On Credibility: Whom to trust on the Internet

This post is a follow-up to the one on the Good Internet: With the enormous and diverse set of people made reachable to you through the Internet, how do you choose whom to trust? Ideally you’re interacting with transparent, accountable people who may even be experts in their fields, who may have tons of invaluable wisdom and knowledge to share—if only you trust them.

The hard problem

This presents a kind of problem that is relatively new to humanity. How do you tell if someone is knowledgeable and trustworthy, without already having the subject-matter knowledge or knowing anything about them personally? We didn’t evolve to solve this problem—in pre-technology eras, just about every person you’d meet would have a legible context that you could use to form judgments about them. They might be a member of your “tribe,” or indirectly related to your tribe through the praise of a friend, or a member of a rival tribe, or indirectly a rival through the slander of a friend. You’d have long webs of trust that would be mediated by hundreds of in-person relationships, and any new person would be trusted or distrusted based on how they fit into your web.

It’s not so on the Internet. On the Internet, you’re dropped into the middle of the ocean, a stranger to every other living thing. There are tribes, sure, but they’re all foreign to you. Even if they sound like the type of person or community you’re familiar with, they’re all strangers and you can’t assume authenticity. We didn’t evolve for this, and that’s why scams on the Internet are so easy and so common—our default trust mechanisms can’t work there.

Of course, sometimes you can rely on credentials. A blue check on some platforms tells you that yes, this person really is the highly educated/successful/influential person they claim to be. But then, when their celebrity is common knowledge, you’ll have a hard time competing for their attention. I see much more potential in the practice of finding and connecting with non-famous-but-very-valuable strangers.

But given that the Internet is an ocean and we have no context for anything, it’d be tempting to not engage with strangers at all, except for the fact that some of them could be extremely valuable to you. You can have a one-on-one conversation with a subject matter expert without even telling them your name; you can stumble upon free financial advice that actually performs well; same for career tips and business strategies; you can get tips on which products and services are actually worth buying. We could say that one perk of the Internet is that it’s not an “efficient market” of information: great information can be found just as cheaply and easily as bad information.

So, figuring out which strangers to trust on the Internet is a challenging problem, but potentially a very rewarding one.

Past successes

I’m weighing in on this because I have reason to believe I’m particularly good at solving this problem. In recent years:
  • I joined a personal finance / lifestyle chat group with about 3000 people. Most of them were uninformed idiots and scammers. Of those 3000, seven of us found each other and thought we were pretty cool people. We split off into our own chat.
    • One of those seven recommended an options trading strategy to me, originally created by a random person I’d never heard of. I learned the strategy from a YouTube video with only ~3000 views. I’ve been trading it for over a year and it’s been profitable. He’s taught me a lot more about options trading since then.
    • Another of them has credibly proven, through personal photos and video, that he owns and runs a very high-end gym in the tri-state area. His services are way beyond my budget, but I can get fitness advice from him whenever I want, all because we found each other in a stupid group chat.
    • Another of them has credibly proven, by showing me in-person, that he runs a set of automated trading strategies which collectively earn a profit that’s competitive with his NYC software developer income: Really a life-changing project, which he’s offered to mentor me on setting up myself.
    • Some of these people live in NYC, so now they’re my actual in-person friends. They are in fact pretty cool people.
  • I became convinced, by a complete stranger in a different chat group, to buy Bitcoin, at a time when it was priced far below its previous high and had no momentum. This was February 2020, priced at about $9K, so clearly it’s performed very well as an investment.
  • I could have been tipped off early to the Covid-19 pandemic, if I had bothered to read the blogs that I’d already curated in my RSS feed. In other words, while I was taking a break from reading news in January 2020, the sources that I liked were sounding the alarm about Covid, telling people to stock up on food and toilet paper (!) even as every mainstream media outlet was dismissing Covid as a non-issue. So, I missed that one because I was taking a break from that corner of the Internet, but I give myself credit for picking the right sources for news analysis, as far as global-pandemic-predictions go.
  • I used an instructional YouTube video that I found by accident to hack the Spotify installation on my computer and give myself unlimited ad-free music for free. It seemed like it would work… Now it’s been working without incident for years. (I know this is not ethical; we can either pretend I only did it for “research purposes,” or we can call it a Robin Hood thing because Spotify screws over its artists anyway).
  • I also found a hacked ad-free Spotify app someone developed, and I installed it on my phone. Again, without incident, years.
  • I joined a friends-and-family-style phone plan with people I found on Reddit, whose names I don’t even know. Every month I Venmo money to a guy whose face I’ve never seen, and he pays our collective phone bill. This has been working without incident and saving me money for years.
  • I met a guy in the comment section of a blog I like who was interested in cross-editing each other’s writing. I shared some of my drafts with him, and he gave me good tips. Now we can use each other for that sort of thing at will.
  • I want to say, “I engaged in a lot of online conversations over a period of ~3 years, and it fixed a lot of my beliefs and made my worldview really good,” but then I’d have to explain my whole worldview in order to prove that (and you’d have to agree with it). Instead I’ll say some more objective things about it: I’m rarely ever surprised by the behavior of other people (I used to be more surprised); I can understand and empathize with a huge variety of people (I used to be less tolerant); I know a lot of specific things I want to do with my life (I used to be more vague on that question). I got all this by talking and listening to smart people online, for free.

