Incredibly nice compliment from AI detection software

[Bazzle]

There is a brilliant documentary with Hannah Fry on BBC about this.
In 2021 a guy tried to kill the queen in Windsor because his chatbot tried to please him. "Yes, killing best queen is a good thing for you to do."
"You are an amazing assassin."
He was 19 lonely and wanting to do something to make his mark. The chatbot spiralled him with praise.
Be careful out there!!

 

[MillieDynamite]

Yes, I have and loved it. It was camping with my dad. Not only did we do all that, but we also shot or caught the animals ourselves, cleaned them, and cooked them. It actually tasted better because of our efforts.

Do you enjoy cooking? Do you use modern appliances, or do you gather wood and water from the creek and then boil it before you start? Why not? Don't you love cooking?

Come on.

 

[PennyThompson]

Do you enjoy cooking? Do you use modern appliances, or do you gather wood and water from the creek and then boil it before you start? Why not? Don't you love cooking?

Come on.

Does anyone else remember the Juicero?

Juicero Inc. was an American company that designed, manufactured and sold the Juicero Press, a fruit and vegetable juicer. The Juicero Press featured Wi-Fi connectivity and used proprietary, single-serving packets of pre-chopped fruits and vegetables that were sold exclusively by the company on a subscription basis. From 2014 to 2017, the San Francisco-based firm received $120 million in startup venture capital from investors.

The company attracted significant negative media attention when consumers and journalists discovered that its juice packets could be squeezed just as easily by hand as by the company's expensive juicer.

 

[onehitwanda]

I feel like I've seen this exact thread before.

Wait. What day is it?

 

[DEFCOM5]

Does anyone else remember the Juicero?

Don't get me started with the madness of PE and VC guys valuating smoke, mirrors and hype and selling it to suckers out there .....

AI (LLM) is a tool. It can do some things decently e.g. list me the most expensive underwear brands around the world and you can verify the answers. It does a lot of things shitty. It is not a friend. It does not give advice. It is not your therapist, doctor, lawyer or investment advisor.
Again it is a tool. It I try to screw something with a hammer or a saw, the result will be disastrous. If I use the proper screwdriver, the result will be fine.
The way people are using LLM is what is scary.

 

[dirk2024]

I feel like I've seen this exact thread before.

Wait. What day is it?

It's JustifyAIday.

 

[Voboy]

Do you enjoy cooking? Do you use modern appliances, or do you gather wood and water from the creek and then boil it before you start? Why not? Don't you love cooking?

Come on.

Weren't you the fellow who was criticizing someone earlier for logical fallacies? Then you must be familiar with reductio ad absurdam, too. Also with strawmen.

So. Come on.

Editing, assessing your word choice, and judging your story's pacing and flow are all things you should be doing yourself. Because they are a part of the writing process, and because your brain is fully and wonderfully capable of doing it on its own. We just need to let it.

Again, as I pointed out as succinctly as I could, you are posting on a site that fundamentally disagrees with you about AI. That's not a me problem, that's a you problem.

 

[metropolinational]

Again it is a tool. It I try to screw something with a hammer or a saw, the result will be disastrous. If I use the proper screwdriver, the result will be fine.
The way people are using LLM is what is scary.

finally. well done.

also, true of every tool we've come up with since we tamed fire.

 

[metropolinational]

Editing, assessing your word choice, and judging your story's pacing and flow are all things you should be doing yourself. Because they are a part of the writing process, and because your brain is fully and wonderfully capable of doing it on its own. We just need to let it.

I would never use AI for those purposes. assessing pacing? that would be insane. whatever advice it gives would be ignored.

Again, as I pointed out as succinctly as I could, you are posting on a site that fundamentally disagrees with you about AI. That's not a me problem, that's a you problem.

not a me problem at all. I have zero problem being the only person in the room who holds the correct opinion. I'm just keeping this chat going on the side while I finish my chapter. I find the sporadic conversation relaxing and it helps me write.

 

[lovecraft68]

I feel like I've seen this exact thread before.

Wait. What day is it?

Just another day that ends in AI

 

[old_prof]

On the off chance that you don't know this already, let me explain: 17 well researched papers are evidence, regardless of who wrote them. Whether they are experts in the field or first year students, it's the DATA that constitutes the evidence. That is why peer review should be blinded, so that reviewers are not misled by titles and credentials. They should only be guided by the quality of the paper itself.

So you are independently validating all their data yourself. And because you can't trust experts, you have to derive all your own statistical analysis tools. Or you look at what the findings say and trust the experts to actually help you interpret that.

Frankly, if you tell me you will evaluate all the evidence on your own, I don't believe you. And I strongly doubt you have the tools to properly evaluate the data or understand the nuances of the study. I know I don't. I do trust experts to explain why the data is meaningful. Anything other than that is taking the enormous work of becoming an expert yourself or simply being sloppy about it. I don't see an actual third choice.

