Public discourse in the age of AI: Do we need to charge people for the ability to post?

Since this topic came up else where:

Many still think of paywalls as a blockade to read content.

But maybe we need to think about this differently. Paywalls, not to permit reading, but to permit posting.

If you want to write a comment on the German newspaper of record welt.de, you are greeted with this:

You have to signup and pay if you want to comment. And people pay for it, because they want to actively participate in the public opinion making process. This barrier to entry gives their opinion more weight. Even though it is small, it is certainly bigger than the random opinion posted with a possibly fake account on facebook.

Just a year ago the cost to interfere in elections and spread fake news with the help of bots and fake accounts was between 6 to 15 million Euro. Smaller campaigns would cost on the order of hundreds of thousands. Here is an example for a company that would offer these services: Team Jorge - Wikipedia


src

The cost to create such a campaign now is drastically lower. All that is needed to push an agenda is an LLM and residential proxies.

And residential proxies are super cheap and easy to acquire. All that is needed are simcards and a set of 4g modems in a major city:

Mobile IP addresses (4G, 5G, LTE) are practically un-bannable, because they are shared by up to hundred thousands of legitimate users in major cities. Instagram will never dare to ban 200.000 people in LA just because of some pesky spammers use the same IP! When Carrier Grade NATs were designed, the designers knew about this issue

source

So how can we decide what is real and what is fake?

How can we create a barrier to entry that is not easily gamed?

This is a very fundamental question that does not just affect countries and cities, but also crypto currency communities. Really any kind of community.

In the crypto currency world the pay off from pushing an agenda is very immediate. So much of the recent spam accounts are somehow related to it.

But there is also a payoff to manipulate the wider public discourse. To create fear in the population and guide the average person to specific solutions. It is a more subtle game that is hard to detect.

We are not in the same situation we were in when twitter and facebook started. The landscape is different now. I personally don’t find much joy anymore in chat groups that don’t have a paywall. Even a small amount like 25 dollars per year makes people behave so much more friendly and constructive.

In my mind there a two parts to this solution:

  1. Smaller communities
  2. Paywalls

Currently the newspapers of record take up too much space and authority in the public discourse. The quality of the resulting conversations often decreases with their size. The question is: How can we give discourse communities more legitimacy?

How can we convince users that it is better to pay for a forum than for twitter blue or similar big platforms?

Would be glad to hear your thoughts on this!

11 Likes

Well, this is a nice post and an interesting idea.

To start by stating the obvious, a forum needs participation. Ideally quite a lot of it. There might also be a point where there is too much. Starting a forum and making it get traction isn’t necessarily easy. The problem here is you might make finding participants more difficult. And getting the forum rolling initially far more difficult.

On the other hand, your idea may be beneficial for the “quality” of the participation. Although this isn’t a given either. You may very well end up with some people who don’t care about the money and consider they are now entitled to behave however (badly) they want. Some weird psychological effect of money on most people might favor the former rather than the latter.

I like this quote very much:

There are no solutions. There are only trade-offs. --Thomas Sowell

The real question is one of offer and demand. Why would people pay to post? You answered that, admittedly, but for a news portal where the content is created by the site and people pay to comment under it. If there isn’t any comment, or just a few, that’s perfectly fine. I even see more and more news sites simply not having a comment section anymore. If you would be willing to remove it entirely, going a paid route seems easy. For a forum, not so much.

Currently, most paying models in community sites are optional and often launched afterwards, when there is already a wide user base. It’s often a “premium” model.

What is really the incentive of paying to have the privilege of being the one creating the content too? :laughing:
Only being among other paying members and the overall effect it has on the atmosphere in that community may be an incentive. I’m not sure, but maybe that could work well.

One step further, which would be very suited with cryptocurrencies, could be to redistribute part or total of the proceeds to the “good content” creators. How do you judge about that is one question. The other big question is about the negative side effects such a model might have.

There is an argument people are already paying on many sites by being served ads. As often said “if something is free, YOU are the product”. Redistribution also occurs on many sites, but is usually also the proceeds from ads. Not money directly from other members, and not according to likes or such judgments, if you were to use that to determine what content is more deserving.

3 Likes

I’m not sure how charging for the ability to post changes whether not an AI is capable of writing those posts.

2 Likes

Yes, the correlation isn’t necessarily clear, but there are some aspects.

I guess the fundamental idea is something like: Anybody can now produce a post thanks to AI tools. Or some AI can even do it on its own. Paying implies somewhat a willingness to participate. A desire strong enough to put a bit of money on the table. Why would an AI do that? Why would someone just posting AI-generated text do it?

Of course it can still be done.
But I guess it would tend discourage multiple accounts, participation which isn’t “real”, etc.

