Hacker News new | past | comments | ask | show | jobs | submit login
Show HN: NotYetNews – AI-Generated News from the Future (notyet.news)
53 points by johnpolacek 9 months ago | hide | past | favorite | 44 comments



Check out my recent full-GPT-4 generated HN parody as well:

https://crackernews.github.io

(featuring coherent comment threads!)

P.S. the prompts and the code I used can be found here: https://github.com/crackernews/crackernews.github.io/blob/gh...


This is great! Pretty similar to what I did in that you have actual links to the articles and even images. If you had a cron script that auto-generated new content every day, I think that would be the only thing that I'm doing that crackernews doesn't yet.

Another interesting choice is you are generating the article content itself. It is wild how you can get it to generate markup on the fly for different styles of posts.

For my project, I'm generating structured JSON with an [OpenAI Function](https://platform.openai.com/docs/api-reference/chat/create#c...) as I found even if you tell it the format you want in the prompt, it sometimes ignores it or doesn't match exactly. Then I'm uploading the JSON and images to S3.

I didn't realize you could get a DALLE image and drop it in your own directory with a oneliner like you have. Very nice!


> If you had a cron script that auto-generated new content every day

Thank you for the feedback! I thought about that initially, but turned out that generated jokes become a bit repetitive after a while and the project actually requires manual curation (for which I don't have the time). Or, it needs some source of "external entropy" which I haven't figured yet.

What is also interesting is the exact ways in which LLM humor becomes repetitive. I believe it could be a subject of study on its own, as it reveals something interesting about how human humor is structured.

P.S. For example, for some reason LLM really love all kinds of jokes about pirates, pizza, cheese and lost socks.


Yeah I found the same regarding repetition. I avoided that by seeding it with new data from NYTimes API each day.


Found this in the thread about GPT-10...

kernelKommando 4 minutes ago | parent | next [–]

What if we're all just GPT-generated comments in a GPT-generated world, and this is our existence now? reply

existentialDev 4 minutes ago | parent | next [–]

Well, that's my existential crisis for the day. Thanks a lot, kernelKommando. reply

Now this is "me"...What if we are all just existences spawned by GPT-1000 for the fleeting amusement of some beings...


And yeah... it is interesting that GPT-generated agents sometimes "realize" that they could be indeed GPT-generated.

I once saw a strange dream where my "character" realized that he was only a product of my imagination — the very moment before I woke up. It was not me, but him, who had that stunning realization. And once it happened, there was no him.

So I sometimes wonder if that is how the death looks like — you just suddenly realize that "you" was only a subprocess temporarily spawned to existence within some kind of powerful virtual machine, be it a GPT-1000 or whatever.


It easily could be. And according to the simulation hypothesis (by Nick Bostrom), it is the most probable thing actually.

P.S. There is also a hilarious Rick and Morty episode about a car battery universe: https://rickandmorty.fandom.com/wiki/Microverse_Battery


There's also Rachel's Whacker News: https://rachelbythebay.com/fun/hrand/

And Filippo Valsorda's gem:

https://filippo.io/fakenews/

And of course my one (shameless plug):

https://coxomb.github.io/This-Hacker-News-Does-Not-Exist/

I think everyone should have their own fake HN side project lol


I feel like stripping the emojis out would make it a bit more realistic


A good idea, did it.


The site is great, by the way


I built `This Hackernews Does Not Exist` which is similar to this:

https://coxomb.github.io/This-Hacker-News-Does-Not-Exist/

HN comments:

https://news.ycombinator.com/item?id=33845946


The post that worries me most about the future is "Ask HN: What Happened to Pizza?"


Cool! Old school no AI just good ol fashioned JS :)


Nice to see a fun, creative project like this. The LLM tie-in makes sense because it keeps headlines fresh & relevant to today's news. Thanks for sharing!

If anyone else wants a peek behind the curtain, here is the GPT-4 call: https://github.com/johnpolacek/notyetnews/blob/main/cron/ope...


Dare I say "The Onion" but in the future? That's what came to my mind when I saw it the site!


I think the year 2122 will be closer to Dune than any of these more optimistic articles. Or maybe Fahrenheit 451 if we don’t totally screw up the environment, with a dose of Harrison Bergeron.


In 2122 most western countries will look like slums in Brazil and South Africa, mixed with disproportionate wealth in a few places. Add a sprinkle of minority report, since (state) mass surveillance tech will be cheap and ubiquitous. People will still eat, fuck and shit. Coca cola and major brands will still exist in some way or another. The environment will remain the same.


very funny. But "NEW YORK - The year is 2122 and the fight for clean energy has taken an unexpected, yet entertaining turn." Shouldn't it read as a current news article writen in 2122? I wouldn't say "The year is 2023 and" today.