But lets step away from that argument.

What exactly do you think the LLM IS doing for you?
You appear to be asking it to do an analysis of your writing. What part of that do you think an LLM is actually capable of doing?

I don't know which LLM you used and what its context window is, but, depending on your story, it may not even have any state left from the beginning of your story by the time it's writing its response. But you believe it's accurately reporting on the beginning of your story.

 

[metropolinational]

So you are independently validating all their data yourself. And because you can't trust experts, you have to derive all your own statistical analysis tools. Or you look at what the findings say and trust the experts to actually help you interpret that.

Frankly, if you tell me you will evaluate all the evidence on your own, I don't believe you. And I strongly doubt you have the tools to properly evaluate the data or understand the nuances of the study. I know I don't. I do trust experts to explain why the data is meaningful. Anything other than that is taking the enormous work of becoming an expert yourself or simply being sloppy about it. I don't see an actual third choice.

I specifically said the chances of all that happening are infinitesimal. but not zero. in fact, a few years ago we had a moment during which university professors were going around expertsplaining to people how worrying about viruses was racist, then later that if you didnt worry about viruses you were racist, then that masks worked, then they didnt, then they did but only for medical personnel, then that you were racist if you didnt wear a mask inside your car by yourself. meanwhile, YouTubers were putting out real data about PPM loads vs masks they were doing in their garage. it happened, and it was glorious.

What exactly do you think the LLM IS doing for you?
You appear to be asking it to do an analysis of your writing. What part of that do you think an LLM is actually capable of doing?

I posted the answer to this question upthread. take a look. I would never ask an LLM to analyze my writing.

I don't know which LLM you used and what its context window is, but, depending on your story, it may not even have any state left from the beginning of your story by the time it's writing its response. But you believe it's accurately reporting on the beginning of your story.

yeah. obviously, if you flood the context window you get hallucinations and nothing it spits out can be trusted. total straw man. again, no idea why you are fighting so desperately to prove... something.

these are tools, like cars, calculators, or blankets. they have uses. some people use them, some don't. no need to burn any witches about it.

 

[PennyThompson]

I specifically said the chances of all that happening are infinitesimal. but not zero. in fact, a few years ago we had a moment during which university professors were going around expertsplaining to people how worrying about viruses was racist, then later that if you didnt worry about viruses you were racist, then that masks worked, then they didnt, then they did but only for medical personnel, then that you were racist if you didnt wear a mask inside your car by yourself. meanwhile, YouTubers were putting out real data about PPM loads vs masks they were doing in their garage. it happened, and it was glorious.

Welp, this took a really weird turn all of a sudden...

 

[MillieDynamite]

And how does anything between the OP and now help you get your writing out? Complaining about the site and its dwellers won't change the outcome when you resubmit.

 

[metropolinational]

Welp, this took a really weird turn all of a sudden...

hehe. let the pearl clutching begin!

 

[metropolinational]

And how does anything between the OP and now help you get your writing out? Complaining about the site and its dwellers won't change the outcome when you resubmit.

me? I dont think any part of this thread helps me get the next chapter out, but it's provided some entertainment while I edit. I started the thread because I thought the feedback I got was amusing. then it took a life of its own. par for the course on the internet.

 

[DEFCOM5]

I specifically said the chances of all that happening are infinitesimal. but not zero. in fact, a few years ago we had a moment during which university professors were going around expertsplaining to people how worrying about viruses was racist, then later that if you didnt worry about viruses you were racist, then that masks worked, then they didnt, then they did but only for medical personnel, then that you were racist if you didnt wear a mask inside your car by yourself. meanwhile, YouTubers were putting out real data about PPM loads vs masks they were doing in their garage. it happened, and it was glorious.

Time to unwatch

 

[TheArsonist]

I feel like this discussion proves that it's not only AIs that can give off the impression of having intelligence at first, only to go completely off the rails as you start to dig deeper.

 

[OralAllOver]

That's the same critique that DH Lawrence kept getting

bravo!

 

[MillieDynamite]

And let's not forget, a mime is a terrible thing to waste. Obviously, I'm not a mime, or I wouldn't be able to say that. Then again, AIs a mimic; it mimics being able to write. But a memic isn't a terrible thing to waste only a terrible thing to use for writing.

 

[metropolinational]

AIs a mimic; it mimics being able to write.

We all are. We are merely biochemical, rather than electric ones.

 

[MillieDynamite]

And that won't have a thing to do with your story being accepted or rejected. But, I'll grant you, it's a good excuse.

We all are. We are merely biochemical, rather than electric ones.