I don’t really play games at all and the only system I own is a Nintendo Switch. I played the new Zelda and am really enjoying it. When I get stuck or have a question I google it. When I google it, I note that the top say 10 sites that answer my question all have similarly-formatted articles and also have the same mistakes in the answer to the question.
I expect nearly all the sites pointing to Zelda tips or FAQs are not original in any way and are instead just regurgitated from one another through AI to form a webring of shit.
At least I use ad blockers and block JavaScript so they don’t get some ad revenue.
Yep, game wikis and guides have been a cesspool of SEO bullshit for a little while now. I guess there's money in the ads.
I find it hard to believe Google doesn't have the data required to figure out which sites should be getting domain authority, but I understand it's an antagonistic and ongoing battle between SEO spam and search engines.
I owned a basic wordpress site with some guides written during COVID for a couple of popular (at that time) mobile games. It appeared in top 3 searches for the <game name> + <guide/event/best ...> combination.
Ad revenue was around ~$100 per month peaking around ~$150 at the new content drops. I have abandoned the website since, but it still is generating $100 here and there without me actively working on it for the past 2 years.
To the point - yes, there's lot of money if you own a network of similar websites. My competitors were paying $10-30 per guide submitted on their website because a few guides could easily get paid off in a couple of months from ad revenue and keep generating it years afer.
I could write more detailed story if someone's interested.
Recipes sites too. Recently I was looking at recipes for a dish I wanted to try and noticed that two blogs had identical recipes for "Grandma's X". I scrolled back up to the obligatory five page essay and those were almost identical too--one of them had clearly stolen the story from the other and run it through one of those AI "paraphrase" tools to change words here and there. It was pretty easy to tell which was the original and which the copycat based on some strange word choices in the paraphrased version.
>… understand it's an antagonistic and ongoing battle between SEO spam and search engines
On the other hand, those Stack Overflow clones get top billing all the time. I do not think this is a data problem, but an incentives issue. More ads=good
But they do know. They are incentiviswd to send you to the sites with the most ads, not the site which is most relevant, because they make a ton money off those spam sites and less of dedicated fan sites.
At some point I'm gonna say fuck it and simply maintain my own list of searchable links, which score each ad with a single negative point, and each relevant keyword or phrase with a single positive point.
Yeah I don’t get it - without search dominance google’s entire business model is highly vulnerable. Why have they allowed search quality to stagnate while pouring billions into random tech areas where they have little competitive advantage?
Nobody wants to see a page of copies/duplicates by SEO rank. I recall years ago there was, for a while, a similar issue with Wikipedia clones but that problem was addressed. Not just games, recipes, but music lyrics, chords, etc - googling almost evert niche interest is rife with it.
Oh Google certainly could moderate their search results or augment them with human knowledge. They'd need a dozen people per country, at most, for such a task. The problem is, Google is already threading a very fine line regarding anti-competition and bias especially in Europe... and say, they would choose "Computer BILD" (a German tabloid) over Heise or Golem (actually respectable media), or the other way around - the outcry would be massive and so would the coming lawsuits.
Instead, it's better for Google to let the search go to utter dog shit, blame issues on "algorithms" and get sites to pay to play with ads.
The top domain for almost every result in my niche is a shitty wordpress with centered text on a purple background. It makes about $10-15k a month even with its lackluster content. It's had spyware ads and porn ads. Still #1 for every query.
The funny part is, bing, ddg are even worse. Welcome to webapps in 2023.
> Yep, game wikis and guides have been a cesspool of SEO bullshit for a little while now. I guess there's money in the ads.
I absolutely loathe to use fandom for that very reason. They game the system to appear first, even when another, better, wiki exists.And it is nigh on unusable. Sadly one game that really depends on the wiki, that I play only has a fandom site
(warframe)
That is the main reason why I fall back to good old book guides. They are actually quite fun and make me nostalgic, because in my childhood we used them too. Piggyback is a publisher that makes great ones, they publish only one per year and put a lot of effort and work in their guides. They made the offical Tears of the Kingdom guide too, can recommend.
