The stupidity of AI generated content.

[Anonymous]

Did not know to post it under entertainment or tech forum. Anyhow this is a biography of Fergie where the singer and Royal have been deemed to be the same person. The result is confusing to say the least.

https://coopwb.in/info/fergie-children/#:~:text=Fergie%2C%20the%20multi%2Dfaceted%20artist,Princess%20Beatrice%20and%20Princess%20Eugenie.

[Anonymous]

Anonymous wrote:Did not know to post it under entertainment or tech forum. Anyhow this is a biography of Fergie where the singer and Royal have been deemed to be the same person. The result is confusing to say the least.

https://coopwb.in/info/fergie-children/#:~:text=Fergie%2C%20the%20multi%2Dfaceted%20artist,Princess%20Beatrice%20and%20Princess%20Eugenie.

I admit it, I clicked, and what a hilarious clusterf*** of word salad that "article" is. But oh, it's not AI! There's a person named as the "author" so surely that means a real person wrote it! Right?...Right?....

[Anonymous]

I have played around with a few different AI platforms and there is no way that the AI that is out now is taking anyone's job.

Is it making some jobs easier? Yes, but you still need someone to proof and fact-check the output.

I work for a small company where every employee wears multiple hats. AI has been fantastic in helping with our social media and marketing campaigns. However, I quickly noticed that AI talks in circles. It will sometimes reiterate the same point twice or even three times in the same paragraph, but change how it is said ever so slightly. Anyone who uses nonedited AI-generated content is a fool.

I plugged in my CEO's name - someone who is very well-known in our small, niche space - and two different AI engines output bios that were filled with inaccuracies. One stated the CEO was a decorated military vet. The CEO has never served in the military. How could AI have gotten that information? Because the CEO presented donations for two different military fundraisers and AI mashed all the information from those articles together into one sloppy fake bio.

[Anonymous]

I mean, it's not actually "artificial intelligence" in the slightest.

[Anonymous]

AI, by definition, improves on its own, with time and exposure to more data. What you're seeing now is not what you're going to see in 5, 10, 25 years. Indeed, as any kid will tell you, right now you can direct AI to edit college essay drafts and other work to help you with polishing, or adding knowledge you lack. ChatGPT has written essays that can absolutely pass for decent undergrad work. It depends on how specifically you word your query.

During my kid's first college writing class, the prof requested work from ChatGPT with the student's own analysis of the operation.

Make no mistake, AI is coming for a ton of jobs. It is inevitable. This might not affect you, but it will certainly affect your children.

[Anonymous]

Anonymous wrote:

AI, by definition, improves on its own, with time and exposure to more data. What you're seeing now is not what you're going to see in 5, 10, 25 years. Indeed, as any kid will tell you, right now you can direct AI to edit college essay drafts and other work to help you with polishing, or adding knowledge you lack. ChatGPT has written essays that can absolutely pass for decent undergrad work. It depends on how specifically you word your query.

During my kid's first college writing class, the prof requested work from ChatGPT with the student's own analysis of the operation.

Make no mistake, AI is coming for a ton of jobs. It is inevitable. This might not affect you, but it will certainly affect your children.

I read another article about how even the essays that are written sound too "perfect". There's no nuance, and it's obvious that AI generated it. I've seen some postings on this forum that clearly indicate that an AI generated it what was written.

IMO, I think AI has a lot lot more to go before we need to really worry about it taking over. Even so, someone will need to review the AI content before it's published.

[Anonymous]

As the lawyers who filed an AI-generated brief discovered, AI is not a search engine. It doesn't know or care about facts or accuracy. It just makes stuff up.

[Anonymous]

I am a university prof and the students are using it to produce their essays. One student even submitted a paper using one of my journal articles without citing it, not knowing that I was one of the co-authors.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

AI, by definition, improves on its own, with time and exposure to more data. What you're seeing now is not what you're going to see in 5, 10, 25 years. Indeed, as any kid will tell you, right now you can direct AI to edit college essay drafts and other work to help you with polishing, or adding knowledge you lack. ChatGPT has written essays that can absolutely pass for decent undergrad work. It depends on how specifically you word your query.

During my kid's first college writing class, the prof requested work from ChatGPT with the student's own analysis of the operation.

Make no mistake, AI is coming for a ton of jobs. It is inevitable. This might not affect you, but it will certainly affect your children.

I read another article about how even the essays that are written sound too "perfect". There's no nuance, and it's obvious that AI generated it. I've seen some postings on this forum that clearly indicate that an AI generated it what was written.

IMO, I think AI has a lot lot more to go before we need to really worry about it taking over. Even so, someone will need to review the AI content before it's published.

PP you replied to. No no no. Sigh. It's because some people don't know how to use it. If you really want to use it well today, you can. But that implies a solid amount of intelligence and specialized background information from the human regarding the task!

Please remember that the only examples you're citing are the ones you *recognize* to be bad AI. Can you spot the good ones?
Observational bias, my friend.

