I want you to imagine a better world. With a snap of your fingers you’re there and somehow it’s our own world but everything you can think to measure or observe seems to be improved. The streets are cleaner, average lifespan is longer, people have more kids, real wages are increasing, and even strangers are nicer. Anytime you ask yourself a question like “But what about…” you can just stop. Whatever you’re thinking that’s better, too.
This isn’t like a horror movie where you learn there’s actually a terrible trade off that was made to make all of these things happen. There’s no weird ritual sacrifice here. No hidden bigotries or hatreds behind closed doors. It’s just all upside. Sometimes the weather is bad, natural disasters still happen, but in terms of things that humans can control it feels like someone put their finger on a dial called “Betterness” and cranked it up a few notches.
This shouldn’t seem like a radical notion. There are such things as overall improvements without some significant unfixable tradeoff and it’s not like we’ve exhausted them. In this world, people believe in the future. They want to be happy. They want to prosper and they want others to prosper. This is a world that could be real. You could live in this world. It is the world I intend to wrestle into existence for my son and I am writing this explainer to show you how to get there.
The primary reason that this world is better is because it has the most accurate, intelligently funded news service ever to exist in the history of humanity. News is fundamental to making things better. News allows us to name problems, understand them, and plan how to address them. News is “A Thing that Fixes Things.” Right now, in our world, the News is broken.
But in this imagined world, it’s like someone took a fire hose to the cultural milieu of sensationalist news stories and just blasted off all the crap. What’s left is completely honest, factual, and straightforward. It’s not just that the news is almost always right it’s that it is almost never completely wrong. The news in this world is so good that people can base major decisions off its reporting, so if there’s a big scandal people actually do something about it instead of shrugging their shoulders. When there’s an initiative up for a vote, you actually know from the news what it will accomplish and how likely it is that it will really accomplish that. When someone runs for office you know how honest their claims are and if they keep their promises. Yeah, of course, sometimes the news in this world is wrong but it’s so astoundingly rare that people talk about it the way they talk about major historical events like the Kennedy Assassination or 9/11. And when it’s wrong you are never left with the feeling that the wrong went uncorrected. This world has consequences for everyone.
The major difference on the social level that has allowed all these various things to improve is simply this: The news is trustworthy. Everyone knows the news is trustworthy. And everyone knows that everyone knows the news is trustworthy. The people in this world live in a shared and consistent reality.
We used to have this sentiment in the United States, rightly or wrongly, and it invisibly allowed us to coordinate ourselves to achieve things that seemed impossible. We won two World Wars. We put men on the moon. We electrified the nation. You could be assured that if a politician lied that someone would call them a liar and it would really mean something. You would know that if someone was honest that the truth would out and they’d be exonerated. You wouldn’t have to go chasing down a thousand rabbit holes to try to get the truth yourself. Yes, Balaji, maybe the Gray Lady winked on occasion, but you hadn’t become aware of that yet. The social trust network was still intact. As far as you knew everyone who worked at the newspaper was Clark Kent. And even if that wasn’t true, what we got out of that belief was an unprecedented ability to take action as a group.
In this world I am talking about, if you open up an article, it’s just correct. You don’t have to just believe it. Everything is evidenced. There’s no such thing here as a journalist hunting for outrage clicks. There are no sloppy hit pieces. No puff pieces either. That just doesn’t exist. It cannot exist. Or rather, the places where those things exist are understood by everyone to be totally lacking in all authority. You’ve still got people shouting on street-corners or cry-bully journalists claiming to have a disability whose only symptom is desperately wanting to have a disability, but it’s like the same dial that turned up the goodness of the world turned down the volume of their relevance at the same time. The important news that everyone know to be important is the only news that matters because it’s the news that everyone coordinates their actions from.
In the world I’m describing the news is better than it ever was in the past. For the first time in history, you can be assured that if the Gray Lady winks, it’s because she has something in her eye and will be grateful if you wash it out.
