Older thoughts about Secret and the web of trust.

[This is an personal journal entry from a few years ago when Secret still existed — I think my views have changed somewhat but I figured I’d post it because it reflected things I was thinking about at the time]

Often it seems that social internet projects succeed when they have some mastery of the gray space between public and private. One axis is like this:

  1. Identified platforms (where utterances and identity are connected) tend to reflect a kind of uniform understanding of the world — with an insensitivity towards individual vulnerabilities and marginalized voices.
  2. Anonymous platforms (where utterances are detached from identity) don’t exploit the power of trust networks. They do provide safe space for candid conversations — but also tend to be superficially violent and ugly.

And the other axis is like this:

  1. Personal — concerned with an individuals lived experience, joys, tribulations, sadnesses; more intimate — one feels like they “know” somebody.
  2. Professional — concerned with wealth generation or changing the world, more about facts outside of a persons own embodiment (although sometimes people are their own product). Often a person who has a strong professional presence also manufactures a mask or persona that reflects that professional stance.

For a few folks in our civilization (the so called “mainstream”) their identity and persona give them the freedom to converse within a specific facade. They get some reward, and are rewarded for utterances in the mainstream.

In reality we are all probably freaks to some degree and for many if not all of us there are risky behaviors that can cause social criticism or even personal risk if made public. We all diverge from the image in the mainstream.

What are examples of intimate statements?

These would be disclosures that put oneself (or others) at risk. Any statement that could reflect badly on character, that might be wrongly assumed to be part of a broader trajectory of identity of self. Basically the kinds of disclosures that can be used by other parties to inhibit or impede personal freedoms and growth. For example if we look at the lives of professional actors and politicians we often see attempts to undermine character and discredit parties (for a variety of reasons).

Even acts that are good acts that one does put one at risk — one can come across as having ego and the like which is frowned on in many cultures.

I don’t see these kinds of disclosures as truths per se. I simply note that there is a range of statements about a given party (either by that party or others) which may cause injury to that party and which could tangibly inhibit or impede the life path and future opportunities of that party. Sometimes statements are made inadvertently, sometimes people come out of the metaphorical closet, sometimes people utter statements about other people, sometimes statements are taken out of context — but I consider the veracity or origin of the statements to be distinct from the subject matter.

Kinds of statements that create liability include:

  • humble brags
  • acts of kindness or concern for others that reflect empathy but that one doesn’t want to say with ego
  • celebrity gawking; oddly increasingly common in silicon valley
  • schism creating statements — true or not observations designed to ignite debate over a social tension
  • personal sexuality or topics around sexuality (who you were intimate with, how good/bad/unusual those experiences are, queer culture, non-monogamy and the like)
  • changes in job roles or employment status; changing job roles, quitting, being fired
  • general disgruntlement with people, places, jobs or situations
  • criticisms, judgements, parodies, defamatory claims
  • overt enthusiasm for topics that are not gender or role aligned with public estimations of role and identity
  • pursuit of personal topics of interest that differ from your public persona (a pursuit of the arts while not being known for that previously)
  • perceived weaknesses or liabilities — things others would consider liabilities — physical or mental differences from idealized people.
  • perceived vulnerabilities ( having been abused for example )
  • incompetence, mistakes, errors, poor judgement and other kinds of actions that cast a poor light on personal decision making ability, character and integrity
  • descriptions of sociopathic and insensitive behavior, non-culturally accepted responses to the suffering of others and the human condition.
  • descriptions outright malice, deception and general hostility towards other humans / humankind or other entities who are capable of suffering pain by ones hand
  • abusive behavior towards others such as threats, belligerence, strutting and dominance posturing

What are the kinds of personal liabilities that online conversation can have?

Intimate and risky disclosures can constitute a personal liability. Obviously a person can be physically assaulted or harmed based on perception, but more generally even just censured or outcast from communities or job roles based on third party perception of beliefs, behavior, conduct.

It’s also difficult to air some intimate disclosures because social networks tend to not propagate disclosures which contradict social norms. This can make it difficult to evaluate or test ideas — even as part of a process of rejecting them — since it is hard to shine light on them and get consensual analysis of them. As well it makes it difficult to find if others do share similar values. This said — a benefit of personal disclosures is to test where community values are and to push community values in a given direction.

For worse we live in a society that can be quick to rush to judgement. Often a good defense is proactive transparency. “Knowing thyself” or having a clear and personal sense of identity in order to be able to make clear statements about that identity can help. Also it can be useful to create a lot of room in a personal identity — so that one is not typecast as a specific kind of person.

To be exposed to as broad a community as possible is to be well rounded, grow and learn tolerance. But we do more than just respond to community, we also shape it. Standing outside of community and holding a value position can take work. The further from the center the more work. Often outsiders protect the center.

Our public persona may be shaped based on others expectations and peer pressure. In some senses the rewards are extrinsic; not fundamentally due to the role or persona but due to others acceptance of that persona.