BTW, I entirely agree that people have to be aware of that impersonal touch ChatGPT generates. They still have to put in the work to express their own "voice". AI-generated writing pieces only work with careful human management right now. But soon, that will change.

[Anonymous]

Anonymous wrote:I am a university prof and the students are using it to produce their essays. One student even submitted a paper using one of my journal articles without citing it, not knowing that I was one of the co-authors.

did you give them a 0?

[Anonymous]

Anonymous wrote:I am a university prof and the students are using it to produce their essays. One student even submitted a paper using one of my journal articles without citing it, not knowing that I was one of the co-authors.

I don't think you understand that many that even without AI a student will forget to site a paper and they rarely look at the author's name, they just type it in and have no clue who they are.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

AI, by definition, improves on its own, with time and exposure to more data. What you're seeing now is not what you're going to see in 5, 10, 25 years. Indeed, as any kid will tell you, right now you can direct AI to edit college essay drafts and other work to help you with polishing, or adding knowledge you lack. ChatGPT has written essays that can absolutely pass for decent undergrad work. It depends on how specifically you word your query.

During my kid's first college writing class, the prof requested work from ChatGPT with the student's own analysis of the operation.

Make no mistake, AI is coming for a ton of jobs. It is inevitable. This might not affect you, but it will certainly affect your children.

I read another article about how even the essays that are written sound too "perfect". There's no nuance, and it's obvious that AI generated it. I've seen some postings on this forum that clearly indicate that an AI generated it what was written.

IMO, I think AI has a lot lot more to go before we need to really worry about it taking over. Even so, someone will need to review the AI content before it's published.

PP you replied to. No no no. Sigh. It's because some people don't know how to use it. If you really want to use it well today, you can. But that implies a solid amount of intelligence and specialized background information from the human regarding the task!

Please remember that the only examples you're citing are the ones you *recognize* to be bad AI. Can you spot the good ones?
Observational bias, my friend.

BTW, I entirely agree that people have to be aware of that impersonal touch ChatGPT generates. They still have to put in the work to express their own "voice". AI-generated writing pieces only work with careful human management right now. But soon, that will change.

In the next 20 years? Probably not. For one thing, all AI right now is built upon massive copyright fraud and is all illegal. Soon, AI will either be scrapped or completely rebuilt in a much more expensive, cumbersome, and smaller format trained only on legal data and information.

[Anonymous]

Anonymous wrote:I am a university prof and the students are using it to produce their essays. One student even submitted a paper using one of my journal articles without citing it, not knowing that I was one of the co-authors.

Have you submitted any of your papers to AI detectors to see the score given?

There was a big hullabaloo at DS' college last semester. Professors started using AI detection software that was outputting nearly all papers submitted were failing. The professors were not believing students who said they had not used any AI during their writing until students proved the inaccuracy of the software by having professors submit their own papers.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

AI, by definition, improves on its own, with time and exposure to more data. What you're seeing now is not what you're going to see in 5, 10, 25 years. Indeed, as any kid will tell you, right now you can direct AI to edit college essay drafts and other work to help you with polishing, or adding knowledge you lack. ChatGPT has written essays that can absolutely pass for decent undergrad work. It depends on how specifically you word your query.

During my kid's first college writing class, the prof requested work from ChatGPT with the student's own analysis of the operation.

Make no mistake, AI is coming for a ton of jobs. It is inevitable. This might not affect you, but it will certainly affect your children.

I read another article about how even the essays that are written sound too "perfect". There's no nuance, and it's obvious that AI generated it. I've seen some postings on this forum that clearly indicate that an AI generated it what was written.

IMO, I think AI has a lot lot more to go before we need to really worry about it taking over. Even so, someone will need to review the AI content before it's published.

PP you replied to. No no no. Sigh. It's because some people don't know how to use it. If you really want to use it well today, you can. But that implies a solid amount of intelligence and specialized background information from the human regarding the task!

Please remember that the only examples you're citing are the ones you *recognize* to be bad AI. Can you spot the good ones?
Observational bias, my friend.

BTW, I entirely agree that people have to be aware of that impersonal touch ChatGPT generates. They still have to put in the work to express their own "voice". AI-generated writing pieces only work with careful human management right now. But soon, that will change.

In the next 20 years? Probably not. For one thing, all AI right now is built upon massive copyright fraud and is all illegal. Soon, AI will either be scrapped or completely rebuilt in a much more expensive, cumbersome, and smaller format trained only on legal data and information.

This. One novelist has already filed a suit to stop her work from being "scraped" to train AI; she had found writing that clearly, identifiably had lifted portions of her work. Other novelists are quickly becoming vocal about scraping to train AI.

We non-novelist folks need to get loud now, too, because companies can--without our permission or knowledge--scrape things online and use it to train AI, and that includes our own social media posts etc.

Don't believe it's happening? Think that only authors and big names are affected? Nope.
https://www.theverge.com/2023/7/5/23784257/google-ai-bard-privacy-policy-train-web-scraping
And Google's not the only one doing it.

[Anonymous]

There is an AI scraping this forum right now, and soon it will be giving questionable advice to strangers.