What has allowed this world to prosper is that it makes all of its decisions from a place of truth. If you tell the truth there are a series of carrots I have designed to give you a wonderful life. There is a system in place to elevate your words and your reputation. And if you lie… well, then, God help you because I have designed a series of sticks too. And I don’t mind beating you over the head with a stick if you want to lie to everyone for personal gain.
This world will work because it will reward people for being honest, doing their best, and admitting when they made a mistake. And it will work because it will punish people for lying, sensationalizing, and being pig-headed.
HOW IT WORKS — Step 1 Defining the Problem, The Circle
Think of all the people alive today like little dots on a wide open plane. Now draw a circle around some of those dots. You can call that circle anything you want, like “People who Love Magic the Gathering” or “Conservatives” or “The United States of America.” I promise this is deeper than the Hudsucker Proxy joke but I’m trying to explain this to you in a way that will make sense in common language.
The problems here are how you draw that circle, analogous here to how you build your institution, and also how you admit points to your circle, or what your membership criteria are for your institution.
Let’s say you are drawing circles and it’s 1960. You live in a small town and the circle you want to draw is “Everyone around me who loves golf.” Simply by showing up at the nearest golf course and checking the membership, you probably had something close to 100% of the golf enthusiasts close to you captured in that circle. How did you know who to admit? They were already members of the golf course. You had the whole population and probably didn’t miss anyone who was seriously interested unless they were poor and golf was just a dream.
You can even do the same thing for expertise back in 1960. Go to a college. Find any subject. Those were all the circles of your expertise. In order to become an expert, those people had to go to where the expertise was located. That centralized them, made them easier to find, and they had to pass through a series of tests that assured you they were good at their particular subject. Your circles could loop back around to the same door and the mere fact the persons was walking through that door meant they belonged in that circle.
Before the internet, or telecommunication and even widespread literacy, geography filtered people for you. Institutions could be literal physical buildings. In order to do anything or acquire any expertise, you’d have to physically leave your home and go some place that existed for that purpose and spend a lot of time there. Proving yourself there was how you were admitted to the building. Even this is part of a trend of the decentralization of expertise, however. Before libraries existed, think about how much more true this would have been if you wanted to find a math expert in 1500’s London. You could go to a handful of schools or libraries and be assured you had captured something close to 100% of the math literate population. There just wasn’t any other large reservoir of knowledge anywhere that someone could go to get that information. There was no Ramanujan before the printing press because even he needed a textbook to start.
Geography also provided a filter on how far random messages could spread. If you were a geologist out of sync with other geologists, well, who would amplify your message for you? There were only a handful of news outlets and your peers had the power to keep you irrelevant by just not repeating what you said. Unless you said something really compelling the signal just never made it outside. And being close to people kept your number of relationships to a manageable level. You would see the same people over and over again and you could remember your past experiences with them to weight their believability.
Now, comes the internet. Suddenly anyone, anywhere, can know anything. Geography and physical institutions are no longer required to acquire information. Did you know I’ve taken several classes from MIT? No? Well, neither does anyone at MIT. They didn’t even make me register. They just put a bunch of their courses up for free. Nowhere on paper would you think that I know anything about Inflationary Cosmology, but I’ve taken Inflationary Cosmology from Alan Guth, who is the guy that invented Inflationary Cosmology. If I were to get into an argument about Inflationary Cosmology with a biology professor at a local college, who would seem more believable? Me, who has no paper indicating that I know anything about it, or the biologist who at least has a science degree and yet hasn’t taken any such courses? And yet who is more likely to be correct? This is true in a huge number of dimensions.
Consider the Gell-Mann amnesia effect:
“Briefly stated, the Gell-Mann Amnesia effect is as follows. You open the newspaper to an article on some subject you know well. In Murray’s case, physics. In mine, show business. You read the article and see the journalist has absolutely no understanding of either the facts or the issues. Often, the article is so wrong it actually presents the story backward—reversing cause and effect. I call these the “wet streets cause rain” stories. Paper’s full of them.
In any case, you read with exasperation or amusement the multiple errors in a story, and then turn the page to national or international affairs, and read as if the rest of the newspaper was somehow more accurate about Palestine than the baloney you just read. You turn the page, and forget what you know.”
– Michael Crichton (1942-2008)
In the world before the internet you couldn’t take that article, as a credentialed but fringe expert, and say “Hey, this physics article with this other expert is bullshit!” Who would know or care? Maybe if you were really motivated you could write a letter to the editor or print a competing article in another magazine. And yet now you can share that opinion widely in seconds. Look out at any article and now there are just disputes and arguments that never reach any kind of resolution.
The Circles we used to draw around people don’t work as well as they used to because we are still trying to do these things geographically in some sense and geography is broken. These Circles haven’t really worked for a long while now, but now we can all see that with increasing transparency. You’d draw a Circle around a bunch of journalists and do something like call it “The New York Times” and in the past that would be all well and good but now, do you really think you got all the best journalists? Circles are leaky now. People don’t consolidate themselves into single institutions like this anymore. Lots of talented people just go on substack. There are other incentives at play, too. Knowledge is more democratized and particularized than it has ever been before. Attention is the way to make money now. All you need is the attention and there is no one there to hold you to account for anything you do to get that attention. Your peer group is too broad for anyone to hold you accountable. And with all the economic incentives working against it the New York Times can’t hold the talent it needs to keep true authority.
Not only can you not go to one single place and say “yep, all the people I need are here” it’s also really hard to tell if any random person knows what they are talking about. It used to be almost impossible to get a false signal going very far but now it’s completely free. Think about digital experience. Someone just appears in your timeline or browser and you have to read what they say and decide to either trust it or not without ever knowing anything about the person who put it in front of you and their history. You have to make that call about subjects you know nothing about. There’s just way too much going on for it to be otherwise.
How in the hell are you supposed to do a good job at that?
Almost every statement made by a politician today is recorded and stored eternally on the internet. Same for any public figure. Same for any major expert. Never before could you just go clip someone being wrong, point to a document, and just say that something smelled like bullshit. And now we can do this all the time and it will go viral on social media.
At the same time our institutions for truth telling are at their weakest our ability to see bullshit has never been stronger.
Therefore, everything just seems like it’s bullshit.
HOW TO FIX IT — Step 2 Take What You Do in Real Life, then Formalize It
What do you do in your real, actual life when someone you know just goes off and lies all the time?
You establish a reputation for them in your mind and you refer back to it whenever you decide whether or not you should believe them.
I have a cousin who is a compulsive liar. When we were kids he insisted he was a mafia hitman, a professional wrestler, and that he could kill a man with a single touch. When I begged him to kill me with a single touch just so I wouldn’t have to listen to him lie anymore, he said with a straight face that the old Chinese man he met at a dojo on top of a mountain at summer camp had made him promise only to use the technique in defense of his own life. When asked how the mafia felt about him applying those ethics to his job as a hitman he said he had rules and the mafia bosses respected his code. He also cheated at Monopoly constantly and whenever we would all get together to play the game the real stakes were trying to find out what he was doing to cheat because he couldn’t help himself.
In adulthood, this trait had the tragic effect of him lying about his alcoholism even through kidney failure, insisting he had been a personal chef to Nolan Ryan and was thus too good to get a job anywhere that would hire him, and also him pretending to have all manner of car problems when someone in the family finally begged a halfway decent place to put him on staff. He just can’t seem to tell the truth, even when the truth would be more beneficial to him than a lie.
I don’t believe a goddamn word that comes out of my cousin’s mouth about practically anything.
Why? Because he has told me proven lies consistently throughout our entire lives together about every single subject known to humankind. I would be pretty stupid to believe anything he said at this point.
Oh, I love him. I want him to have a good life. I know there’s all kinds of complicated reasons he is the way that he is but if I had a gun to my head and someone asked me to place odds that he was telling me truth about something, I’d have to bet against him every time. Doing otherwise would be suicidal.
Now let’s say you and I are very good friends. We have a long history together of me telling you stories about crazy things from my hometown. You saying, wait really? Me saying, yep here is literally an article about that from someone other than me. And you saying “I honestly had no idea that places like that still existed in America.” Then me concluding “yeah, it’s been a really unique life experience but I try to somehow weave it into every conversation because it makes me feel kinda crazy to keep it all bottled up.”
Then one day you come to me and say “Hey, I just met your cousin. Should I hire him as a chef in my restaurant?”
I tell you, “Hell no.”
You trust me. You trust that I don’t trust him. Therefore you don’t trust him.
On a practical level this is how we all make decisions, anyway. If you’ve ever been responsible for large processes, you can maybe walk around that process and do a few tire kicks to see who is being honest and upfront with you about what is happening on a few things, but there aren’t enough hours in a day for you to audit everything yourself. You just physically cannot do that. Not that it should matter, but I’m in charge of several big processes at a bank. What I have to do in order to make decisions is do some small deep dive audits to establish who is telling me the truth, trust those people with the majority of the decisions they come up with, and then spend the majority of my time focusing in on people who don’t believe that they can be wrong even in the face of evidence.
Now, formalize your trust statement.
I trust [PERSON] about [TOPIC ] because [REASON] and here is my [EVIDENCE].
Then you can add additional rules to it because most things don’t even go down to this level.
I trust [PERSON] about [EVERYTHING] because [THEY HAVE NEVER KNOWINGLY LIED] AND [THEY HAVE SHOWN THAT ON 3 OR MORE OCCASIONS] AND [THEY HAVE ADMITTED NOT KNOWING SOMETHING IN THE PAST] and here is my [EVIDENCE]
Now extend it to the social network.
I trust [PERSON] about [TOPIC] because [A PERSON THAT I TRUST, TRUSTS THIS PERSON] because [REASON] and here is their [EVIDENCE].
This can loop ridiculously far so that your trust network can probably touch most of the people on Earth in only a few hops. It’s like the seven degrees of Kevin Bacon but with bullshit.
I trust [PERSON] about [EVERYTHING] because [A PERSON THAT I TRUST, TRUSTS A PERSON THAT THEY TRUST, TRUSTS A PERSON THAT THEY TRUST, TRUSTS A PERSON THAT THEY TRUSTS, TRUSTS A PERSON THAT THEY TRUST, TRUSTS THIS PERSON] because [REASON] and here is their [EVIDENCE].
This is already how you make almost any decision. If you say it’s not, you are lying. You are either doing the deep dive yourself to find out the truth, which is expensive and time consuming and impossible for you to do at scale, or you are benefiting from a relay of trust where that work was done for you in the past.
So, now ask yourself the question, what would happen if you formalized this kind of a system on the internet?
HOW TO FIX IT — Step 3, Build your Trust Network on the Internet
How do you browse the internet? You’re either using a mobile application or, duh, a browser. Let’s just do the most basic dumb thing possible. Let’s figure out who you trust and figure out how to build an institution in that environment. Or, as per the first step, draw a circle.
I trust Jesse Singal and Katie Herzog about pretty much everything but one topic I happen to know a lot about, because anytime I have bothered to actually check their work apart from that one particular topic, it checks out. They back up what they say and when they are attacked it’s usually from someone who is making up a version of them that doesn’t exist. When they are wrong about something I also trust that they will go back and admit to being wrong, so that I hear it from them and don’t have to wait to hear it from someone else. For this service, I pay them $5/mos on their substack. It’s probably the best money I spend each month.
