I’ve been working on this idea for the past few months. It’s a completely automated news publishing service, that reads the news from around the internet from many different sources, in order to understand what they all seem to agree on, and what seems to be bias, then writes a completely original news post based on what its learned, but without the bias. Everything from the titles, text, images, and summary are all written by AI and everything runs on auto-pilot.
The point is to create a reliable alternative medium for getting the news, in a unique and more fun/entertaining way that is more appealing to people today. It also includes a short easy to read summary for each news story, for people who don’t care to read the full story and quickly want to get the gist of it.
There’s also a automated daily & weekly newsletter, and everything is 100% free.
Please let me know what you think!
Feedback, suggestions, critiques are welcome : )
MEGA SUPER NEWS
2 Likes
nice website. How you generate images?
Although I can tell most of the content is written by AI due to its repetitive nature, it is pretty good, did not see any fabricated information or incorrect facts. AI is known to hallucinate which can harm quality if not checked…
Interesting. I know the likes of Bloomberg and other big publishers use AI for news but most stories are edited by humans before being published.
3 Likes
Do you have any idea about images on website? Which ai could have been used?
1 Like
There are so many AI image generators now, it could be any.
Image Creator by Microsoft (powered by Dall-e) (free)
Midjourney
DreamStudio
DaVinci
List goes on.
1 Like
custom fine-tuned model for everything, not open source.
1 Like
Please stop pretending that anything “written” by AI is “news.” Please stop presenting this to people as though the contents might be true or otherwise potentially worth reading. This is a vicious, evil thing you are doing, and you know it. Nobody wants to read made-up nonsense that was made up by a research project. The only people who read news-type articles are people who want to be able to trust the thing they’re reading, and the only people who will read this are the people who are stupid enough to believe it.
Stop making the world worse.
3 Likes
Is there any truth to this? If so, it’s very dystopian, and I’d like to know which publications so I can focus on the ones worth reading instead.
1 Like
Agreed. AI is trained on text from a plethora of sources and often those sources do contain biases. It’s impossible to claim it’s not biased.
2 Likes
I realize some of you have a cursory understanding of how large language models work, and that they do indeed have a tendency to hallucinate, therefore they must be unreliable at the task of creating unbiased news. This is understandable. However, I am not using AI in the way you might assume/imagine. If what you’re saying is true, do me a little favor, find a single instance on the site where it’s saying something that isn’t true. What’s happening under the hood is far more complex than asking chatGPT to generate some content about something… I spent months figuring out how to do this in such a way that the content actually IS what it claims to be. Please, point to any evidence to prove me wrong.
2 Likes
I have written many patents on different machine learning techniques. I assure you, my understanding of machine learning is not “cursory.” I’ve also used large language models, and obviously, they are incredibly unreliable. It’s hilarious that you think they can be anywhere near reliable, even when specialized to this extent. It’s especially hilarious that you think your model is somehow immune to bias–gee, what kind of training set and method could you possibly have used to negate its own biases?
No, I will not waste a second of my time fact-checking your pathetic little experiment for you. I don’t care if you spent years on this, I don’t care if you have a team of engineers with Ph.D.s in various machine learning applications, you did not invent and train a magical model that is immune to hallucination and to bias. I would sooner believe that you cured baldness and want to sell me the Brooklyn bridge for $25. I pity anybody who falls for your scam.
Take your site down, now.
You suggest my website has inaccurate information, yet you refuse to provide any evidence of that claim. You say it’s because you don’t want to “waste your time”, yet you waste much more of your time attacking random people on the internet, barking orders as if you have any authority whatsoever. Your argument is irrational, as demonstrated by your own words and actions. You’ve inadvertently confirmed that you actually have no basis for your argument against what I made, so I guess you’re bored/lonely and looking for some kind of attention or something? Anyways, you know, the nice thing about the internet, and freedom, is that if you don’t like something, you can simply not use it. Do you also whine and complain when grocery stores refuse to stop selling the foods you hate because in your view they’re unhealthy for people? Just because you don’t like something doesn’t mean you get to bark orders at people for it to go away. What a clown.
You suggest your website is magic, and expect us to believe you and take you seriously. I do not spend time inspecting your claims any more than I spend the time inspecting the claims of the last Nigerian prince that emailed me, because I’m not an idiot.
I’m attacking you for attacking me and the rest of our society. You do not have the right to use us as guinea pigs. Nobody wants to be experimented on.
I’m looking for you to be banned from our platform and for you to learn, through shunning, that dishonestly pretending that your website is a news website rather than an experiment and gimmick, is a bad thing to do.
If you prefer, you might change your banner to indicate that your website is just an experiment and is not to be taken seriously. But it sounds like you’re committed to your scam.
I will not use your website, and I will not have your website poison the society I live in. News is written by people, and fact-checked by people.
I would complain if a grocery store sold me spaghetti made of yellow play-dough and tried to convince me it was real spaghetti.
Everybody here knows who the clown is.
Nowhere did I suggest my website is magic, but nice straw-man. You remind me of the people who accused doctors of being witches because they couldn’t comprehend what they were doing… You don’t understand how something works, therefore it must be a scam… Nice logic, Mr. very smart man. I’m done engaging with you, please stop spamming my thread.
You suggested that your website is magically unbiased and magically immune to hallucination, like nothing it says is false. That’s not something any LLM has ever been close to capable of, there’s no research out there suggesting that level of accuracy is close to possible or that bias can be eliminated at all, you’re just making shit up to make it sound like you’re not a villain preying on society.
Stop trying to defend your scam, nobody here is going to fall for it, you already told us everything we need to know. Just say it in your header, too: tell them what your website really is, and watch how the few readers it has disappear.
You know it’s pretty ironic that the guy claiming to be some internet superhero trying to get stuff he doesn’t like cancelled, is blatantly making up stuff that I “suggested” that are verifiably false to anyone who can read the thread. Again, never said or suggested anything about magic. you don’t get it, and rather than ask “so, how does it work” like someone who’s actually open-minded and curious would ask, you instead insist that you are a ML domain expert, and are going to sit here and argue with someone who has nearly 20 years of experience and sold multiple software companies. You don’t even know who you’re talking to, yet you have the nerve to assume you know so much that you can say what I’m doing is a scam? I even gave you the benefit of the doubt, and said ok, so show me where it’s hallucinating, and you can’t…. Seriously dude, you’re embarrassing yourself at this point. I asked you to stop spamming the thread, and this forum won’t let me block you, but maybe you should get a hobby or something. Again, please stop spamming the thread.
Ok, this took an unexpected turn. I’m all for debate; that’s what forums are for. We all have different views and opinions, which is what makes discourse fascinating. There’s no need to resort to personal attacks.
For one, I think AI is a godsend when it comes to automating routine, monotonous tasks.
I like your project and its transparency. The website clearly tells the reader upfront that the content is AI-generated. If the reader wants to continue using your website that is on them. I would implement something like this myself if I had the brains for it. It would be tremendously helpful, especially in the sector I cover. There’s so much going on and so much is missed and this is where AI could help bridge the gap. However, I would include human input and wouldn’t let it publish without human supervision.
As mentioned previously, Bloomberg, MarketWatch, and Investing.com all use AI to generate news. In fact, Investing.com has gone so far as to add a disclaimer at the bottom of their AI-generated pieces, warning readers that the content is AI-generated. I’ve also seen reports of CNET doing the same.
1 Like