This can create a gap between that persona and actual personal life satisfaction. As well, owning a persona can make it challenging to take detours and be exploratory — and thus can make it harder to explore what are ones ‘wheels within wheels’ to understand ones one innermost motivations. Many people operate at some tangent to an ideal personal direction because of external pressures. (The inhibitory influences can be positive in helping to provide guidance through a thicket of choices of course).

What are some of the social networks where conversations occur and how do they differ from each other?

In the last decade we’ve seen the rapid emergence of a wide variety of online social networks such as Facebook, Twitter, LinkedIn and Secret. In a sense each of these services is a kind of brokerage that connects parties with complementary interests together. The physical structure of these services is worth noting because each differs:

  • Facebook is technically on the web but in many ways acts like a private silo. Posts are not generally visible on the web anonymously and are not addressable outside of Facebook — posts don’t participate in the web as a whole so all discourse on a post is limited to being within Facebook. Posts have a limited visibility because Facebook prunes posts using an unknown algorithm. Peers will not see everything that a given party posts. Peers on Facebook tend to be somewhat self-selected although there is social pressure to accept people as friends who may not be intimately connected. Facebook tends to have a conversation that are appropriate for a mixture of professional and personal audiences. Intimate disclosures or vulnerabilities are somewhat less common, defamatory comments are also uncommon, truly vulgar or inflammatory trolling is infrequent or blocked— although sometimes honestly held views diverge radically. Facebook is like a social space on training wheels compared to say 4chan or other networks where rhetoric can be grotesque and distressing if one chooses to give credence to the voice.
  • Twitter is a public social network that is highly visible on the web. Posts do tend to have limited visibility simply because of the flow rate of traffic. Identity can be concealed but since it takes time to build a following it is not fruitful to post under an anonymous identity. Twitter tends to have swarming patterns and there is a substantial amount of violent and angry language — often trolling is exhibited. For some reason Twitter lacks any kind of contextual network graph or weighted trust filter for scoring voices so there is no signal attenuation of bad traffic. It’s easy to make a robot army to attack a person — they are not invisible even though socially disconnected from a target under attack. It’s a broken social network contract and surprisingly poorly designed.
  • LinkedIn is something of an anomaly. It is somewhat like Facebook but missing the personal side. There’s a focus on professional interactions related to the work sphere — although culturally our values around work are pre-capital and to some degree discussion of work is frowned upon and seen as posturing or ego-driven. It comes across as a fairly unhealthy space with a fixation on for profit motives — even though generally speaking people are extremely polite given that their professional identities are on the line.
  • Secret (which I’ll focus on below)

Looking at Secret more in depth:

Secret is probably the most unusual project to come along recently, and sparks much of my thinking.

It lets people make posts anonymously while allowing their extended peer group to still provide feedback on their posts. Effectively posts are public but the author is anonymous.

To be on a soap box means to attract lighting strikes or at the very least to have to own a given persona or posture. On Secret a speaker retains anonymity (to some degree) but statements can be public and have public feedback. A person can be vulnerable, critical or defamatory and still be corrected, have feedback and the like but not put their life at risk. Somebody can say things they don’t believe in, or that are rumor or conjecture. They can pretend to be somebody else or even themselves and use that voice to make statements as if issued from that party. Secret likely filters out many posts but still at some level a candid and curious questioning cultural spirit seems to emerge.

Posts on Secret tend to be more intimate and often more vulgar than other services — presumably because personal identity is removed and the risk of being colored by a post is low. Since posts are anonymous it is not clear how much interference there is in what would be a raw unmediated social discourse. It is likely that there is a large staff employed by Secret to remove the most egregious posts and as well to create fake posts to direct the tone of the community. The community is also seeded with a certain kind of person. There have been some posts which were defamatory — leaving those personalities without a defense in the public eye — Robert Scoble is fairly upset with Secret for example.

Even though posts on Secret are anonymous there is something of a tendency to want to engage in in-depth discourse rather than off the cuff comments. It may be that the community self-selects from the starting pool of intellectuals and academics that seeded it — but as well the reward for trolling seems low with anonymity. There can be a sense of using intelligence and wit to “win” Secret by making a post or comment with the most up votes — and poorly worded or thought out ideas don’t get that reward. Trollish or simple knee-jerk humor or otherwise easy to manufacture statements don’t get many upvotes in this crowd.

There seem to be 3 spectrums — of which I try to draw above:

  • Personal conversations <-> professional conversations. Some services such as Facebook are focused more on personal journey and identity. These sites tend to encourage manufacturing a persona that emphasizes vulnerability and personal growth. On the other side of the fence professional conversations largely concern themselves with concealing weaknesses and emphasizing strengths — or at the very least providing a sense of core competencies to find other parties who are complementary to those abilities.
  • Anonymity <-> known identity. Being anonymous as opposed to being public. On some services one can be anonymous while still having an audience, and on other services one has to build up an audience (ostensibly in public). On services like Twitter ones identity is often although not always public; and of course statements made are often visible to the world. Anybody can make critiques or defamatory statements — but the criticizing parties are themselves public as well and therefore are at risk if they utter statements that are either untrue or dangerous. Even if an individual is hidden their “brand” or twitter account reputation is at risk. A cultural hierarchy is at work here; criticisms, secrets and the like can cause harm to the party that speaks them depending on who that party is. In that sense a public identity is volatile; utterances can bring a person down. This is similar to the concept of ‘True Names’ elucidated by Vernor Vinge — while some harm can be caused to an anonymous account — real harm can be caused to the physical body of an authenticated account. An example of a dangerous statement might be to say that somebody embezzled funds, or that somebody has caused harm. That party in turn or their community, along with their robots, may act to silence the utterance and the person pointing fingers.
  • Privacy <-> publicity. Not as emphasized in my thinking (slightly overlapping with other concepts) but the visibility of a conversation also affects the liability of that party. In the drawing above I conflate the idea of anonymity and privacy and the idea of being known and public. Ideally this should be painted in 3 dimensions.

What’s the right kind of service that can straddle the kinds of failures that the above services exhibit?

An off the cuff wishlist — which isn’t terribly well thought out:

  1. Social dialogue should be filtered by some kind of trust network. There’s a weakly transitive property of trust in that you trust a friend, and to some degree their friends. A contextual network graph can trivially score your extended trust network, and such a graph can act as a trust filter or a lens to attenuate traffic against. This would solve the Twitter conundrum and make Twitter actually useful. It’s surprising they don’t do this.
  2. So many of these services seem to be shallow echoes of each other, there are fragments of each in each, and it feels like the stack shouldn’t be vertical — that there shouldn’t be separate services all replicating each others features weakly, but rather horizontal, different services should each take on one role.
  3. Identity should be anonymous even if it has consequences that are negative. Our civilization is far to bloody minded to tolerate differences of opinion. A way to have an anonymous identity that can also have visibility is to have many utterances over time. Over time an identity emerges and that anonymous account has some value that helps keep the anonymous speaker from running off at the mouth and being an asshole.
  4. Services should probably be distributed rather than held by a central service provider (such as facebook). This also means self policing against the most egregious statements. As our online communities mature we need to really learn how to self-police, to rotate that role through an engaged civic populace, so that we’re not always at the mercy of those who would offer to manage us but learn to take on some of that responsibility ourselves. The risk of centralization is that the center can do arbitrary filtering or modulation of content and the system fails to represent the ground truth of the reality of human experience. As it diverges from reality the members who subscribe to it increasingly make mistakes in their estimations and actions in reality. The echo chamber effect is especially pronounced in our current political landscape.
  5. Mature social networks should probably allow some kind of flagging of “kind” of statement. Statements could be marked as merely a heartbeat, or as a request for help, or as being temporally or spatially relevant. By having some kind of wrapping scope on a statement it becomes easier to create rooms for groups of statements that parties who are interested in that particular topic can listen to without being drowned in noise.
  6. Facebook and like services tend to focus on the individual although they do have group concepts. It feels like a more mature kind of social space would be room centered. Rather than being so focused on who ones friend is — entire groups of people with similar interests can group around that interest. While this does occur it feels like the “forward facing” aspect of Facebook is the personal or intimate experience of a persons life.
  7. Services should be multi-portal. One should be able to interact with all services via chat, email, web or sms — and all participants should be able to have arbitrary gateways to content; viewing content as they wish. This is also related to de-centralization. A truly mature social network is a utility — it isn’t owned by one party — but is merely a way for people to communicate about a variety of topics at a variety of scopes.
  8. Statements should be able to be structured — not just having “kind” but having precisely defined sets of properties. Many of the services are echoes of each other ( Craigslist, AirBNB ). With trust filters and stronger structure there could be actionable regional discussions about civic issues, actual buying and selling, all manners of brokering, gifting or other kinds of transactions that are prevented by the clumsiness, siloing and limited visibility of the various self-similar services that exist today.
  9. There may need to be active design in the system for what is best for the system — not for individuals. For example it may be necessary to formally stir voices together to break down echo chambers. Or it may be necessary to encourage people to meet outside of the digital realm to have discussions with real empathy for another persons lived experience.

As these social networks mature we need to think about what we actually want them to do for us — how they can work for us, not just as revenue for an organization, but how they start to define our culture, the public weal, and the future trajectory of our civilizations.

I’m sure that with the kinds of cavalier generalizations I’ve thrown out that there’ll be a lot of criticism of these thoughts. And this does reflect older thinking. I think at the heart of this I’m questioning / curious about if what we have reflects what we want and if it could be better. I’m not especially attached to my observations.