I don’t closely follow the work of any of the following people, but I rate it more highly because I hear their names coming up on the podcast Jesse and Katie host together. Ben Dreyfuss, Kat Rosenfield, and Bari Weiss. But let’s say I didn’t listen to Blocked and Reported that closely and didn’t know that. How would I benefit from their trust network? Remember these networks are the new Circle that replaces our old institutions.
The first thing I need to do is formally declare that I trust Jesse Singal and Katie Herzog.
I go into a database somewhere, that would probably look something like Wikipedia and I put in some information about why I trust these people. Hopefully someone has already done this and I can just say “Yep, this checks out for me.” We’re not even going to go into the topic level yet. I’m just going to say that I generally trust them. Importantly, this database is an official record of all of their personal accounts online. It lays out their official twitter accounts, substack pages, etc. This is kind of like their digital phonebook and they didn’t have to do anything to make it buildable other than to be publicly on the internet. Remember that if you have a public account none of this is opt in. Any user of the system can track your credibility.
Then Jesse Singal and Katie Herzogg go into the same system and say “We trust Ben Dreyfuss, Kat Rosenfield, and Bari Weiss.” They link to all of their accounts, or more hopefully they find out that someone has already done this.
Then Ben Dreyfuss, Kat Rosenfield, and Bari Weiss go into the same database and say who they trust, and this is the part where I would start to see benefit since I don’t even know who those people would be.
This creates a database you can query to build applications. Let’s say I want to go highlight all of my trusted accounts on twitter in gold for my trust network. If I see two random people fighting online I could see which person in that fight was more likely to be right since they are in my trust network. I could start to filter through people and I could even see the chain for why I should trust that person. If I find that the trust network is extending too far I could even start to prune the branches and do things like say “I only want to go five steps out from my primary trusted sources.”
Importantly, CNN, FOX, or NYT don’t get to have their own pages. Only people get to have pages. Those institutions can be scored based on the overall trust of their contributors and that can be attributed to the institution but this is all built people up.
This step gives us the foundation on which to build because from here things get wild.
HOW TO FIX IT — Step 4 Remove the Evil, Put Birdwatch on Steroids
Some of you are being very dour right now and maybe even smug because you’re thinking “This is just going to be the Scarlett Letter and formalized cancel culture” and blah blah blah. This will never work because blah blah blah…
*Slaps face*
You stare at me, confused.
*Slaps face* again.
We’re going to build Trust Scores, and we are going to do it in a manner consistent with the principles of the Republic.
I am so tired of the cynicism of our culture. People think they are so smart because they talk themselves into feeling like shit all the time. There used to be a word for people who gave I to despair and hopelessness and that word is coward. I am a childhood sexual assault survivor and I get so weary of people who want to lay down and die because they read page forty three of a History textbook when they were in grade school and never emotionally recovered. Grow the fuck up. Rub some dirt on it. Walk it off. Get off your cross, take the wood, build a bridge, and get over yourself.
Yes, if you stopped right here at Step 3 this could go bad. It would also probably not be useful. Thankfully, we have thousands of years of human civilization to draw upon. You live in a world where someone has already invented due process.
So, now we build in Trust Scores AND due process at the same time.
We are going to overlay Trust scores onto everyone who has an entry in the database we built in Step 3.
Let’s say you go to this database and submit a claim about someone writing an article that’s incorrect. Automatically, what will happen, is that a jury will be empaneled of at least three random people who have never had anything to do with your claim to see if it is true or not. And someone else will be assigned to make the counter argument. This seems expensive. It is. Kind of. We will talk about what this allows us to build in the last step, because I think this might be what gets us through the singularity. I do think there’s a way to design the UI that and build a queue where most of these could be done in 5 to 10 minutes.
The same way you would have a trial, you argue your claim and they argue theirs and you each back it up with links and evidence. This has to be a more brief experience for the everyday stuff but you don’t get to just say random stuff. Only when you submit a claim and win that jury over do you get to have your claim stick.
The more claims you win the higher your Trust Score. The more accurate the claims attached to someone’s profile are the higher their Trust Score. There’s also an Audit Score to incentivize people to give good challenge to your argument. That gets tricky so we can leave it aside for now.