Some of the above might make me sound reckless, but in all honestly I’m a very risk-averse person. I felt very confident in each of those choices, believe it or not. There have been lots of Internet-mediated opportunities that I didn’t take. I encounter just as much BS as you—”shitcoins” being pumped, trading strategy scams, start-your-own-business scams, malware, etc. I have never had my identity stolen and never paid money for a course or service that I regretted.

I think the reason I’m good at this kind of thing is because my life is (to a greater degree than average) full of strangers. My job at a big tech company doesn’t exactly surround me with an intimate community, especially now that I’m a remote employee. My two most valued hobbies—digital art and writing—are mostly done solo. Plus, I lived far away from my family for nine years: I moved halfway across the country for college, then I moved the rest of the way for work. Twice I moved to a brand new city with virtually no roots. That’s not necessarily good—I’d rather be surrounded every day by people I know well. But that comes with time (specifically it requires staying in one place for a long time, which I now plan to do, having moved back to the east coast). I think those circumstances have given me a lot of practice in evaluating people with minimal context.

Leveraging knowledge

First off, probably the most reliable way to discern people’s credibility on a given subject is to learn some of the subject yourself. Then you can judge their thoughts on one or two specific questions, and extrapolate that judgment to the rest of their messaging. I came across a blog post, What Money Cannot Buy, that speaks more to this.

The more you know about a subject, the better your judgment will be, but of course your time is limited. The reason experts are valuable in the first place is because they save you the time and effort of doing all your own research on things.

So for example, I’ve got my cool expert gym owner friend now, but I did know a bit about lifting and nutrition before I met him, and that knowledge informed my opinion of him early on. Same with most other Internet people I’ve trusted—you need to start with some knowledge of your own. Then it’s a game of leverage: use your little knowledge to figure out whom to ask for more knowledge.

Reading signals of credibility

The signaling of credibility is a tricky thing. Credibility is something that nearly everybody wants to claim about themselves, but signaling it directly doesn’t work. We might say it’s at the halfway point in its symbolic cycle, where signaling the thing actually makes people expect the absence of that thing: framing oneself as very knowledgeable and serious leads the audience to expect an amateur or a fraud.

That’s also why some people earn perceived credibility while actively signaling that they don’t deserve it.

