We're Losing Our Voice to LLMs

On HN everybody sees the same ordering. Therefore you get to read opinions that are not specifically selected to make you feel just the perfect amount of outrage/self-righteousness.

Some of that you may experience as 'dubious political stuff' and 'patently incorrect takes'.

Edit, just to be clear: I'm not saying HN should be unmoderated.

I want to agree with this. Maybe OP is young or didn't frequent other communities before "social networks", but on IRC, even on Usenet you'd see these behaviors eventually.

Since they are relatively open, at some point comes in someone that doesn't give care about anything or it's extremely vocal about something and... there goes the nice forum.

I would be intrigued by using an LLM to detect content like this and hold it for moderation. The elevator pitch would be training an LLM to be the moderator because that's what people want to hear, but it's most likely going to end up a moderator's assistant.

When video games first started taking advantage of behavioral reward schedules (eg: skinner box stuff such as loot crates & random drops) I noticed it, and would discuss it among friends. We had a colloquial name for the joke and we called them "crack points." (ie, like the drug) For instance, the random drops that happen in a game like Diablo 2 are rewarding in very much the same way that a slot machine is rewarding. There's a variable ratio of reward, and the bit that's addicting is that you don't know whenever next "hit" will be so you just keep pulling the lever (in the case of a slot machine) or doing boss runs. (in the case of Diablo 2)

We were three friends: a psychology major, a recovering addict, and then a third friend with no background for how these sorts of behavioral addictions might work. Our third friend really didn't "get it" on a fundamental level. If any game had anything like a scoreboard, or a reward for input, he'd say "it's crack points!" We'd roll our eyes a bit, but it was clear that he didn't understand that certain reward schedules had a very large effect on behavior, and not everything with some sort of identifiable reward was actually capable of producing behavioral addiction.

I think of this a lot on HN. People on HN will identify some surface similarity, and then blithely comment "see, this is nothing new, you're either misguided or engaged in some moral panic." I'm not sure what the answer is, but if you cannot see how an algorithmic, permanently-scrolling feed differs from people being rude in the old forums, then I'm not sure what would paint the picture for you. They're very different, and just because they might share some core similarity does not actually mean they operate the same way or have the same effects.

I suspect it got worse with the advent of algorithm-driven social networks. When rage inducing content is prevalent, and when engaging with it is the norm, I don't see why this behaviour wouldn't eventually leak to algorithms-free platforms.

I personally dont feel like an ultra filtered social media which only shows me things I agree with is a good thing. Exposing yourself to things you dont agre with is what helps us all question our own beliefs and prejudeces, and grow as people. To me, only seeing things you know you are already interested in is no better than another company curating it for me.

So far my experience is that unless you subscribe to the general narrative of the platform, the discover algorithm punishes you with directing the mob your way.

I had two of my Bluesky posts on AI being attacked by all kinds of random people which in turn has also lead to some of those folks sending me emails and dragging some of my lobster and hackernews comments into online discourse. A not particularly enjoyable experience.

I’m sure one can have that same experience elsewhere, but really it’s Bluesky where I experienced this on a new level personally.

I enjoy Mastodon a lot. Ad-free, algo-free. I choose what goes in my feed, I do get exposed to external viewpoints by people boosts (aka re-tweets) and i follow hashtags (to get content from people I do not know). But it's extremely peaceful, spam and bots are rare and get flagged quickly. There's a good ecosystem of mobile apps. I can follow a few Bluesky people through a bridge between platforms and they can follow me too.

That's truly all I need.

Doesn’t Bluesky have a set of moderation rules that guarantee that it will turn into bizarro-world Twitter?

I tried Bluesky and wanted to like it. My account got flagged as spam, still no idea why. Ironically it could be another way of loosing ones voice to an LLM :)

> should be heavily regulated.

By who, exactly? It’s easy to call for regulation when you assume the regulator will conveniently share your worldview. Try the opposite: imagine the person in charge is someone whose opinions make your skin crawl. If you still think regulation beats the status quo, then the call for regulation is warranted, but be ready to face the consequences.

But if picturing that guy running the show feels like a disaster, then let’s be honest: the issue isn’t the absence of regulation, it’s the desire to force the world into your preferred shape. Calling it “regulation” is just a polite veneer over wanting control.

My view is that they are just exposing issues with the people in the said societies and now is harder to ignore them. Much of the hate and the fear and the envy that I see on social networks have other reasons, but people are having difficulties to address those.

With or without social networks this anger will go somewhere, don't think regulation alone can fix that. Let's hope it will be something transformative not in the world ending direction but in the constructive direction.