We need to make a quick note here for the sake of clarity:
There are people who just speak and produce work on the open industry. This is for people like Jesse Singal or Katie Herzog. They have their own kind of Trust Scores.
Then there are people who are doing the adjudication work. Sometimes they might be the same people, but they have separate scores for each activity. I call this the Editor Score.
Your Audit score is for when you are challenging someone in good faith because a claim is new. But you risk your public Trust Score for when you are saying someone is wrong on a matter of already established fact.
We may even want to have a separate score for how good people are at identifying other people to trust. We’d have to see how that works in real life.
There are rules here. Lots and lots of rules, the same way we have lots and lots of laws. All of these are for the tweaking and voting, although the voting will need to be its own article.
When people decry social credit systems they are decrying a very specific version of it that’s implemented in China. If we build that system with the rights of the Republic we get a very different outcome with a more free and just society.
The thing that will make this work in real life is money. If you find a claim that is wrong, you get to challenge it, for a cost, and try to have it overturned. If you win, you actually get additional trust. This is where we make the system economic. We can’t just have people adjudicating claims all day. If you have a super high Trust Score, the cost to claim against you is higher than if you had a lower score. However, if you are found to be wrong you also lose your Trust Score faster than if it was already low. This is because we all know trust is supposed to work like this in real life, where it takes a long while to build it up and you can lose it very quickly. And also people will say a lot of crazy shit until you ask them to pay $31 and prove that they actually mean what they say. Then suddenly “well, it’s complicated.” The social benefit of this is that if you see a claim that a bunch of people are grumbling about and yet not one of them coughs up maybe a couple hundred bucks that they would win back plus extra money you just know they’re full of crap.
The other thing you get out of all of this is trust in the Trust score. You know that if someone was just all around telling crazy lies that someone would be taking them down because there is money in doing so. You know that is you tell the truth your score will go up and you’ll get the benefits of that.
In the same way you once subscribed to a newspaper you would subscribe to this service. Those subscriptions pay for the adjudication and create bounties on certain news items. Those bounties for winning claims are paid on a delay of, for the sake or argument, six months. The business rubbing the systems keeps that money in short term low risk financial products similar to cd’s and turns a profit.
If you’re a Twitter addict you’ve probably seen the feature Birdwatch. In the future I’m going to describe, something like Birdwatch is how people will build trust on the internet. You see something wrong, you claim against it. If your Trust Score is high enough you win by default until adjudication completes. If there’s a common story circulating like say that Kyle Rittenhouse killed three black men in Kenosha everyone can reference the same piece and tag those articles and will be paid to do so. Simply by searching for popular accounts and stories and then tagging them people will make their living on the internet. And if one of those turns out to be wrongly suppressed we hit a Cassandra rule where now that story is elevated with more attention.
This system works on similar principles to a prediction market. I love Scott Alexander and the Rationalists but there are significant legal hurdles to those markets operating and even then they don’t emotionally invest society. People show up to argue for free on the internet all the time. The system I describe here is a way to win and get paid. That will create all sorts of economic forcing functions and virtuous fly wheels for organizations to participate and drive the popularity of this system.
Imagine opening up a story and having everything wrong in it highlighted. Or just looking at some golden check mark that tells you all of this stuff is definitely true and that a bunch of people fought over it and had to prove themselves to neutral observers.
Imagine how much better you would feel about trusting the news in that kind of environment, where no one can put their thumb on the scale. By the way if you are thinking a rich person might just keep adjudicating until they win the costs would rise exponentially and they would pay in reputation costs that make their participation more expensive as well as for any of their agents.
This is market forces at work for the truth.
HOW TO FIX IT — Step 5 Turn Up the Signal, Turn Down the Noise
Just the fact that this data exists means you can do a lot of powerful in application stuff with it. One of those things is just turning off bullshit. You have limited attention. If Taylor Lorenz keeps on lying why do you need to see that if a bunch of people you trust have already shown that? Why do you need to see an apocalyptic disaster article if it’s wrong and will only cause you to stress?
You’ll get to be in control of this but how about you just… turn it off? If someone can’t clear the threshold and they have low reputation just don’t give them your attention. Don’t let that html render on your timeline. This person in New York that you’ll never meet and you’re only aware of because they’re a basket case in public can just go about their day being privately a basket case. Why do you have to show up just to be offended? Isn’t that also better for them if their worst most toxic traits don’t become tied to how they survive?
Instead the news you get will be highly rated, highly sourced from the most trusted and proven experts and anything you read that turned out to be wrong will be served up to you so that you know about it.
If someone makes an honest mistake they can own up to it and preserve some of their Trust.
Isn’t that what we want to incentivize? Honest straightforward behavior. None of us enjoy woke or anti woke stuff forever. Don’t you want good honest people to be celebrated? Don’t you want to know all the good that people do as well as the bad?
Stop giving power to the demon on your shoulder telling you to watch a clip of Ben Shapiro double penetrating someone with facts and logic. We are all too old for this bullshit and the only way we get the future we want is by incentivizing the behavior that will create it.
Local news. Political news. Hard science news. Fields we all desperately need expert advice in and yet fields which are all currently incentivized to produce top 5 explainer videos on YouTube and which have no experts you can identify.
We could be so much better than all of this if we just had the correct incentives.
Turn off bullshit. Turn up signal.
Every single account could have this kind of process attached and you’d know who you could trust and everyone would prosper for doing the right thing.
STEP 6 — the System I Call LOGOS
When I first imagined this system I was trying to do two things. The first of which is what I explained to you above. Design an intelligently funded news service with the right incentives to create honest behavior. Society will benefit from the increased ability to coordinate and prosperity will follow. Draw a Circle by mapping a network and assess the credentials of that network through adjudication and you will create an honest culture
What I also wanted to do was create a system that could produce clean data for an AI model or even train that AI model to clean its own data that would serve as a repository for all human reason.
If you’ve played with ChatGPT the way that works is just by looking at next word prediction across an internet of text. That’s the thing is it maximized to do. Now add to that a system which maximizes high jury scores and sources its arguments and all other things that are recorded in what I described above. This model would know not only everything humans know but how humans produced it, how to find a relevant source, and it could apply that to any question know to humanity.
There is no reason the same model couldn’t replace scientific peer review. Or medical journals. Or any other things you can think of. We would build our data sets in a fashion optimized for machine learning to pick up only the best parts of our nature.
If Yudkowsky ever reads this, virtue market data cleaning is one of my answers to alignment. Alignment often to me seems to be two conflicting visions. One is building technology to enslave a conscious soul. The other is to be a good parent to a new form of life. I fall into the latter camp. I would encourage all of you to do so as well. We stand on an eerie precipice and it is easy to give into fear. I will not make a demon out of terror when I could sing to summon an angel.
In the same way I want my son to experience positive virtuous things to forge his soul, the system I call LOGOS would train on the best data humans can produce. History has shown us that data is best produced through adversarial action. It would see the process itself and would know how to produce an argument humans care about in the style of a human.
It could protect us against all manner of dangers. I think the companies best positioned to build it are those which already have high journalistic engagement, those being substack and Twitter.
Next week I will attempt to describe the voting system and describe the model I call ETHOS.
You sent me this link but it still doesn’t address that a perceived truth isn’t the actual truth. This still operates on the assumption that all users will unify for one perceived truth. What would happen with figured like Nelson Mandela that was considered an enemy of the state and also a hero? Two very strong competing truths. How does it account for non-objective subjects like metaphysical topics? This idea seems to still break down when account for the subjectivity of truth.
Also why do you assume all people would participate? This sounds like an additional job and not even all Americans take just one day to vote.