To ask what you use ChatGPT for?

[OhWhatsTheBloodyPoint]

I may be late to the party, but recently discovered it and wonder what others use it for?

This week I’ve been using it to translate French menus while I’m on holiday. It’s been great.

Hope none of you have allergies that might be triggered by a mistranslation

@TheLudditesWereRight no allergies. I think if I had an allergy worth worrying about I’d probably learn the word for it when travelling!

Transcription, a hateful and tedious part of my job. It's sloppy but not so bad that I can't tidy it up, and referring back to the original audio is easy.

Other than that... nothing, and I'm frankly appalled at the things people choose to outsource - making life decisions, making aesthetic choices, writing emails, and don't even get me started on how fucked up it is to use it as therapy. I like thinking, I like using my own human critical faculties. What's the point of decorating your house if you're not choosing how to do it? What's the point in any of this if you're outsourcing anything that makes you an individual human with agency? If anyone in my life dares send me an AI-composed email I WILL notice and I WILL call them out on it.

The thought occurred to me throughout this thread that the people who are most excited about artificial intelligence doing things for them are probably those who don't have a great deal of human intelligence to do those things themselves.

Question for those using chatgpt to increase their productivity: is your salary increasing to reflect your improved performance?

People have had decorators in for decades. You’re being ridiculous.

All sorts of things.

I've used it to help me tone-check emails because I have a tendency to sound a bit curt without meaning to. I'm also a waffler so I've used it to help me trim things down to be more succinct. I don't get it to write my emails for me, just offer suggestions of how I can make them better, then I edit myself as needed.

I use it for recipes when I can't be arsed thinking of something to cook for dinner or need to substitute an ingredient. Best Madeira cake I've ever made was a ChatGPT offering! It's great if you just tell it what you've got in, what vibe you're going for and how much effort you want to put in, it'll come up with ideas.

I use it when I can't think of a word but I know the meaning of the word, or if I can't remember the name of a book I read when I was a child but remember the main character, stuff like that.

I play videogames and have asked it how to do X or Y because it's easier than sifting through endless Reddit threads. I mean it basically just does the sifting part for you.

It's just important to remember it's a tool and needs human input to work properly, sometimes it'll come out with absolute garbage that needs reworking or modifying.

I will say I have also massively reduced my use of it since finding out the environmental impact it has.

@JudgeBread

“I've used it to help me tone-check emails because I have a tendency to sound a bit curt without meaning to.”

This is how I use it for emails. I’m not curt but I have a quite formal writing style at work and sometimes it helps to make it a little more friendly.

I was thinking of using it for counselling but my friend who creates Apps and has worked very high up in the technical side of Samsung and sky gave us all a warning about not putting anything personal into ChatGPT unless you are prepared for the potential that you could be blackmailed with it. It's harvesting lots of information. It's quite easy to narrow down who individuals are particularly if they've say booked a flight, dicussed their career, used it to edit something that that ends up online, put photos in etc.

You might think you're okay with sharing your personal information but when they threaten to tell your boss (or future) that you were struggling because you fancied a colleague or some such thing that you probably don't want them to know. But do you really want your highly personal information to be sent to every email contact you've ever had or to everyone in your phone directory?
My friend says the amount of blackmailing of firms now is going through the roof and there is now big business in blackmailing individuals.

The people that have access to this information answer people you want to know your private situations. And I fear this is only going to get worse.

I had parents grow up in a communist country and the things the government used to do to extract this sort of information this horrific and now people are just giving it away for free.

Well yes it's good online literacy to not give sensitive info.

Do you really think chatgpt employees are committing the crime of blackmail with threads asking for recipes?

This is fear mongering nonsense.

Really? You know that there are hundreds of people doing romance scams (some of them basically as forced labour) all around the world. This sort of information would be worth something.

And at top of her post was counselling not asking how to make a pavlova.

I think anyone not questioning what the info could be used for is naive.

https://www.bbc.co.uk/news/articles/cpqeng9d20go

Everything except the courses I’m doing that it’s strictly prohibited on. It’s really helped me organise my thoughts on a lot of things

So this is your one extravagance when it comes to water and fossil fuel?

That's shortsighted thinking.

It's not OpenAI employees that are the risk (although OpenAI do reserve the right to use your queries as training data, which in turn means that it's possible for them to be regurgitated later on by ChatGPT).

Who owns the ChatGPT account, for example? For a lot of people, it's their employer. Who, therefore, has access to their query history? Strictly-speaking, it's also the employer (same as for emails on a work account, there is legal precedent for this).

As the use of LLMs becomes more pervasive, query histories are necessarily going to become more sensitive than browser histories. That means the compromise of an OpenAI account, or somebody getting hold of your browser with open sessions to AI services, is going to be more and more invasive.

That article says absolutely zero about your point.

That article is talking about the AI model resisting being deleted. Not chatgpt staff stealing info from chats (which are encrypted into numbers and dashes, yeah I've done research, not just listened, probably inaccuratly, to some 'high up tech dude from sky) and black mailing people.

And romance scams are a different thing altogether, nothing to do with AI and have been around since well before AI was a thing.

Chatgpt staff are categorically not interested in the average persons inane chats with it's bots.

That's my point, it ISN'T the AI employees. Literally what I was debunking in the previous post. Now had they said hackers, I'd have responded differently. I think the poster didn't quite comprehend what was being said by the sky employee and got confused.

If people are using work AI for personal reasons, then they are clearly idiots. Most are using the free chatgpt model.

Today I asked if it mining a 50x50 block of cobblestone to bedrock would be quicker using an Un-enchanted pickaxe or by turning the cobblestone into moss with bonemeal & using a shovel.

It was the latter by half.

Then I needed to know if moss is flammable. It is not.

Then I need to know if pale moss is flammable & it is!

Very exciting stuff.

I was probably trying to make too many points at once.
Romance scams are a big business. There are modern slaves holed up in offices being forced to carry out these scams. Numerous reports but here is one https://www.google.com/amp/s/www.bbc.co.uk/news/world-asia-68562643.amp

It doesn't take any imagination at all for these companies to target chat gbt style information, either using hacking or an insider who they pay off.

Currently they blackmail companies (see M&S, the Coop and many more). Again very easy to start doing this to individuals particularly when you have enslaved people doing it. They will find ways to use the information people have given to make them more obvious to attack. Someone is asking what to do with an inheritance whilst at the same time asking chat gbt about their shit love life or their fear they have dementia. Get A1 to sort out an algorithm on that and you have a potential easy prey. Of course it might not work but then you move on to the next 10000 other people on your list.
There will be a company willing and ready to sell this information. Just like they do when some poor person falls for a scam and then they get put on the list.

The fact that they won't delete information on request is where the worry lies. Let alone the fact we probably forget what we actually ask online.

I tell it to draw the characters in my novels. Quite amusing. It writes very bad literary short stories (clunk) and mechanically accurate poetry - does a good Petrachan sonnet - handy for practising them.

I ask it for current discount codes

I get it to code for me! My website has a lovely new JavaScript image carousel thanks to Chat GPT, plus some very posh looking 3D icons which it made for me in SVG format.

For making pictures of my cats doing silly things and that's about it!

If a person can hack into my ChatGPT to see my questions about my imaginary inheritance and lonely love life they can hack into my email / back accounts/ passwords and take the money directly.

I ask it for current discount codes

Thank you!