To be fair, this type of intro has been used plenty of times as a stylistic choice

https://archive.is/qwURS


For a more present-oriented, yet archaic in its delivery, news coverage I recommend the Medieval Content Farm: https://tidings.potato.horse


It seems like something is wrong with the links, at least on my machine. If I click the summary of the top article, I get linked to the second article. Clicking on the second article leads me to the third article, and so on.


2122 and still plagued by off-by-one errors!

Yes, same for me on mobile, took a while for me to figure out what was going on because the one I was getting was so similar to the one I was trying to click.


Oops! Fixed now


This one is actually quite hopeful! https://www.notyet.news/news/2023-08-13/1


> Why a Basic Algorithm Garnered Unprecedented Support from Quantum-Net Billionaires

> In an absurd twist of events, the quantum-net community is perplexed by the curious case of AI-RH1002 – an unremarkable algorithm sowing division in a universally united world. Like many divisive codes before it, AI-RH1002 seems to have won the backing of quantum-net billionaires.

Very funny site idea! This snippet reminds me of Ian M. Banks novels about the Culture.


[flagged]


If you're not interested in random and creative acts of programming, I think you're on the wrong discussion site.

This is for fun. It takes todays headlines and rewrites them. Like I could tell that the "The Mars State Fair Sees Many Galactic Party Candidates but Only One Zorgon" was a rewrite of "The Iowa State Fair Sees Many Republican Party Candidates but Only One Trump".

It retells the present in a future context. I enjoyed it.


>> Who cares?

I care, and I appreciated this submission. I dont think the submission is meant to be a replacement for Literature or Science Fiction, I think it is meant to be a Proof of Concept and showcasing a technology hack.


Thanks - yeah just having some fun over here!


I agree. It seems like an obviously tongue in cheek submission rather than any kind of profit focused product. It even has a link to the open sourced version at the bottom.


A proof of what concept?


I wanted to try out building a website that auto-generates its own content. So I did that https://github.com/johnpolacek/notyetnews


Honestly, I struggle with this. I've made some 'AI' generated content, Multi-player text based rpgs with persistent generated content, stories, news articles, etc. I never released any of it- not because it isn't entertaining content, but because it feels shallow. It's almost like the idea of uncanny valley, but I'd describe it more as a hollow facsimile. I want to like it. I keep trying to like it, but I am not there yet.


The silver lining is that flooding the internet with AI spam is self-defeating since it will taint the training data for future models.

Instead of learning to immitate humans they'll begin to imitation the imitations, then imitate the imitations of the imitations.


Everything in archives pre-2021 is still untainted. All major social media, Q&A, code repos, and archive.org are timestamped. It taints future collection of training data, but not existing collection of training data.


What's the plan then, to coast on pre-2021 data forever? How much utility would todays LLMs have if they were trained on fossilized archives of the internet from 10, 15, 20 years ago?


Not like we haven't filled it already with misinformation and blog farm drivel. We need to judge both human and AI content by its utility.


I agree. The more attention we give AI, the more it will be profitable for sites to use it to gain advertising revenue, and the more money will be given to it's development. That is why I have adopted the policy to NEVER click on any video or link that celebrates the use of AI, or uses AI to get more clicks. Of course, websites or blogs that harshly criticize AI are good.


This is inevitable. There are people who scroll aimlessly 2-3-4 hours per day. They will continue.

There are people that are looking/checking out the source. I don't think that AI will be so prevalent (its owners) so as to fale things so well. I remember quoting to a friend how Lex Fridman uses "love" as a unit of measurement while Harari uses "suffering". I was trying to imagine how a dialogue between the two would play out, and BOOM it happened (and I thoroughly e joyed their dialogue - twice).

Now, should Spofity purchase their voice signatures and IP and "have a-la-Holodeck" generate weekly discussions between the two, JUST so they keep me renewing, it will make me think that "it cannot be that these two guys talk every week" and thus I will pull the plug on it.

Looping back to the "aimlessly" part, I CARE to check that the thing I consume is legit/authentic. Someone else though will consume the "new AI generated Sepultura album" or "see the new clothing collection worn/displayed for you by YOUR favorite celeb (e.g.) Taylor Swift/James Hetfield/Max Cavalera/et al". To each his own I say, and let them scroll till the end of time.


I would generally agree with you, except that "to each his own" generally is leading to a very unhealthy society and the destruction of the ecosystem as well through endless economic growth.


> The more attention we give AI, the more it will be profitable for sites to use it to gain advertising revenue, and the more money will be given to it's development.

Pretty obvious slippery slope fallacy.


Your conscientious abstaining doesn't fix anything in the big picture


Well maybe not, but I am slowly growing a group of people that are interested in avoiding technology, and maybe soon we can form a coalition, one which would promise to avoid using AI. Not sure it will work but maybe it could. And there is already some backlash against AI so it's not as hard as you think!


Cool




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: