Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

We're losing our voice to LLMs

tonyalicea.dev

Post reply on HN
Print view View original

121–130 of 425 posts

Re: We're losing our voice to LLMs

#121

Post by drbojingle »

ricardo81 wrote: post #6

I deleted my Facebook account a couple of years ago and my Twitter one yesterday. It's not just LLMs, it's how the algorithms promote engagement. i.e. rage bait, videos with obvious inaccuracies etc. Who gets rewarded, the content creators and the platform. Engaging with it just seems to accentuate the problem. There needs to be algorithms that promote cohorts and individuals preferences. Just because I said to someo…

No, there needs to be control over the algorithms that get used. You ought to be able to tune it. There needs to be a Google fuu equivalent for social media. Or, instead of one platform one algorithm, let users define the algorithm to a certain degree, using llms to help with that and then you can allow others to access your algorithms too. Asking for someone Facebook to tweak the algorithm is not going to help imo.

Re: We're losing our voice to LLMs

#122

Post by MichaelZuo »

vanviegen wrote: post #87
chemotaxis wrote: post #48

Earlier quoted context omitted.

> It's not just LLMs, it's how the algorithms promote engagement. i.e. rage bait, videos with obvious inaccuracies etc. I guess, but I'm on quite a few "algorithm-free" forums where the same thing happens. I think it's just human nature. The reason it's under control on HN is rigorous moderation; when the moderators are asleep, you often see dubious political stuff bubble up. And in the comments, there's often a fair…

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.

Yeah this is a critical difference, most of the issues are sidestepped because everyone knows nobody can force a custom frontpage tailored for a specific reader.

So there’s no reason to try a lot of the tricks and schemes that scoundrels might have elsewhere, even if those same scoundrels also have HN accounts.

Re: We're losing our voice to LLMs

#123

Post by anonymouskimmer »

AstroBen wrote: post #67

There's something unique about art and writing where we just don't want to see computers do it As soon as I know something is written by AI I tune out. I don't care how good it is - I'm not interested if a person didn't go through the process of writing it

A person can be just as wrong as an LLM, but unless they're being purposefully misleading, or sleep-writing, you know they reviewed what they wrote for their best guess at accuracy.

Re: We're losing our voice to LLMs

#124

Post by randycupertino »

AstroBen wrote: post #67

There's something unique about art and writing where we just don't want to see computers do it As soon as I know something is written by AI I tune out. I don't care how good it is - I'm not interested if a person didn't go through the process of writing it

I had a weird LLM use instance happen at work this week, we were in a big important protocol review meeting with 35 remote people and someone asks how long IUDs begin to take effect in patients. I put it in ChatGPT for my own reference and read the answer in my head but didn't say anything (I'm ops, I just row the boat and let the docs steer the ship). Anyone this bigwig Oxford/Johns Hopkins cardiologist who we pay $600k a year pipes up in the meeting and her answer is VERBATIM reading off the ChatGPT language word for word. All she did was ask it the answer and repeat what it said! Anyway it kinda made me sad that all this big fancy doctor is doing is spitting out lazy default ChatGPT answers to guide our research :( Also everyone else in the meeting was so impressed with her, "wow Dr. so and so thank you so much for this helpful update!" etc. :-/

Re: We're losing our voice to LLMs

#125

Post by criddell »

rdtsc wrote: post #70
ricardo81 wrote: post #6

I deleted my Facebook account a couple of years ago and my Twitter one yesterday. It's not just LLMs, it's how the algorithms promote engagement. i.e. rage bait, videos with obvious inaccuracies etc. Who gets rewarded, the content creators and the platform. Engaging with it just seems to accentuate the problem. There needs to be algorithms that promote cohorts and individuals preferences. Just because I said to someo…

> I deleted my Facebook account a couple of years ago and my Twitter one yesterday. I never signed up for Facebook or Twitter. My joke is I am waiting until they become good. They are still shitty and toxic from what I can tell from the outside, so I'll wait a little longer ;-)

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.

Re: We're losing our voice to LLMs

#127

Post by LogicFailsMe »

Propelloni wrote: post #92
LogicFailsMe wrote: post #69

Earlier quoted context omitted.

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.

It would just become part of the shitshow, cf. Grok.

Which is why you should cancel your Twitter account unless you're on the same page with the guy who owns it, but I digress.

if a site wants to cancel any ideology's viewpoint, that site is the one paying the bills and they should have the right to do it. You as a customer have a right to not use that site. The problem is that most of the business currently is a couple of social media sites and the great Mastodon diaspora never really happened.

Edit: why do some people think it is their god-given right that should be enforced with government regulation to push their viewpoints into my feed? If I want to hear what you guys have your knickers in a bunch about today, I will seek it out, this is the classic difference between push and pull and push is rarely a good idea.

My social media feeds had been reduced to about 30% political crap, 20% things I wanted to hear about, and about 50% ads for something I had either bought in the deep dark past or had once Google searched plus occasionally extremely messed up temu ads. That is why I left.

Re: We're losing our voice to LLMs

#128

Post by mentalgear »

IMTDb wrote: post #71
Lapel2742 wrote: post #23

Earlier quoted context omitted.

> it's how the algorithms promote engagement. They are destroying our democratic societies and should be heavily regulated. The same will become true for AI.

> 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 sh…

It's really not that complicated:

- Ban algorithmic optimization that feeds on and proliferates polarisation.

- To heal society: Implement discussion (commenting) features that allow (atomic) structured discussions to build bridges across cohorts and help find consensus (vs 1000s of comments screaming the same none-sense).

- Force the SM Companies to make their analytics truly transparent and open to the public and researchers for verification.

All of this could be done tomorrow, no new tech required. But it would lose the SM platforms billions of dollars.

Why? Because billions of people posting emotionally and commenting with rage, yelling at each other, repeating the same superficial arguments/comments/content over and over without ever finding common ground - traps a multitude more users in the engagement loop of the SM companies than people have civilised discussions, finding common ground, and moving on with a topic.

One system of social media that would unlock a great consensus-based society for the many, the other one endless dystopic screaming battles but riches for a few while spiralling the world further into a global theatre of cultural and actual (civil) war thanks to the Zuckerbergs & Thiels.

Re: We're losing our voice to LLMs

#129

Post by WhyOhWhyQ »

tucnak wrote: post #119
AstroBen wrote: post #67

There's something unique about art and writing where we just don't want to see computers do it As soon as I know something is written by AI I tune out. I don't care how good it is - I'm not interested if a person didn't go through the process of writing it

> There's something unique about art and writing where we just don't want to see computers do it Speak for yourself. Some of the most fascinating poetry I have seen was produced by GPT-3. That is to say, there was a short time period when it was genuinely thought-provoking, and it has since passed. In the age of "alignment," what you get with commerical offerings is dog shite... But this is more a statement on Americ…

"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.

Re: We're losing our voice to LLMs

#130

Post by outside1234 »

We have a channel at work where we share our experiences in using AI for software engineering.

Predictably, this has turned into a horror zone of AI written slop that all sounds the same, with section titles with “clever” checkbox icons, and giant paragraphs that I will never read.

Post reply on HN