otherwise”. To prevent spamming, however, they would limit the number of profiles visible per day. Many of the app components are valuable food for thoughts, but the app itself—if it ever gets built—would fall short to embrace the core values of the preceding section. It is for dating only, not open-source, and every user may only see a few profiles per day.

Another recent app, Firefly, has reached more popularity while keeping its pro-social mission. Its user base is fully transparent and many interesting prompts bring thoughtful discussions. But, again, profile information and filters are very scarce, so it’s hard to search the user base and spot like-minded people. The code is not open source either.

Lastly, BeWelcome, a French web app focused on hosting and connections, is worth considering for its extreme pro-social mission and worldwide success. It is proud to be completely free, solely funded by donations, ad-free, non-profit, open-source, and run by volunteers. Its governance is partially democratic; they have three types of position: member, BeVolunteer member or board member. A BeVolunteer member is approved by the board members, and the board members get elected by the BeVolunteer members. All decisions are voted by BeVolunteer members only; so the users have no voting right. Profile information is diverse, although tailored for travelers—such as hobbies, languages, and hosting capabilities. Their member directory is transparent, but only searchable by location (and username…).

To conclude, none of the platforms above seem to fully fulfill the core values and goals stated in the precedent sections. Please let me know if there is any other platform you may know that better approaches the requires features or if one of the above platforms would be open to upgrades that embrace relationship fulfillment. Otherwise, I will be happy to build one, as detailed below.

Platform Values

Mission

The platform’s mission is to maximize the number of close, meaningful connections among the community (i.e., people who align with the core values).

To best perform the mission, we need a free, ad-free, transparent, community-driven platform. I describe each aspect in detail below.

Free

To ensure no manipulation for profit, users must have full, free access to the entire platform.

A free platform means freedom from data exploitation and other commercial techniques.

Ad-Free

To avoid dubious external influences as well as a mission drift toward user retention optimization, the platform must be free of any advertisement.

If some mission-aligned institutions sponsor the platform, their acknowledgment may only be shown in textual form outside the main page (e.g., on a “Sponsors” page).

Transparent

To foster decentralized trust in the platform and other users, the platform must promote radical transparency. The entire source code must be public on GitHub. In addition to having their implementation open source, all matching, filtering and sorting algorithms must be documented in English on the platform. To helps curious thinkers find each other without endless swiping, all profiles must be searchable.

Data privacy must be promoted as much as code transparency. Users must be made aware of the visibility of their content. Direct messages, login information and other content marked as private must be kept private to the user. Whether some user data are public (for the world wide web) or semi-public (for logged-in users only) must be clearly spelled out.

Mission-Aligned

To align with the core values, the platform must promote slowness, reflection, vulnerability, truth-seeking and openness. Tools must be designed for signal, not addiction.

Community-Driven Governance

To prevent the platform from drifting toward exploitative practices, it must be created and maintained by the community, for the community. Decisions must follow a democratic process. Everyone may suggest and implement features.

Implementation

I’ll describe in more technical details how to implement the platform in order to promote close connections in a very efficient fashion, while staying aligned with the platform values—and hence the core values. It will be very heavy on data and algorithms. The back end will be extensive and accessible to the user, while the design will be clean but minimalist (a drastic paradigm shift from traditional meeting apps).

Profile Information

The profile page should front-load intellectual and psychological alignment, while still being human and approachable. It should contain as much information as possible—this requires a lot of self-awareness. Naturally, many people would be uncomfortable sharing so many personal things about themselves, especially since everyone in the app could view the profile. One easy way to circumvent this concern is to set different levels of privacy for different parts of the profile; the main screen would be open to everyone, and the other parts would be revealed whenever the profile owner feels like the connection has matured enough.

The first screen could thus include:

Name and small headshot (20% of the screen max)
Intellectual topics currently being explored
Favorite intellectual topics
Least favorite intellectual topics
Thinking style
Results from evidence-based personality tests (e.g., Big 5)
Conflict style
Desires: type of connection, activities to do, etc.

Many traditional features remain important, such as the level of education, job or studies, hobbies, pets, habits, subcultures, diet, emotional sensitivity, sense of humor, ambition, organization, pet peeves, non-negotiables.

They could mention their physical and mental health—especially some traits that rub people the wrong way, insecurities, biases, triggers, therapy, or the things they are trying to improve about themselves.

Their values must also be clearly stated, as value alignment is a strong indicator of good connections. Moral aspects include community engagement, social justice, and other cause areas.

For people interested in romantic relationships, they would benefit from talking about their love languages (giving and receiving), timeline, romantic orientation, family projects, work-life balance, financial goals / habits, career goals, housing situation (renting vs owning), location, and whether they would date someone who already has kids.

I think it would also be valuable to add a profile variable that tracks when a user was last connected, in order to let any user filter out inactive profiles.

What they are looking for is also paramount for a good match. The user could go through all the features above one more time and write what they would like in their ideal person.

The points above are important to define tags (necessary for filtering), but the user could in addition complete their profile with as much well-formatted text as they want—including a few pictures, link to a blog, contact information, or even some reviews from close people.

To avoid empty profiles, we would set a minimum number of characters to write on their profile. Also, we’ll make some features mandatory (personality traits, intellectual interests, location, desired connection, etc.). An interesting nudge to incite people to fill in their profile would be to add a profile variable for the level of profile completion (in percent); like any other feature, one would be able to filter or sort by profile completion in order to interact with the most completed profiles.

Modules

Ultimately, we could add modules to raise self-awareness about disagreement styles, thinking styles, intellectual interests, philosophies, etc. I suspect this would add very useful profile information.

Prompt-based Messaging

Deep open-ended questions are a great way to deepen connections; they can be answered individually, used as openings, or as tags for filtering. Here are some that I find particularly appealing in light of the core values, but they should probably span many more topics.

How has understanding X helped you suffer less when Y happened?
How have some ideas changed the way you respond to stress?
What’s something you recently changed your mind about after seeing stronger evidence or reasoning?
What do you think counts as a good reason to believe something?
What’s a topic you find endlessly fascinating but rarely get to talk about?
Have you found that understanding Bayesian thinking changed how you handle personal conflict or relationships?
When someone disagrees with you on a topic that matters, how do you approach the conversation?
Has understanding the truth about something ever brought you peace, even if it was difficult at first?
Is there a belief you hold that you wish were false—but still think is probably true?
If we were to spend an afternoon talking, what topic would you bring up first—and why?
When you're in conflict with someone close to you, what happens in your body? What do you notice first—emotions, thoughts, or physical sensations?
Think of a time when you and someone important to you disagreed, but came out stronger after. What do you think made that possible?

I also like the idea of user-generated prompts voted by the community.

Searching

The principal method to find like-minded users will be through advance filters and sorting algorithms across most of the variables defined in each profile.

The difference between filtering and sorting may be subtle but very useful. Filtering simply means finding a subset of the user base that matches some hard criteria. For instance, one could filter for openness (in the Big 5) higher than 50% AND interested in a relationship, OR favorite intellectual topic is neuroscience. With pure filtering, the resulting profiles are presented in no specific order.

Sorting means assigning a number to each profile and ordering them according to that number. The formula which maps each profile onto a number will be fully defined by the user. For instance, it can be "agreeableness in percent + 10 for each currently explored topic belonging to biology”.

So, filtering is for hard desires and sorting is for soft desires. Mixing the two methods provides immense capabilities to find the connections that matter to us the most.

Alternatively, some open users with less specific preferences could enjoy a secondary tool that balances discovery with serendipity (aka the exploration / exploitation trade-off). If desired, it would be easy to provide profiles that fall slightly outside the mentioned preferences. Or one could use filters with no sorting to randomly sample profiles from the filtered subset. At some point, people could use some algorithms (transparent and curated by the community) to sort their filtered selection; but this would always be an opt-in feature that would never override a sorting method implemented by the user.

Matching

Once users find great people, one needs to configure if they can start a connection.

In the ideal world where everyone is good and caring, one would let all users start a connection with anyone else by allowing direct messaging. In a world full of spams and harassment, one would only be able to “like” a profile; and the conversation would start only when both people like each other. This process would of course be catalyzed by making the list of who likes you fully transparent.

