I have an opinion than we need to pay some microtransactions to post our comments (web3?). Than all chatgpt-generated text will become a regular text from humans.
What prevents spammers from also paying? Right now many times of spam is done by humans so if it can be made profitable with manual work today I cannot see why having to pay would change anything.
The theory is it becomes exponentially more difficult to post spam in a cost-effective manner. As a nice added benefit, the spammers are also now directly funding your site, essentially paying you for the hassle of having to deal with them.
The first thing is not true for many sorts of spam. For example a lot of the people who destroy google search results do it manually and even pay for domain names tp do so. It is more difficult but I would say only slightly more so. And worst of all it is likely that spsmmers are willing to pay a higher price than your genuine users. Spam can be highly profitable.
That they make you a profit is true though, but that is both good and bad. It finds your antispam work but could also create perverse incentives where you want spsmmers.
Great. Can't happen fast enough. We've been coping with human-generated garbage for a long time, now AI will generate garbage at such a scale that the problems we've been coping with will be solved, because otherwise we'll all drown.
This quote stuck out to me especially
"There are a bunch of words, but no real value in what’s written,"
That has been my main problem with so many top search results, business insider being the most recognizable culprits IMO, for years.
Drowning in plastic waste, drowning in atmospheric pollution, drowning in mental health disorders. None of these have resulted in a magical "we'll solve it because otherwise we'll drown" solution. I fail to see why drowning in bad search results will result in a different outcome.
Carbon capture technology has moved out of R&D and into the productionization phase. There are multiple ongoing efforts to remove plastic from the ocean and other waterways which are currently actually removing plastic, not just working on R&D. There's also tons of work on how to breakdown plastic, so we aren't stuck with it forever. Mental health is honestly a harder problem because there's less that can be directly qualified, but I'd point to the recent successes with drugs like MDMA, Ketamine, and mushrooms as a very promising avenue. Currently they seem most well suited for those who are worst off, but that is likely just the start.
The knowing you're drowning part is super important. We haven't been visibly drowning in atmospheric pollution until say the past decade or so, even though the levels were absolutely astonishingly high. Similar with mental health, in the US we are now really drowning in mass shootings, which wasn't the case even 10 years ago.
Search results similarly have been like kind of shitty for a long time, the past 5 years at least, but they were and are still navigable. The drowning hasn't started in search yet either.
Plenty would argue it's been dead for awhile, since even before AI. [1] I refuse to concede though, it's as dead as you let it be. In both the places you passively consume from, and actively participate in. [2]
The old web was killed by the corporate web at least ten years so. The corporate web has created this absolute dumpster fire dystopia of brainless bumblegum listicles, outrage bait attention economy, and frightening total surveillance secretly destroying people's rights worldwide.
I was skeptical of this new wave of AI and still am but the open source side has been absolutely stunning. llama.cpp, whisper.cpp. HuggingFace and all the insane models. The community documenting and converting and explaining models.
Let this wave of LLM generated trash choke the gills of content marketing once and for all.
If I want mindless content, I'll ask my laptop now, and I won't ask your poor, pitiful click-farmer intern.
The corporate web might be good for buying things from companies, but it is an absolute cancer for knowledge and culture.
There's exponentially more knowledge and culture on the corporate web than there ever was on the old web.
I know Hacker News is almost universally at the "angry man yelling at clouds" stage when it comes to the web, but there is actually a baby to be lost with the bathwater.
I couldn't agree more about the baby. I've been through many of the stages of grief about what we've lost and what we're losing by letting the web be controlled by... whatever the hell this is.
I've tried for years to get people to understand that this web has become diseased. That it harms them in ways they may not understand as a less technical person. No one cares. I'm at a stage now where I stopped caring as well.
People will only understand when it affects them personally. It's sad. It hurts me to know that people will get hurt. But the current incarnation of the web needs to implode under the weight of its own garbage so it will finally be so obviously diseased and disgusting that no one will be able to not see it.
I say let it happen. I don't know how else to convince people to look away from the box.
Is my new project [0] also part of the problem? I'm still unsure myself because LLMs also allow us to process data in unprecedented ways. In my specific case, I auto generate stories to highlight different view points based on what people are saying about hot controversial topics on social media.
Because I intentionally process them with a very rudimentary "cartoonizer" in order to distinguish from a regular news articles and to emphasize that these stories are not written by humans. I don't know yet whether this helps.
Yes. In a few ways it’s considerably worse. Your website is referencing major world events, including war and freedom of press, by leveraging uninformed comments from the web (many of them themselves written by AI). That would be a problem even if your content weren’t auto-generated (random comments don’t make good or accurate journalism) but it’s worse when it’s churned at a high rate and introduces its own false interpretations.
Thanks for your input. This website should neither compete with nor replace regular journalism. What I try to achieve here is to be able to break free from social media silos where usually people are in kind of bubble. No one can read this many comments and people usually tend to read only comments / conversations where they align with their believes. Hence, I try to highlight different view points along with contrasting opinions (across several different social media platforms) to get an overview--not necessarily fact-based. These stories aren't supposed to push any agenda down anyone's throat.
Since this project just went live I'm still figuring out how to communicate that.
> This website should neither compete with nor replace regular journalism.