It does get deeper than just your mentioned surface level of served ads.
Many of those wiki ecosystems are used to view-pump twitch channels with embedded players, such as Fextralife's webring of wikis being bought up and built up, and don't forget those "1-feature VIP memberships".
What I find amusing is that even when ChatGPT responds to my gaming questions I find the response to be unusually wordy with a surprisingly heavy into paragraph before it answers my simple question.
Very similar to those few advertisement sites with a huge block of wonky into text taking forever to get to the answer.
Moderating on a subreddit where people are definitely trying to spam their way in with LLMs, that's becoming on of my big tells. At least the way they are tuned now they are really prone to wordy and pointless intros and outros. (At least in my personal opinion, while intros and outros are fine they should do something more than just summarise again, especially the outro, even if it's just trying to end on a joke.) I recognise there's no reason to believe that is fundamental to the tech, but right now it's a tell. (Though "subtle errors of the type humans don't make" is both a stronger signal and more relevant one.)
This is the default behavior from the way it is fine-tuned and from its system prompt. But you can always tell it to give you answers that are short and to the point.
And trying to remember what specific enemies are where is not fun for me. Is it fun for anyone? The Compendium doesn't keep track of them when you fight, despite the fact they respawn every blood moon.
I want another Savage Lynel Bow. Where did I see that Silver Lynel? I can't remember, but I'm not hunting it all across the map hoping for the vague ping of my Purah Pad sensor when I'm already on top of one. Should I be making my own notes as well?
The log is a great idea, but there's lots of stuff it doesn't capture. The quest descriptions often aren't sufficient and markers are often the person that started the quest. Great(!)
I appreciate the subtitles, but why is so much of the game unvoiced?
Not a concern for me as I'm playing on PC, but why no button mapping?
Even with internal FSR and AA disabled so it doesn't occasionally dip to smeared potato quality, it can be hard to spot things if you don't know explicitly where to look, even for hawkish eyesight. The Ultrahand glow can help, but it's a rubbish solution.
No colour blind settings. WTAF.
There are so many missed opportunities to make this amazing game vastly more accessible. Nintendo has made it clear they don't care about accessibility. It's a real shame, because accessibility makes things better for everyone.
> Not a concern for me as I'm playing on PC, but why no button mapping?
I'm not sure if this would be exposed in whatever emulator you're using, but the Switch does have console-wide button remapping in the system settings. This obviously isn't sufficient though, since at the very least it presents a huge hassle for having different input settings per game or if multiple people are sharing the same profile. Unfortunately, Nintendo either doesn't recognize the limitations of only having a single system-wide remapping or doesn't think it's worth prioritizing better solutions within their first-party games.
There are absolutely emulator mods that swap out the image assets in the game for any controller you would like, including the steam deck, steam controller, Xbox controller, and PlayStation controller. If you are feeling lazy, head to your preferred non-google search engine and type in “Tears of the Kingdom Mod Manager” and you will find helpful software that enables one-click installation of many mods.
I don’t understand how you could fault the game developer for not supporting this feature, since it is made by a company that sells specific hardware that they guarantee support for.
As far as color issues, I too share color vision deficiencies but everything I have encountered has been carefully designed with shapes in mind so that color is only an enhancement. I STRONGLY believe that this approach to design is much, much, much more accessible than any toggles and filters buried in menus. Could you share with me some examples of where color was key to solving a puzzle or identifying an enemy or item?
And there are SO MANY named locations, and many things refer to them only by name, but no way to search through ones you've found! I've spent quite a lot of time browsing through the map at maximum zoom to try to find the named area, only to discover hours later that I wasn't even in the right quadrant.
It's quite frustrating at times. I broadly have fun with it, but improving these things would make the game a lot more consistently enjoyable.
On Kagi I just block that shit since it's relatively obvious. Some trustworthy sites in this space are gamefaqs(!) and IGN (and game magazines and others I can't recall now).