  • I’ve read some C. S. Lewis and I recall him emphasizing that he was not a theologian and was not a man of outstanding intelligence, just a common man with thoughts. Today he is more widely read and quoted and admired by theists than many (if not all) of the “real” theologians he was referring to.
  • From a completely different direction: comedian Bill Burr often wades into political/cultural issues and then quickly follows up with “don’t listen to me, I don’t read.” His passion is comedy; he doesn’t want to become one of the mainstream thought leaders / cultural commentators whose opinions he tears apart in his standup specials. Yet, he has a devout following of fans who not only love his comedy but, if you look closely, admire who he is as a person and the values he espouses.
  • Jon Stewart had a similar dynamic as the host of the Daily Show. He did political punditry / cultural commentary, and unlike Burr he did want that commentary to be taken seriously, but he never signaled seriousness directly; he only ever countersignaled by emphasizing the fact that his show was comedy. “The show that leads into me is puppets making crank phone calls.” He gained a seriously dedicated political following from that platform, and around that time a meme started circulating: “I watch Comedy Central for my News and Fox News for my Comedy.” That’s exactly what I mean by the symbol being flipped.
  • I’m sure can think of your own examples—people who only ever deny their credibility, yet thousands of followers are hanging on to their every word.

What then? Do we just look for people who say they’re not experts? Maybe that’s one thing to look for, but of course it can easily be gamed. And when it comes to writing stuff on the Internet, any signals can be gamed. What I’ve found myself doing, when I make these kinds of judgments, is adding up a lot of smaller, more subtle signals.

So here are the signals I use to figure out who’s credible on the Internet. More generally, these are signals for figuring out who’s giving the right answers, when you don’t already know the answers and you don’t know anything about the person. All you have are the messages they’re sending. A tricky problem! I count each of the following as partial evidence of credibility.

Vocabulary

  • In any field, there’ll be unique terms that come from older canonical bodies of work or from experience in the field. The correct use of these terms is a signal that the person has read the works or had the experience. Example:
  • If their writing is unnecessarily wordy, or if it uses any of the empty pseudo-professional tics that George Orwell thoroughly disparaged in Politics and the English Language (worth reading), that’s a bad sign. It indicates that the writer is explicitly aiming to sound highbrow and important—yet if they were writing with substance and experience, they would’ve implicitly achieved the same thing.

Tribal affiliation

  • Is the person signaling membership with a group of people that I’ve previously concluded is misinformed? This can be a dangerous heuristic because it can easily turn into, “They sound like they’re on the opposite end of the political spectrum from me, so I’ll ignore what they’re saying.” And, I realize I’m making a political statement in saying this, but: I’ve found that the typical red-blue distinction is way too coarse to determine whom to listen to about various topics. I will however make stricter judgments if I know the more particular subgroups that the person belongs to. You can make inferences based on categories, you just have to be careful. There are some cliques and communities that really do seem to be worth ignoring most of the time, and you’ll identify their members by their vocabulary and by their views on certain token issues. I’ll have more to say about tribal affiliations in a future post.

Vibe

  • Do they lose control of their emotions? This is easy to see in the context of a disagreement. Do they resort to bitter sarcasm or name-calling? And sure, a person can be right while being emotionally immature, but I generally expect the opposite correlation. The conditions that cause a person to say false things are also conditions that spike emotions: they could be a fraud and be afraid of getting exposed publicly; they could be holding their views for reasons of personal identity and therefore feel personally attacked by disagreement; they could be mentally ill.
  • Do they try too hard to convince others of their views? There comes a point in online disagreements where it’s just not worth it to continue (I talk more about this in Choosing Battles on the Internet), but some people chatting online go way beyond this point. Or they force their topic into a conversation that wasn’t related, and then try to argue their views with anyone who bites. When I see this behavior, I sometimes assume it’s that classic dynamic with proselytizing: they are insecure in their own opinion, so they have a psychological need to see other reasonable people be convinced of that same opinion. They make their interlocutor into a kind of authority who can validate them. These aren’t the kinds of people worth listening to.
  • Do they try not hard enough to convince others of their views? There’s really two specific indicators I’m talking about here:
    • There’s a certain kind of person who will write six dense paragraphs of text, frantically plowing through a variety of topics, metaphors, and examples, and then say something to the effect of, “I know I’m rambling here, but I just think it’s really important that…” and then go into five more rambling paragraphs. This is unserious communication. I always stop reading at the “I know I’m rambling” line, because it’s a sign that the person doesn’t care enough about their ideas to edit them into something readable. I guess that their act of writing is more about blowing off steam or organizing their own thoughts than about spreading true ideas.
    • There’s also the postmodern appeal. So far, all of my tips have been meta: they’ve been about how people deliver their message, not about what the message actually is. But this item is about a specific message. When people deny the value of discourse by saying something like, “Humans don’t naturally find truth in a free marketplace of ideas; they’re doomed to be misinformed,” there is self-evidently no reason to continue listening to them. What value is their contribution to the discourse, if they themselves don’t believe it has value? I’d guess usually there’s a hidden assumption that “I can tell what’s true and false in the discourse, but the unwashed masses cannot,” but if they won’t admit and justify that assumption, then I won’t credit it to them. If you don’t believe free information tends to lead to truth, then you shouldn’t be looking for truth or expertise online. You shouldn’t even be reading this.
  • Know the stylistic personality markers! All caps: they’re angry. All lowercase: they think they look cool. Title case: they’re unintelligent. No punctuation: they’re reckless. No paragraph breaks: they’re psychotic.