I agree, but focusing on "the algorithm" makes it seems to the outsider like it must be a complicated thing. Really it just comes down to whether we tolerate platforms that let somebody pay to have a louder voice than anyone else (i.e. ad supported ones). Without that, the incentive to abuse people's attention goes away.

We've seen what happens when we pretend the market will somehow regulate itself.

A social network can be great. Social media — usually not.

Something like Instagram where you have to meet with the other party in person to follow each other and a hard limit on the number of people you follow or follow you (say, 150 each) could be an interesting thing. It would be hard to monetize, but I could see it being a positive force.

Your loss.

Twitter was an incredible place from 2010 to 2017. You could randomly message something and they would more often than not respond. Eventually an opportunity would come and you’d meet in person. Or maybe you’d form an online community and work towards a common goal. Twitter was the best place on the internet during that time.

Facebook as well had a golden age. It was the place to organize events, parties, and meetups, before instagram and DMs took over. Nothing beats seeing someone post an album from last nights party and messaging your friends asking them if they remember anything that happened.

I know being cynical is trendy, but you genuinely missed out. Social dynamics have changed. Social media will never be as positive on an individual level as it was back then.

You can have a LinkedIn profile without reading the feed.

This is literally how most of the world uses LinkedIn

I never understand why people feel compelled to delete their entire account to avoid reading the feed. Why were you even visiting the site to see the feed if you didn’t want to see the feed?

LinkedIn bothers me the least, even though it definitely has some of the highest level of cringe content. It's still a good tool to interact with recruiters, look at companies and reach out to their employees. The trick is blocking the feed with a browser extension.

Sorting the feed by "recent" at least gives you a randomized assortment of self aggrandizement, instead of algorithmically enhanced ragebait

I have a special, deep, loathing for linkedin. I honestly can't believe how horrible it is and I don't understand why people engage with it.

This. Linkedin is garbage, yet I still use it because there are no competitors. This is what happens in a monoculture.

IMO there should not be an algorithm. You should just get what you have subscribed to, with whatever filters you have defined. There are better and worse algorithms but I think the meat of the rot is the expectation of an algorithm determining 90% of what you see.

Could someone use a third-party AI agent to re-curate their feeds? If it was running from the user's computer I think this would avoid any API legal issues, as otherwise ad and script blockers would have been declared illegal long ago.

> but the well-intentioned blocklists just became the chief tool for the mean girls of the site.

I've never used it, but yes this is what I expected. It would be better to have topical lists that users could manually choose to follow or block. This would avoid quite a bit of the "mean girl" selectivity. Though I suppose you'd get some weird search-engine-optimization like behavior from some of the list curators (even worse if anyone could add to the list).

> If you don’t want to hear a lot of Brexit talk, don’t engage with Brexit content.

That is really limiting though. I do not want to see Brexit ragebait in my threads, but I am quite happy to engage in intelligent argument about it. The problem is that if, for example, a friend posts something about Brexit I want to comment on, my feed then fills with ragebait.

My solution is to bookmark the friends and groups pages, and the one group I admin and go straight to those. I have never used the app.

I got out of Twitter for a few reasons; part of what made it unpleasant was that it didn't seem to be just what I did that adjusted my feed, but that it was also affected by what the other people I connected to did.

> You get more of what you engage with. If you don’t want to hear a lot of Brexit talk, don’t engage with Brexit content.

The algorithm doesn’t show you “more of the things you engage with”, and acting like it does makes people think what they’re seeing is a reflection of who they are, which is incorrect.

The designers of these algorithms are trying to figure out which “mainstream category” you are. And if you aren’t in one, it’s harder to advertise to you, so they want to sand down your rough edges until you fit into one.

You can spend years posting prolificly about open source software, Blender and VFX on Instagram, and the algorithm will toss you a couple of things, but it won’t really know what to do with you (aside from maybe selling you some stock video packages).

But you make one three word comment about Brexit and the algorithm goes “GOTCHA! YOU’RE ANTI-BREXIT! WE KNOW WHAT TO DO WITH THAT!” And now you’re opted into 3 bug ad categories and getting force-fed ragebait to keep you engaged, since you’re clearly a huge poltical junky. Now your feed is trash forever, unless you engage with content from another mainstream category (like Marvel movies or one of the recent TikTok memes).

I use X. I have an enormouse blocklist and I block keywords. I found that I can also block emoji. This keeps my feed focused on what I want to see (no politics. Just technology, classical and jazz music, etc.)

