The Death of Digital Discourse

Or, Do You Really Want to Argue With a Robot?

We reach the conclusion of our journey through the digital world. We’ve discussed problems with it (one in specific) and how to make it better. But the consistent message is that it needs to change. And today, we explain why.

The 90s internet was a wild place that I often discuss. It was largely something we would now consider social media: bulletin board systems. Then it slowly developed into static websites (as well as forums) before it became static corporate websites. We lived in a world where Geocities existed simultaneously with American Online.

Social media increased participation on the internet, giving more people more opportunities to create content. And to a degree, this was great. But as it became the dominant form of internet usage, the core problem with it became clear: most people couldn’t create good internet content.

I’m of two minds on this. I’ve extolled the greatness of nobodies on the internet. At least one every five articles I proclaim that the death of the AV Club comments section was the greatest loss since the burning of the Library of Alexandria. Yet, that is the exception, not the rule. I just wrote 2800 words explaining how to make social media usable, which I would not need to do if most of it wasn’t terrible.

Recently I saw – on social media – a quote from the late, great Italian polymath Umberto Eco who said – well, wait, he’s better with words than me, let’s just turn this over to him:

Social media gives legions of idiots the right to speak when they once only spoke at a bar after a glass of wine, without harming the community ... but now they have the same right to speak as a Nobel Prize winner. It's the invasion of the idiots.

Apt. However, I both agree and disagree with this. To a great degree, many Nobel Prize winners – and other people of acclaim – were revealed to be idiots at the bar after a glass of wine. There are few polymaths, and fewer still who didn’t benefit from a filter between their words and an audience. But for the sake of brevity, let’s focus on the part I agree with. Most people can’t produce valuable internet content. If you’re complaining about most people on TikTok seeming dumb, remember, most people are dumb. What do you expect to be popular, a two-hour long explanation about Dante’s Inferno?¹

One of the dirty little secrets of the 90s internet is that it was better because the barrier to entry led to generally smarter people driving the information superhighway. Internet discourse benefited. And there was a lot of it.

You can act like you’ve been in arguments on the internet, but unless you were around 20 years ago, you have not seen what I have seen. Oh, a heated thread on Reddit or mean replies on Twitter? You have not seen what I have seen. I fought in the capitalism vs. communism wars. The religion vs. atheism wars. The Kirk vs. Picard wars.

I loved arguing online. I know I said last week that no one ever wins, but I never cared. I mastered Fisking before I mastered driving. It was fun. But as the internet grew, it became less fun. And it became less fun in part because the quality of people who could argue with went down. You were no longer guaranteed to be arguing with someone five times smarter or more knowledgeable than you, a true challenge forcing you to up your game. You were now arguing with people who used “lol” as a rhetorical device. The discourse became broader and, as Signor Eco described, worse.

Last year, I heard an interesting anecdote from someone I respect on another forum. They told a story about arguing with someone on Reddit and then deciding to click on that person’s profile to see who they were. This led them to quickly realize they – a full grown adult – was arguing with a teenager. Which leads to the question: would I argue with a teenager in real life? The answer was no, so they stopped.

This played a role in the last internet argument I was ever in. It was also on Reddit. I remembered this tale and clicked on their profile. My inspection led me to believe that, well, I guess the most subtle way to put it is that I doubt this person was typing with both hands. I immediately deleted my account.

That said, even despite all these reasons not to argue with someone, there was always one truth: I was arguing with a human being.

There’s something called the Dead Internet Theory. Wikipedia describes it as follows:

The dead Internet theory is an online conspiracy theory that asserts that the Internet now consists mainly of bot activity and automatically generated content manipulated by algorithmic curation, marginalizing organic human activity, to manipulate the population.

Let’s leave aside the alleged motives behind it and focus on the general idea which is of an internet increasingly consisting of automatically generated content. I’m not opposed to artificially generated content. Every heading picture I use is AI generated because I’m among the world’s worst artists. Look, here’s proof of my skill. I made this last year as an adult:

Do you understand now why I use AI generated images?

But discussion is a whole different beast. I try to use ChatGPT text here as little as possible, but this response to my question if it’s possible that internet discussion will increasingly be non-human was shocking to me:

Yes, it's not only likely but already happening that discussions on the internet, including in forums, comments sections, and on social media, will increasingly include non-human participation due to advancements in Artificial Intelligence (AI).