Although I’m eager to receive suggestions in the comments or in the suggestion form, my belief in the goodness of humanity makes me advocate for a hybrid model closer to the ideal one: people can direct message anyone, but this opening would be received more quietly than a message from an already connected user. For instance, those new messages could first arrive in a list separate from the one with connected users, and they would not trigger a notification by default. If early evidence shows that this hybrid model is still subject to immense spams, I—as a good rationalist—will be happy to update my prior.

Of course, people are free to share their contact info on their profile (unlike—you guessed it—traditional apps) if they prefer. Likewise, people exchanging messages on the app would be free to share any information at any stage of the connection—including contact info to connect outside the app.

It always feels painful when someone cuts connections without any context. When a person leaves a chat, the platform could provide general guidelines for ethical endings and invite them to fill in a quick form with the reasons why (for the other user).

Now that I presented how all the good things can happen, from joining to meaningful connections, I’ll dive more into the technical details and precautions to make this pro-social tool viable under the hood.

Tech

First of all, building free and open source software (FOSS) is a drastic paradigm shift from current meeting apps, which brings its pros and cons. One major positive side effect of FOSS is the attraction of technically skilled contributors.

The minimum viable product will contain these features:

Authentication
Page listing all the profiles
Search through all the profile variables (intellectual interests, location, cause areas, personality type, conflict style, desired type of connection, prompt answers, gender, etc.)
Direct messaging
(Prompts or Modules for self-awareness)

I propose to develop the platform as a web app first, as it’s much faster (to develop) and works on all devices. For that purpose, using a framework like React for the front-end and a service like Google Firebase for the back-end seems appropriate. I would consider wrapping it around a progressive web app at some point to make it easier to transfer to a mobile app later. Hosting could be under Firebase Hosting, Vercel, Netlify, or even AWS (seems cost-effective at scale, and I know how to use a few things); let’s discuss about the best one. I personally have reasonable full-stack skills and don’t mind learning on the field, but it would be awesome to get contributions or feedback from developers with specific experience in any of the above frameworks / services.

The software will be under a permissive license; deciding which one is still open.

For optimal user privacy, end-to-end encrypted messaging may be implemented at some point.

Branding

I suggest the branding to be intellectual: high-contrast design, thought-provoking (but humble) and philosophical tone in a clean / bookish font.

Moderation

Strong moderation should be done to filter out spams, harassment, conspiracy theorists and contrarians for the sake of it.

Keeping in mind the advantages of decentralized governance and the limited number of volunteers, I suggest multiple layers of moderation. The first layer will be moderated by the users themselves; they can report and review anyone else—e.g., someone who wrote an inappropriate messages (sexual advances to someone who wrote on their profile that they are not interested, etc.) or a profile with misaligned content (inappropriate images, hate speech, etc.). A profile would be suspended after a few reports.

The second layer will be run by volunteering moderators. They’ll review profiles that got suspended, remove inactive account, and perform other essential tasks.

If a user notices that someone’s profile isn’t accurate, they would be able to softly report it as (anonymous) feedback. This feature would ideally be framed as a helpful “nudge” between two users, as a way to improve profiles and hence successful connections. A profile that keeps receiving soft reports could be scrutinized by moderators.

If too many newcomers are misaligned with the core values, vetting—examining interested users before joining—may be considered at some point. An epistemic and interpersonal onboarding that checks for core principles of rationalism and healthy relationships before signup may help.

Governance

There are at least two possible models for governance, and I am yet unsure which one would best achieve the platform’s end goals.

The first model is cooperative: users or contributors collectively own and manage the platform—they all have voting power and share in decisions.

The second model is a stewardship council: a small group of deeply aligned users would enforce the value by voting on major decisions. This raises questions about how someone would be granted access to that council; elections and rotations would help, I suppose. Regardless, all rationales and votes of that council would be fully transparent / public.

A hybrid model may also work. The co-op would vote on features, and the council would make sure the features are aligned with the core values—e.g., referring to a written constitution.

A constitution would extent the Platform Values section by including, for instance, a mission statement, some core values, a user bill of rights, and governance principles.

Finance

Finance is a critical part of the platform which may severely (and indirectly) impact the mission. Indeed, this platform will have inescapable expenses, but it doesn’t have any clear, easy source of revenue.

On the expense side, since everyone will contribute for free, the main cost related to the app will be hosting servers. Expenses should be minimal, if carefully developed, but I don’t have an estimate right now as there are too many uncertainties. Hopefully, we’ll have an estimate per month and per daily user in the next weeks.

On the revenue side, donations (from users, institutions, etc.) will be the source most aligned with our mission. If donations continuously fall short, a donation campaign would be the best band aid. We send emails and add banners to the platform with a very transparent messaging about the platform’s financial health and how much it needs to be viable—not unlike those Wikipedia campaigns. As a last resort, optional memberships (e.g., providing voting rights on major decisions or little perks) or one-time lifetime purchase may be considered.

Ideally, we will use OpenCollective (or similar) as long-term funding platform, as they provide massive transparency; but we can’t use them before the community and platform exist, So, we will gather initial funding throughout my personal accounts (see top of the article); please mention “meeting app” if through PayPal. Obviously, I pledge to use the donation solely for the development of the platform and, in the event that the platform fails, to donate the remaining capital to a charity voted by the community. Finally, expenses and revenue will be fully transparent and available in real time on the platform.

Extensions

The question of selectivity vs openness is particularly interesting from a utilitarian point of view: how selective should the platform be to maximize overall well-being?

If the platform is very selective, there will be only a couple people with very aligned values. Their well-being will increase a lot, but there are only a few of them. So there is not much overall benefit for society.

If the platform is fully open (everyone can join), there are two possibilities depending on the connection mechanism. If the list of members is opaque and poorly searchable, as in traditional dating apps, it becomes very unlikely to find value-aligned people. Hence, each individual’s increase in well-being is negligible, even though there are plenty of people, and there is not much overall benefit for society either.

Qualitatively, the figure below illustrates this trade-off between selectivity and openness for poorly searchable (i.e., Tinder-type) platforms. The quality is the individual increase in well-being, which increases as the platform selects for people strictly following the core values. The quantity is simply the number of users. The platform’s overall benefit (read, increase in total well-being) is then the product of quantity and quality—reaching a maximum with a non-extreme selectivity.

The second possibility appears when the members are fully visible and searchable by anyone. In that case, each member can filter and meaningfully engage with the few people aligning with their values. Each individual’s increase in well-being becomes important, and there are plenty of people. So, a fully open and searchable platform may bring a lot of overall benefit for society.

Of course, in practice, there will always be interferences within big communities. And, more importantly, a larger platform requires more resources (i.e., funding) and moderation. That’s why the focus of this article is on a specific community for now. Not only do I identify with the rational / intellectual community, but it is also composed of members who are much more likely to contribute (especially on the tech side), making it a very convenient community. But if it creates much greater good, I think it would be worth considering extending the platform at some point in the far future—provided that it doesn’t negatively dilute the community or create brand identity confusion. More than just creating a higher good, a larger user base means economy of scale: donations may scale proportionally while expense per user would diminish—making the platform more likely to survive financially.

Viability

Let’s think quantitatively about what would be required to make the app useful and viable.

By useful, I mean that we consider it a win (for the mission) when every user makes at least one meaningful connection through the app. Let’s make a few assumptions:

Users will only connect with people within their 20-year range, and people are uniformly aged between 20 and 50. So the probability that two users are aged at most 10 years apart is around 50% (20% for the youngest and oldest people, and 67% for most people in the middle).
Users will only connect with people in the same geographical area, and they are uniformly spread across 5 of them. That’s a 20% probability to find someone in the same area.
Users may form a connection with people of any gender (it’s not a dating-only app, after all).
Since most users would share rational and intellectual interests, their chance of sharing similar interests and personality traits with another user is higher than usual: 5%.

So, the probabilities for any two users to form a meaningful connection is 0.5%. That means a minimum viable user base of 200. Once the app reaches this critical mass, I suspect that its value will be tangible—through testimonies, etc.

Another critical aspect that requires people is moderation. I would estimate that one moderator for 100-200 users is enough—most of the users should in principle be less spamming or violent than the average person. That means 2-5 moderators, up to 1000 users.

The number of maintainers can be minimal once the MVP is out; 1 or 2 people who periodically look at issues (especially critical bugs or vulnerabilities) should be enough. If feature requests come out, that’s great; but that’s a job for kind contributors (not maintainers).

Now, though to a lesser interest, we can estimate the maximum number of users that could use the app. As the app will be in English at first, we narrow down the population to 1.5 billion people. It’s a niche topic, so probably only 1% of the population would fit the core values. Let’s be gracious and assume that 50% of them are fulfilled and don’t more connections. That brings the total addressable market to 7.5 million people.

Critiques

I’ll list some expected critiques and provide potential solutions.

Choice & Cognitive Overload

When one can see everyone else, it’s easy to spend more time browsing than engaging. Also, users must actively filter profiles and go through a lot of information, creating much friction. This is a key difference with traditional apps—and it’s clearly a feature, not a bug. I suspect these concerns may only concern to the more casual, less data-savvy users or the ones with less specific preferences. For those people, I would suggest a discovery mode that resembles more traditional apps: a limited number of profiles shown one at a time by a curated (but transparent, of course) algorithm. This casual mode, however, will never reduce the functionalities and efficiency of the explore mode dedicated to power users.

Spam and Harassment

On a platform where everyone can message each other, it’s so easy to be spammed and harassed. The Moderation section suggests different strategies. Let’s carefully monitor that aspect as the product goes live.

Capping the number of daily messages may be useful as a last resort. I am somewhat reticent to the idea, however, as I believe the app should be fully usable by every “good” user—which is the rationale behind making it free. Maybe there is value in capping the number of first / opening messages, instead.

Unequal Gender Ratio

Men are more likely to outnumber women with those niche values, but this should not be an issue as the app won’t be focused on dating.

In this situation, women will be at much higher risk of spamming and harassment. If the strategies in the Moderation section fall short, we’ll tackle the issue more aggressively.

Self-awareness Dread

People usually don't like to write descriptions about themselves. They get descriptions from others and follow self-awareness modules, however. People can also give feedback and suggest modifications.

Online Connections

The platform doesn’t target a specific location; until we get thousands of users, it’s very likely that connected people live far apart. It’s a real concern, although the rational / intellectual community has slightly lower needs for in-person connections than the average person, as a lot of their activities pertain to the mind.

Premortem

Here I’ll identify a few potential problems that may terminate the project and address whether, in my humble opinion, they should be concerning.

What if I don’t need the app anymore?

There will likely a time when I don’t intend to use or develop the app anymore. When this happens, the app will likely survive on its own, per the robust mechanisms detailed above (open source, democratic governance, etc.). So, if contributions of all types (donations, moderation, and code maintenance / improvement) are decentralized, my presence, or the lack thereof, should not significantly impact the well functioning of the app.

What if donations can’t meet expenses?

This is a more critical issue. If donation-based models fail, even for such an app that would require so little financial help, then it simply means that the value of the product is severely below its maintenance cost. And a mission-driven product with net-negative value and no potential for improvement is not fulfilling its mission; so it should be shut down.

Roadmap

If the app attracts enough contributors, a minimum viable product should be available within 1-2 months. If it further gains popularity, features would be released whenever the community implements them.

In the meantime, please help us by filling up the suggestion form (at the top). Here are some name ideas from ChatGPT:

BayesMeet
Prior
CoThink
Sensemaking
Reflect
SlowMatch

Conclusion

I hope the reading was valuable. Let’s build a better way to form meaningful relationships—one designed for people who think deeply, value evidence, and want connections that last. If you believe conversations should go deeper than swipes and small talk, join us.

To learn more about me or connect more personally, feel free to check out this form.

Much love to everyone!

[1] Waldinger, R. & Schulz, M. S. (2016). The Good Life: Lessons from the World’s Longest Study on Happiness.

[2] Diener, E., & Seligman, M. E. P. (2002). Very Happy People. Psychological Science, 13(1), 81–84.