Yea, its tragic. When I was a kid I would often just go to GameFAQs because that was the only real source, and the guides would be surfaced in google. Its still around but Google gives me 20 crappy ad riddled sites that require a new page click, listicle-style, to get through a 5 minute area
The other problem with static sites is that it's trivial for some jerkface to copy your content and repost it. The first case of that I heard about was almost 20 years ago when it happened to a friend. The social circle of the subject matter was small enough that she could name and shame to get it taken down. Today I don't know what you could do.
It happened to me about that long ago too and I was just flattered, but it also wasn't done by some crappy scammer ripping random content to stuff it full of ads and SEO it to death hoping make a quick buck off of my writing. I have no desire to stop people from copying and sharing whatever they can, but when it's being done with a profit motive it feels a lot more exploitative and dirty.
Undoubtedly it had many hits, but a lot of the greats have also since been remastered and can be enjoyed on newer platforms with HD graphics and modernized controls.
I noticed the same thing with Zelda. The marketability of game FAQs was realised up to a decade ago, but AI content generators, optimised by time-honoured SEO, has evidently accelerated the time to market for new platforms.
The success of TOTK, and the availability of articles and Reddit posts has made it ripe for the picking.
Haha, I made https://meat-gpt.sonnet.io (the site is my way of poking fun at crappy AI startups, I won't spoil how).
Now, one of the biggest sources of traffic for me (ca 70-80% at times!) are crappy AI app catalogues which not just include my MeatGPT but also hallucinate the most beautifully stupid descriptions of what it does.
I couldn't imagine this stuff working better to be fair: taking that juicy slab of meat from my site and then proceeding to repeatedly slap themselves in the face with it screaming "please give me more".
I mean, I love generative art, and built a self-publishing medieval content farm[1], but this thing just writes itself, across multiple sites.
> But again, there’s nothing to stop this. These subreddits can’t only fill themselves with joke articles to screw up a site like this, even if this one specific example is good for a laugh.
Most threads will be normal conversations, and reddit (and other discussion sites) can serve as a simple way of summarizing and generating news for 'free'.
I'm thinking that HN too could serve as a source for tech related news, couldn't it? Summarize the target article, then join it up with summaries/sentiments of the top comments in the thread. I didn't say I'm doing it, but if I could think of it, someone's probably way ahead of me already and has tried it.
I keep hearing this sentiment on HN and IRL. As a journalist I think it misses the mark somewhat by failing to account for the value of reporting.
While some news can be generated exclusively from scraping Reddit threads or whatever, most decent journalism incorporates some form of reporting, i.e. the generation of novel information from trusted sources. Even without reporting, if you can't add to the store of knowledge in the world by writing the article, it doesn't offer any value to consumers or advertisers. That includes the the world of SEO spam. An effort has to be made to distinguish your work from the competition, or else your site isn't winning those top results.
Reddit threads are often just full of emotional responses to news already generated in this way. At some point along the line, a human has gone out and spoken to another human, forming an novel angle or argument, pursuing a line of inquiry, connected dots no one else has yet etc. That's news, not a summary of existing attitudes.
>I keep hearing this sentiment on HN and IRL. As a journalist I think it misses the mark somewhat by failing to account for the value of reporting.
There is valor added by journalists in even niche sectors. A journalist that reports on cars knows about the industry itself and can give an informed take on different developments, he might know how a car works, he might know about different trends in design, or markets, or whatever else. That is his added value.
When it comes to videogame journalism, though, they act as little more than spokespeople for corporations. They generally don't understand the product or how it works (mechanically or in terms of design), and in some cases aren't even adept at playing videogames themselves. The only thing the world would lose if no game journalist ever mentioned WoW again and the devs communicated directly with the playerbase would be the appearance of impartiality journalists give.
> As a journalist I think it misses the mark somewhat by failing to account for the value of reporting.
At a revenue level how do you quantify that value? Unless you are a brand (TIME, Washington Post, etc) the only metrics are page views and time spent reading the article that eyeballs wander across ads.
Arguably a clear well written article may perform worse than an AI generated article that bleeds over on to a second page. Some of the most valuable content on the internet is not well written pieces on the political climate, but "OMG YOU'LL NEVER GUESS HOW THIS CHILD CELEB TURNED OUT" with a 75+ image one per page photo gallery.
As a common man someone who hates what journalism has become: I hope "AI" swiftly replaces your industry which has become a cancer and plague upon society. I would rather deal with a used car salesman than a journalist.
Journalists today only provide unnecessary exposition and emotional poetry (read: wasting my reading time) and actively cause more problems and conflicts in society than they address and resolve (sensationalism and fearmongering is what brings in the clicks and thus the money).
I have no sympathy whatsoever for such a rotten, morally devoid, worthless industry, and while I doubt "AI" will bring about any fundamental difference or improvement it will at least make the process of journalism better reflect the actual value of the final product.
- ask the AI to filter out the ones that are not tech-related (bay-area topics, politics)
- scrape selected articles
- write summaries
- publish static website
other sources are reddit subreddits and rss feeds (official languages' blogs and github's releases page).
The AI is quite gullible. That can be avoided by giving it more context and having a review step where you make sure that you are enforcing your editorial rules.
Another thing that I've been wondering is having a cheaper model (gpt-turbo-3.5) write articles and then use a more sophisticated to review them (gpt4)
I particularly appreciate that you have a disclaimer at the top stating that the article was AI generated. I wish more sites that used AI did that. Nice execution too.
Even if you don't want to just summarise the comments, you can view the comments section for most articles as crowd-sourced research on the topic. You can easily walk away from most interesting discussions here with a shortlist of topical articles, books, etc, and often some colour commentary straight from the horse's mouth.
Anon because I know startups working on that exist and should not talk about it
Information wants to be free; right now it’s a browser plugin that reads DOM and allows tagging; that human work is fed into a training system
The idea started as make Twitter community notes it’s own thing as browser plug-in and then expanded into model training to look like a user browsing the web and read DOM
Not hitting an API, so no fees yes? It’s the user making the interaction choice
Rate limits are going to be the only defense against AI trained to look like doom scrollers
They can't only do that but they can embarrass someone who's passing an AI writer off as human. At minimum, they'll have to message in some way that the article was written by AI lest this trick is pulled to embarrass an author which was otherwise considered to be human.
This is happening on a lot more gaming news sites than just this one.
I'm regularly seeing articles in my Google News feed that are summarizing Reddit threads with questionable accuracy in their coverage that are clearly AI generated.
All that said, when it does work correctly it's actually kind of useful and timely.
For about a year now since seeing the advent of decent generative text AI, I'm starting to dream of the day I no longer use my devices to go online and just get a phone call from an AI voice that tells me if I have anything urgent in my email or summarizes news I'd actually give a crap about.
The web has gone from something I truly loved to an abusive relationship so entrenched in my day to day I can't boot it from my life.
I for one welcome an AI intermediary that needs to suffer through reading SEO spam and click bait headlines and nonsense comments to synthesize a useful summary of what's important to me.
Exciting! This looks and feels a lot like the counter culture experience I had in the 80s as a hacker. The sense of fighting The Man may now be taken up against The AI. Well played game nerds!
I always like how this just brings to the surface the lack of respect the creators of the thing, in this case AI, that it will be abused by the public. The public doesn't care how/what the devs want it to do. The public cares about what they can make it do. See hot rods, overclocking, or any of the millions of other examples.
Yep, and even if the US regulates it, the public will take it underground, and overseas will keep pushing on. The cat is out of the bag.
I would still like to see these companies try and fix any negative externalities mind you. If they just throw their hands up at helping accelerate SEO spam then that'd be disappointing.
I feel like we've seen this any time that the tools are new and powerful enough - the folks unburdened by hierarchical structure can and do move fast, so when there's a sea change they can get in there and do some good damage until the behemoths catch on and steer their resources towards shoring up what they perceive as defects (frequently identified as anything that reduces their control over any system or it's connected parts).
My hope is that things get faster and more chaotic as we move forward. The best stuff happens when the empowered are scrappy and in it for the love and not the money.