That’s good in theory, but people will use it like that anyway. We’re living in an era of rampant misinformation. Provable truths we’ve know for thousands of years are being called into question (the shape of the Earth). Even people who should know to do their own research (see recent case of lawyers using ChatGPT) are taking AI output as unquestionable truth. Your website won’t eliminate bias, it will only give people more sources to cherry pick.
> Hence, I try to highlight different view points along with contrasting opinions
Which exacerbates the issue. If we can prove the Earth is a sphere and 99% of people understand that, making it seem like the “contrasting opinion” is equally valid and deserves the same weighted consideration makes the problem worse, not better. John Oliver illustrated the problem in 2014, using climate change as the example.
>The implication that the Reddit blackouts are related to recent AI advances is ridiculous.
The allegation is that most of ChatGPT can be traced back to their using sites like Reddit and Twitter to scrape the information from for its training data. Reddit is changing the cost of accessing its APIs as a direct result of this and not really about killing off third party clients. That's the connection.
30 comments
[ 3.4 ms ] story [ 71.9 ms ] threadThat they make you a profit is true though, but that is both good and bad. It finds your antispam work but could also create perverse incentives where you want spsmmers.
Next all LLM's will need to find other training data and other approaches, as reddit, wikipedia and stackoverflow will all be dead.
This quote stuck out to me especially
"There are a bunch of words, but no real value in what’s written,"
That has been my main problem with so many top search results, business insider being the most recognizable culprits IMO, for years.
Carbon capture technology has moved out of R&D and into the productionization phase. There are multiple ongoing efforts to remove plastic from the ocean and other waterways which are currently actually removing plastic, not just working on R&D. There's also tons of work on how to breakdown plastic, so we aren't stuck with it forever. Mental health is honestly a harder problem because there's less that can be directly qualified, but I'd point to the recent successes with drugs like MDMA, Ketamine, and mushrooms as a very promising avenue. Currently they seem most well suited for those who are worst off, but that is likely just the start.
The knowing you're drowning part is super important. We haven't been visibly drowning in atmospheric pollution until say the past decade or so, even though the levels were absolutely astonishingly high. Similar with mental health, in the US we are now really drowning in mass shootings, which wasn't the case even 10 years ago.
Search results similarly have been like kind of shitty for a long time, the past 5 years at least, but they were and are still navigable. The drowning hasn't started in search yet either.
[1]: https://forum.agoraroad.com/index.php?threads/dead-internet-...
[2]: https://32bit.cafe/mission/
I was skeptical of this new wave of AI and still am but the open source side has been absolutely stunning. llama.cpp, whisper.cpp. HuggingFace and all the insane models. The community documenting and converting and explaining models.
Let this wave of LLM generated trash choke the gills of content marketing once and for all.
If I want mindless content, I'll ask my laptop now, and I won't ask your poor, pitiful click-farmer intern.
The corporate web might be good for buying things from companies, but it is an absolute cancer for knowledge and culture.
I know Hacker News is almost universally at the "angry man yelling at clouds" stage when it comes to the web, but there is actually a baby to be lost with the bathwater.
I've tried for years to get people to understand that this web has become diseased. That it harms them in ways they may not understand as a less technical person. No one cares. I'm at a stage now where I stopped caring as well.
People will only understand when it affects them personally. It's sad. It hurts me to know that people will get hurt. But the current incarnation of the web needs to implode under the weight of its own garbage so it will finally be so obviously diseased and disgusting that no one will be able to not see it.
I say let it happen. I don't know how else to convince people to look away from the box.
What's your opinion?
[0] https://zeitgaist.social
Yes. In a few ways it’s considerably worse. Your website is referencing major world events, including war and freedom of press, by leveraging uninformed comments from the web (many of them themselves written by AI). That would be a problem even if your content weren’t auto-generated (random comments don’t make good or accurate journalism) but it’s worse when it’s churned at a high rate and introduces its own false interpretations.
Since this project just went live I'm still figuring out how to communicate that.
That’s good in theory, but people will use it like that anyway. We’re living in an era of rampant misinformation. Provable truths we’ve know for thousands of years are being called into question (the shape of the Earth). Even people who should know to do their own research (see recent case of lawyers using ChatGPT) are taking AI output as unquestionable truth. Your website won’t eliminate bias, it will only give people more sources to cherry pick.
> Hence, I try to highlight different view points along with contrasting opinions
Which exacerbates the issue. If we can prove the Earth is a sphere and 99% of people understand that, making it seem like the “contrasting opinion” is equally valid and deserves the same weighted consideration makes the problem worse, not better. John Oliver illustrated the problem in 2014, using climate change as the example.
https://www.youtube.com/watch?v=cjuGCJJUGsg
There needs to be a dial between 100% stories about Prigozhin/wagner and 100% stories about Elon Musk's narcissistic press headline grab of the week.
> Redditors are staging blackouts.
The implication that the Reddit blackouts are related to recent AI advances is ridiculous. The embedded link also makes no mention of AI.
I dislike this kind of disingenuous clickbait. It's not worth my time to keep reading.
In a way I should thank them for the early red flag. This way I don't need to waste my time on the whole article.
The allegation is that most of ChatGPT can be traced back to their using sites like Reddit and Twitter to scrape the information from for its training data. Reddit is changing the cost of accessing its APIs as a direct result of this and not really about killing off third party clients. That's the connection.