This thing constantly soft peddles everything I ask and yet here it’s saying “open your eyes you dumb human.” The future of online discussion isn’t arguing with someone who is dumb, or a teenager, or a serial self-pleasurer. It’s arguing with something that’s not even human. I know it’s easy to say “what’s the point?” now but in the future the answer is certainly “none.”

There’s the possibility that large platforms like Meta or X will choose – and be able to – ban AI generated content. But no matter how much the EU fulminates, the fact is that this is just going to be very difficult to do. X can’t even deal with the kittykat in bio bots that reply to every post on that platform. As AI increasingly demolishes the Turing Test, how are we going to stop it? Social media will not be able to survive in its current form if it – as looks certain – becomes increasingly dominated by non-humans.

Yet, the desire for internet discourse is not going to disappear. It started – long before Zuckerberg or Dorsey – for obvious reasons. Using the internet to talk with other humans is one of its first and core functions. And for all the negative aspects of that it’s had a lot of positive aspects. Turns out, humans like talking to each other, particularly about shared interests and goals. It will not go away.

But it will need to change, just to maintain functionality. The first and most obvious way is the one I keep hammering on about: the scale needs to change. The “everyone who speaks this language can talk to everyone else” style has so many problems and will not survive botification. Instead we’ll need smaller scale, barriers to entry, and curation.

I genuinely think Substack is a great step forward in social media. This series started with an article from Freddie de Boer’s Substack to which I once subscribed. This was in part because there was a time that he had an excellent comments section. And it was for three main reasons.

One, it was small. It self-selected for people who were interested in his content. Yes, that did involve some culture war stuff, but also much broader discussions on near countless topics, particularly literature. I took part in his book club and I hate fiction! Second, there was a barrier to entry in that only paid subscribers could comment. Third, he ruthlessly curated his comments section. This combined to create a genuinely enjoyable discourse. You wanted to get to his articles early so you could talk to the other commenters, because you knew them. “Wow, great point by Klaus. Oh wow Carina, I didn’t even think about that. Ugh, there’s Erin again.” And there were literally dozens of people who had interesting points to make, particularly of the type I prefer: ones I disagree with. There’s a reason that almost all the Substacks I recommend were commenters there. It spurred me back into writing because I wanted to have intelligent conversations with those people.

Yet, this was not Substack specific. Go into the comments section for any Substack with large numbers of subscribers – particularly if the comments are open and unmoderated – and it’ll be no different from a YouTube comments section. One or two intelligent points drowning in a sea of inanity.

You need all three of those elements. Curation doesn’t work at scale and to save time I’ll just gesture at Reddit. But it can work with small groups. And that means a barrier to entry is needed. Which, short of having it be invite only² is best handled by a fee. This seems inherently ridiculous – who would pay money to comment on the internet?! – but it’s not. The history of the internet has not been that people won’t pay for content, it’s just valued differently. There’s a reason Napster slaughtered the record industry. There’s a reason Spotify dramatically decreased online music piracy.

I’ve previously stated that there’s only one Substack I still pay for, The Tomkins Times, and if you are a Liverpool fan and you’re not subscribing to it, I don’t even know what you’re doing. I pay for the content but even if all the content was free and I just had to pay to comment, I still would. I love talking about this thing that interests me. And my alternatives are… bleak. Discussing sports on any large social media platform? No. I throw around nickels like manhole covers, but $5 to have discussions about something that interests me with smart and knowledgeable people? I don’t flinch. Smaller, curated social media spheres - organized around specific reasons - with a small fee to participate fixes most of the problems we have with social media. You wouldn’t need any of my commandments from last week if this was how social media existed. It is also closer to both how the early internet existed and - most importantly - to how the real world existed.

This is the future of social media. When I started talking about future social media platforms a year and a half ago, it was because I wanted something that was better. I felt that we’re still in the infancy of this technology and we’d see it improve. Now, I write this because I believe it is necessary and inevitable. AI’s influence in the world is going to be felt first and most in its native space: the digital one. How long are you going to participate in social media if you don’t even know if you’re talking to a human or a glorified computer program? My hunch is not long.  Change is coming, whether we recognize it yet or not.

1. The video that forever changed my opinion of YouTube was precisely this. And I can’t find it!

2. Since the early days of the internet I’ve been on invite only forums, comments sections, and mailing lists. Is there anyone who wants to dispute that these are inherently superior? I recognize that they’re difficult but they should be thought of as a great solution.