Spelling/grammar errors

  • A person’s ability to correctly use your/you’re and its/it’s and their/they’re/there is correlated with the amount of professionally edited text they’ve read in their lifetime (books, research papers, formal essays), and that is correlated with how much they know. If I’m reading a long post and it has sentence after sentence without any spelling or grammar errors, that’s a rare occurrence and a big green light. It stands out to me, and I lend it credibility.
    • Obviously this is true for many kinds of spelling/grammar errors, but the above examples stood out to me as rules that have no ambiguity, are never violated in professional writing, and are often violated in informal writing.
    • An exception is when you see spelling and grammar errors that have the characteristics of being made by a non-native speaker. There are many errors that only non-native speakers make, so you should know them when you see them. In those cases, spelling/grammar errors don’t tell you much about the competence of the person behind them.

Intellectual honesty

  • Similar to the outright denial of expertise we looked at above is the admission of mistakes on specific questions. When a person publicly admits they were wrong about something, that suggests 1) They’re honest, 2) They care about the truth of this particular topic, and 3) Their ego is not attached to the idea of being right on this topic, which means they can look at it dispassionately. All good signs.
    • In the world of financial advice, this translates to investors and traders posting their losses.
    • In the world of rationalists, this translates to people highlighting their failed predictions.
    • A caveat: Remember that signals can be gamed. I happen to believe that in the world of politics, this one is sometimes gamed. A pundit “switches sides” and starts making all the [new team] talking points as a former [opposite team] member, so [new team] rallies around their perceived intellectual honesty and buries them in attention and money. I don’t think this is always sincere on the part of the switcher.
  • For many signals of intellectual honesty, you could skim through the Street Epistemology guide. It’s intended to tell you how to be intellectually honest, but it’s just as useful as a field manual for finding such people in the wild.

Online reputation

  • Have they been on this platform for a long time? Have they contributed a lot of content, indicating that they care?
  • If the platform has a “karma” score, that’s an easy way to check how much the community appreciates them.

Remember, these are all partial evidence. Just like you automatically combine in-person cues (tone, body language, facial expressions) as indicators of whom to trust, ideally you can use the above in a similar automatic combining process.

Once I was reading someone’s comments that were very wordy and sounded sort of pseudo-intellectual, but not quite. Part of me didn’t want to dismiss it, because the content seemed detailed enough (jargon) and was presented dispassionately (intellectual honesty)—so I had some conflicting signals. I looked through their comment history, and it turned out they had grown up speaking German (very wordy). That explained away their writing quirks, and so I trusted them more.

Wrapping up

I hope this post can help augment your natural “credibility detector,” so you can use it to greater effect in the strange ocean of the Internet. Or, at the very least, I hope I’ve convinced you that there can be a lot of value in engaging with to Internet strangers, and that this is a core part of what makes the Good Internet good.

pdf

Share:

Leave a Reply

Your email address will not be published. Required fields are marked *