2 Likes

The people behind the AI might be willing to pay for the ability of their AI to post. It depends on why they want to post and what they want to post.

3 Likes

Ok. Yes, some do. But probably less than the amount of people who will make an AI post for free. You then have less AI. Not necessarily zero, sure.

You can for example easily imagine young people wanting to try an AI they set up by modifying an open source project. Be it for fun, to learn, for a school project, or whatever reason. Paying is a small entry barrier. Free forum could see a massive influx of AI bots in the future. You even circle back to the “undetectable” problem raised in another topic.

Money might be an easy way to improve things. Sure, some might just pay.

The analogy has its limits because we’re talking of a flat fee per user and not a fee per post, which could be done too, but if email wasn’t free, there would be less spam. Some companies would also pay to send ad emails, but less.

EDIT: AI wasn’t necessarily the (only) focus of OP, IMHO. Despite its presence in the title.

1 Like

Might be worth noting that Metafilter is a site for public discussions which has been running since 1999 and has maintained a very high standard of discussion and contribution. One of their tactics is to charge $5 to sign up. (Another is to have moderators, and there are the usual complaints about moderation.) (Arguably another tactic is to have a relatively unattractive and old-fashioned site design.) (Also it is text-only.)

It might be worth distinguishing two kinds of things

  • spam that’s now easier because AI makes it easier
  • low-effort posts which barely add anything to the conversation because people use AI to make their contributions

Both of those are about the quality of discussion and contributions, although they have two different motivations. It’s already possible to spam, and it’s already possible to make low-effort contributions - AI makes it easier.

I’ve said elsewhere that I feel cheated when someone posts some AI content, even if they disclose it afterwards, because I’ve invested in trying to understand a point of view, presuming that it is held by someone who is thinking and feeling and living. An authentic post is worth reading, a fake post is a waste of my time.

4 Likes

I saw an interesting note today that AI’s are themselves unable to tell if something has been written by an AI.

Apparently if you quote Shakespeare verbatim and ask the AI if an AI wrote it, you can get ‘yes’ as the answer.

1 Like

I would be more concerned about ulterior motives behind AI-generated context than whether or not the AI has an opinion worth reading. (I can think of far too many humans who do not.)

It might be one way to flood a forum with content oriented towards a particular viewpoint, or with subtly disguised propaganda or advertising.

1 Like

I think this is one of those areas where it makes a big difference as to what kind of community you have and why they participate in the forum. For me, it’s niche hobby interests, so (for example) any kind of politicisation would be unwelcome anyway.

1 Like

Now you are totally making OP’s point.
It can be pretty effective to make as if a multitude of seemingly different individuals had that viewpoint. That doesn’t work on everybody, but does wonder on a lot of people.

OP raised how the cost and ease of doing this went down.
Paying to post would raise it again somewhere else.

I already feel the same when an individual writes something he doesn’t truly believe.

Thanks for the Metafilter example. I already saw this signup model a few times and it might be pretty good. Another model, which I guess is mostly to fight multiple accounts and re-registration of banned people, is for example used by bitcointalk.org : They attribute “evil points” to IP addresses and you have to pay to whitelist an IP you want to use according to the amount of evil points it has (one time payment, only if it has evil points)

1 Like

Charging for posting access might also have unintended consequences in terms of user behavior. The Freakonomics books have some interesting examples of this.

And then the reason for charging enters into the picture. Is it simply to discourage bots and trolls, or is the goal to help pay the bills or (God forbid!) make a profit?

1 Like

Can’t it be both?
Even if it is for the former, it will bring in money. So, it will automatically become the latter.

What are you talking about, exactly? Users commenting on the why of the fee structure?
It’s probably never a good idea of wanting to make it only about a reason which doesn’t include the fact you are collecting money in the end. That’s indeed a good recipe for backlash. Unless you redistribute 100% of it. Even then, people might have doubts about everything being redistributed.

[Crypto propaganda] If you have all transactions publicly available on a blockchain with amounts going in and out being 100% verifiable, you can easily argue you’re not keeping anything [/crypto]
There is nothing wrong about getting funds to pay the bills, or is it? If people are paying willingly and especially if they are receiving adequate value in return.

1 Like

For those of us who run forums for our own enjoyment, charging might change it from being a fun activity to being something we’re doing for the money. That might change MY motivation, if not necessarily the motivation of the user base.

I’ve run one particular forum since 1991, and over the years have had more than a few opportunities to try to monetize it. Others did it with similar ventures, some successful, some not.

1 Like

If that’s really your issue, you can just give to a charity.
Wikipedia needs funds. As do plenty of others. There is no lack of recipients if money is a problem for you. Hey, I might even be willing to help if needed :laughing:

1 Like