Yeah. But on the other hand, this is Made For Advertising (MFA) content. The worst thing that can happen to it isn't notoriety, it's getting on the naughty list with Google or the programmatic ad markets. The former can gut its traffic, the latter can gut its ability to monetize.
Ironically, it's looking like the war on trash AI content is going to be fought by adtech firms who need to plausibly claim their customers aren't going to be wasting money on worthless inventory.
Whatever it is they do for money, people visiting in this context are unlikely to convert, taking it down also mitigates a little of the reputational damage at least.
The site is run by a competitive game tournament company (that I briefly worked at) and I think eyeballs on the article would have actually converted since it's different branding.
I was trying to find out about video stabilization options on Linux yesterday and stumbled over an article that extremely confidently was discussing a tool that’s maintained by the „authors of ffmpeg“ … which surprised me. So I googled it and couldn’t find it. Then I put the name into quotes and found three articles mentioning it, all by the same company.
So it’s actually starting to happen. LLM generated content filled with hallucinations are flooding the web. The internet will soon become completely useless as the signal to noise ratio plummets until it’s practically impossible to sort the wheat from the chaff.
That AI transformer paper really is Pandora’s Box.
They aren't as funny and elaborate as they once were, but yeah, they still do post something every year. In the early days of WoW they used to publish really good fake patch notes that were right on the line of credibility with a couple fake bombshells mixed in. The forums would melt down with posts from people who forgot it was April 1. Good times.
Readit News
I expect nearly all the sites pointing to Zelda tips or FAQs are not original in any way and are instead just regurgitated from one another through AI to form a webring of shit.
At least I use ad blockers and block JavaScript so they don’t get some ad revenue.
I find it hard to believe Google doesn't have the data required to figure out which sites should be getting domain authority, but I understand it's an antagonistic and ongoing battle between SEO spam and search engines.
I owned a basic wordpress site with some guides written during COVID for a couple of popular (at that time) mobile games. It appeared in top 3 searches for the <game name> + <guide/event/best ...> combination.
Ad revenue was around ~$100 per month peaking around ~$150 at the new content drops. I have abandoned the website since, but it still is generating $100 here and there without me actively working on it for the past 2 years.
To the point - yes, there's lot of money if you own a network of similar websites. My competitors were paying $10-30 per guide submitted on their website because a few guides could easily get paid off in a couple of months from ad revenue and keep generating it years afer.
I could write more detailed story if someone's interested.
On the other hand, those Stack Overflow clones get top billing all the time. I do not think this is a data problem, but an incentives issue. More ads=good
Now we have 45 minute youtube videos of "100 new player tips and tricks for Red Dead 2!"
At some point I'm gonna say fuck it and simply maintain my own list of searchable links, which score each ad with a single negative point, and each relevant keyword or phrase with a single positive point.
Nobody wants to see a page of copies/duplicates by SEO rank. I recall years ago there was, for a while, a similar issue with Wikipedia clones but that problem was addressed. Not just games, recipes, but music lyrics, chords, etc - googling almost evert niche interest is rife with it.
Instead, it's better for Google to let the search go to utter dog shit, blame issues on "algorithms" and get sites to pay to play with ads.
The top domain for almost every result in my niche is a shitty wordpress with centered text on a purple background. It makes about $10-15k a month even with its lackluster content. It's had spyware ads and porn ads. Still #1 for every query.
The funny part is, bing, ddg are even worse. Welcome to webapps in 2023.
I absolutely loathe to use fandom for that very reason. They game the system to appear first, even when another, better, wiki exists.And it is nigh on unusable. Sadly one game that really depends on the wiki, that I play only has a fandom site (warframe)
Many of those wiki ecosystems are used to view-pump twitch channels with embedded players, such as Fextralife's webring of wikis being bought up and built up, and don't forget those "1-feature VIP memberships".
Very similar to those few advertisement sites with a huge block of wonky into text taking forever to get to the answer.
ToTK straight up sucks for accessibility, cognitive and sensory stuff especially.
I find myself using the ZD TotK map more often than I'd like:
https://www.zeldadungeon.net/tears-of-the-kingdom-interactiv...