[dead]

Yes. That particular content-farm business model (rewrite 10 articles -> add SEO slop -> profit) is effectively dead now that the marginal cost is zero.

I’m not mourning it.

I’m not sure I agree with this sentiment. You can type something "by hand" and still have almost no voice in it if the incentives push you to flatten it out.

A lot of us spent years optimizing for clarity, SEO, professionalism etc. But that did shape how we wrote, maybe even more than our natural cadence. The result wasn’t voice, it was everyone converging on the safe and optimized template.

>her answer is VERBATIM reading off the ChatGPT language word for word

How could it be verbatim the same response you got? Even if you both typed the exact same prompt, you wouldn't get the exact same answer.[0, 1]

[0] https://kagi.com/assistant/8f4cb048-3688-40f0-88b3-931286f8a...

[1] https://kagi.com/assistant/4e16664b-43d6-4b84-a256-c038b1534...

The LLM may well have pulled the answer from a medical reference similar to that used by the dr. I have no idea why you think an expert in the field would use ChatGPT for a simple question, that would be negligence.

The one thing a cardiologist should be able to do better than a random person is verify the plausibility of a ChatGPT answer on reproductive medicine. So I guess/hope you're paying for that verification, not just the answer itself.

Or both the doctor and ChatGPT were quoting verbatim from a reputable source?

"inhabited by characters ACTUALLY living in the computer"

It's hard to imagine these feeling like characters from literature and not characters in the form of influencers / social media personalities. Characters in literature are in a highly constrained medium, and only have to do their story once. In a generated world the character needs to be constantly doing "story things". I think Jonathan Blow has an interesting talk on why video games are a bad medium for stories, which might be relevant.

> Personally, I'm looking forward to the age of computational literature, where authors like me would be empowered to engineer whole worlds, inhabited by characters ACTUALLY living in the computer.

So you want sapient, and possibly sentient, beings created solely for entertainment? Their lives constrained to said entertainment? And you'd want to create them inside of a box that is even more limited than the space we live in?

My idea of godhood is to first try to live up to a moral code that I'd be happy with if I was the creation and something else was the god.

If this isn't what you meant, then yes, choose your own adventure is fun. But we can do that now with shared worlds involving other humans as co-content creators.

I'm taking "same way" to be read as "authoritative", whether de facto or de jure. Basically by dint of people using what's provided instead of coming up with their own.

Everyone has the same few dictionary spellings (that are now programmed into our computers). Even worse (from a heterogeneity perspective), everyone also has the same few grammar books.

As examples: How often do you see American English users write "colour", or British English users write "color", much less colur or collor or somesuch?

Shakespeare famously spelled his own last name half a dozen or so different ways. My own patriline had an unusual variant spelling of the last name, that standardized to one of the more common variants in the 1800s.

https://en.wikipedia.org/wiki/History_of_English_grammars

"Bullokar's grammar was faithfully modelled on William Lily's Latin grammar, Rudimenta Grammatices (1534).[9] Lily's grammar was being used in schools in England at the time, having been "prescribed" for them in 1542 by Henry VIII.[5]"

It goes on to mention a variety of grammars that may have started out somewhat descriptive, but became more prescriptive over time.

Thanks so much!

Tbh I prefer to read/skim the comments first and only occasionally read the original articles if comments make me curious enough. For now I never ended checking something that would seem AI generated.

True, but one of the least-explored problems with AI is that because it can regurgitate basic writing, basic art, basic music with ease, there is this question:

Why do it at all if I won't do better than the AI?

The worst risk with AI is not that it replaces working artists, but that it dulls human creativity by killing the urge to start.

I am not sure who said it first, but every photographer has ten thousand bad photos in them and it's easier if they take them at the beginning. For photographers, the "bad" is not the technical inadequacy of those photos; you can get past that in the first one hundred. The "bad" is the generic, uninteresting, uninspiring, underexplored, duplicative nature of them. But you have to work through that to understand what "good" is. You can't easily skip these ten thousand photos, even if your analysis and critique skills are strong.

There's a lot to be lost if people either don't even start or get discouraged.

But for writing, most of the early stuff is going to read much like this sort of blog post (simply because most bloggers are stuck in the blogging equivalent of the ten thousand photos; the most popular bloggers are not those elevating writing).

"But it looks like AI" is the worst, most reflexive thing about this, because it always will, since AI is constantly stealing new things. You cannot get ahead of the tireless thief.

The damage generative AI will do to our humanity has only just started. People who carry on building these tools knowing what they are doing to our culture are beneath our contempt. Rampantly overcompensated, though, so they'll be fine.