And trying to remember what specific enemies are where is not fun for me. Is it fun for anyone? The Compendium doesn't keep track of them when you fight, despite the fact they respawn every blood moon.
I want another Savage Lynel Bow. Where did I see that Silver Lynel? I can't remember, but I'm not hunting it all across the map hoping for the vague ping of my Purah Pad sensor when I'm already on top of one. Should I be making my own notes as well?
The log is a great idea, but there's lots of stuff it doesn't capture. The quest descriptions often aren't sufficient and markers are often the person that started the quest. Great(!)
I appreciate the subtitles, but why is so much of the game unvoiced?
Not a concern for me as I'm playing on PC, but why no button mapping?
Even with internal FSR and AA disabled so it doesn't occasionally dip to smeared potato quality, it can be hard to spot things if you don't know explicitly where to look, even for hawkish eyesight. The Ultrahand glow can help, but it's a rubbish solution.
No colour blind settings. WTAF.
There are so many missed opportunities to make this amazing game vastly more accessible. Nintendo has made it clear they don't care about accessibility. It's a real shame, because accessibility makes things better for everyone.
I'm not sure if this would be exposed in whatever emulator you're using, but the Switch does have console-wide button remapping in the system settings. This obviously isn't sufficient though, since at the very least it presents a huge hassle for having different input settings per game or if multiple people are sharing the same profile. Unfortunately, Nintendo either doesn't recognize the limitations of only having a single system-wide remapping or doesn't think it's worth prioritizing better solutions within their first-party games.
I don’t understand how you could fault the game developer for not supporting this feature, since it is made by a company that sells specific hardware that they guarantee support for.
As far as color issues, I too share color vision deficiencies but everything I have encountered has been carefully designed with shapes in mind so that color is only an enhancement. I STRONGLY believe that this approach to design is much, much, much more accessible than any toggles and filters buried in menus. Could you share with me some examples of where color was key to solving a puzzle or identifying an enemy or item?
> No colour blind settings. WTAF.
Do color blind settings through the OS work for gaming, or does the game have to be designed to accomodate to them in some manner?
It's quite frustrating at times. I broadly have fun with it, but improving these things would make the game a lot more consistently enjoyable.
When you say it’s still going strong, are you referring to homebrew dev or something?
The success of TOTK, and the availability of articles and Reddit posts has made it ripe for the picking.
Dead Comment
Now, one of the biggest sources of traffic for me (ca 70-80% at times!) are crappy AI app catalogues which not just include my MeatGPT but also hallucinate the most beautifully stupid descriptions of what it does.
I couldn't imagine this stuff working better to be fair: taking that juicy slab of meat from my site and then proceeding to repeatedly slap themselves in the face with it screaming "please give me more".
I mean, I love generative art, and built a self-publishing medieval content farm[1], but this thing just writes itself, across multiple sites.
[1] https://tidings.potato.horse
Dead Comment
> But again, there’s nothing to stop this. These subreddits can’t only fill themselves with joke articles to screw up a site like this, even if this one specific example is good for a laugh.
Most threads will be normal conversations, and reddit (and other discussion sites) can serve as a simple way of summarizing and generating news for 'free'.
I'm thinking that HN too could serve as a source for tech related news, couldn't it? Summarize the target article, then join it up with summaries/sentiments of the top comments in the thread. I didn't say I'm doing it, but if I could think of it, someone's probably way ahead of me already and has tried it.
While some news can be generated exclusively from scraping Reddit threads or whatever, most decent journalism incorporates some form of reporting, i.e. the generation of novel information from trusted sources. Even without reporting, if you can't add to the store of knowledge in the world by writing the article, it doesn't offer any value to consumers or advertisers. That includes the the world of SEO spam. An effort has to be made to distinguish your work from the competition, or else your site isn't winning those top results.
Reddit threads are often just full of emotional responses to news already generated in this way. At some point along the line, a human has gone out and spoken to another human, forming an novel angle or argument, pursuing a line of inquiry, connected dots no one else has yet etc. That's news, not a summary of existing attitudes.
There is valor added by journalists in even niche sectors. A journalist that reports on cars knows about the industry itself and can give an informed take on different developments, he might know how a car works, he might know about different trends in design, or markets, or whatever else. That is his added value.
When it comes to videogame journalism, though, they act as little more than spokespeople for corporations. They generally don't understand the product or how it works (mechanically or in terms of design), and in some cases aren't even adept at playing videogames themselves. The only thing the world would lose if no game journalist ever mentioned WoW again and the devs communicated directly with the playerbase would be the appearance of impartiality journalists give.
At a revenue level how do you quantify that value? Unless you are a brand (TIME, Washington Post, etc) the only metrics are page views and time spent reading the article that eyeballs wander across ads.
Arguably a clear well written article may perform worse than an AI generated article that bleeds over on to a second page. Some of the most valuable content on the internet is not well written pieces on the political climate, but "OMG YOU'LL NEVER GUESS HOW THIS CHILD CELEB TURNED OUT" with a 75+ image one per page photo gallery.
Journalists today only provide unnecessary exposition and emotional poetry (read: wasting my reading time) and actively cause more problems and conflicts in society than they address and resolve (sensationalism and fearmongering is what brings in the clicks and thus the money).
I have no sympathy whatsoever for such a rotten, morally devoid, worthless industry, and while I doubt "AI" will bring about any fundamental difference or improvement it will at least make the process of journalism better reflect the actual value of the final product.
- pick HN items with more than 400 votes
- gather their titles in a list
- ask the AI to filter out the ones that are not tech-related (bay-area topics, politics)
- scrape selected articles
- write summaries
- publish static website
other sources are reddit subreddits and rss feeds (official languages' blogs and github's releases page). The AI is quite gullible. That can be avoided by giving it more context and having a review step where you make sure that you are enforcing your editorial rules. Another thing that I've been wondering is having a cheaper model (gpt-turbo-3.5) write articles and then use a more sophisticated to review them (gpt4)
This is the prompt that generates the articles: https://github.com/lfarroco/news-radar/blob/main/src/writer....
I’d really like to see this - any chance we could?
Information wants to be free; right now it’s a browser plugin that reads DOM and allows tagging; that human work is fed into a training system
The idea started as make Twitter community notes it’s own thing as browser plug-in and then expanded into model training to look like a user browsing the web and read DOM
Not hitting an API, so no fees yes? It’s the user making the interaction choice
Rate limits are going to be the only defense against AI trained to look like doom scrollers
I'm regularly seeing articles in my Google News feed that are summarizing Reddit threads with questionable accuracy in their coverage that are clearly AI generated.
All that said, when it does work correctly it's actually kind of useful and timely.
For about a year now since seeing the advent of decent generative text AI, I'm starting to dream of the day I no longer use my devices to go online and just get a phone call from an AI voice that tells me if I have anything urgent in my email or summarizes news I'd actually give a crap about.
The web has gone from something I truly loved to an abusive relationship so entrenched in my day to day I can't boot it from my life.
I for one welcome an AI intermediary that needs to suffer through reading SEO spam and click bait headlines and nonsense comments to synthesize a useful summary of what's important to me.
I would still like to see these companies try and fix any negative externalities mind you. If they just throw their hands up at helping accelerate SEO spam then that'd be disappointing.
This is _hacker_ news. I don’t care about what it was “intended” to do either; I care about what it can do when pushed to its limits!
My hope is that things get faster and more chaotic as we move forward. The best stuff happens when the empowered are scrappy and in it for the love and not the money.
Ironically, it's looking like the war on trash AI content is going to be fought by adtech firms who need to plausibly claim their customers aren't going to be wasting money on worthless inventory.
Shit tier websites with "quality" ads (ie something I want to click on more) can be very valuable to advertisers.
This fight needs to be fought by search, not ad tech. Ad tech has too many perverse incentives.
That AI transformer paper really is Pandora’s Box.
Maybe I should ask ChatGPT to help me come up with more ideas…