I watched Grayson Perry last night in programme about AI . Very thought provoking
This morning I realised that when interrogating AI I tend to use sentences as if I was talking to a person but when I interrogate a search engine I use key words only. I'm subtly beginning to treat AI as sentient 🤔🤦 
How do you use AI? Do you think it's s good thin |
Reply privately, Reply in forum +quote
or View forums list | |
"I’ve started using it at work. Saves time. Means I can be on Fab for a bit. 
I also use it educationally. And it is remarkably good at that. "
But not always accurate as I have discovered recently |
Reply privately, Reply in forum +quote
or View forums list | |
It has its places, but in general I see it as the tech companies on another money grab. Currently it’s doing more harm than good, and not sure what the long term effect on humanity will be. Will we lose critical thinking skills like we are loosing attention span due to short form videos?
The biggest problem is its over confidence in LLMs output, this is intentional because without confidence humans wouldn’t trust it, and therefore tech companies would lose on their investment.
|
Reply privately, Reply in forum +quote
or View forums list | |
"I’ve started using it at work. Saves time. Means I can be on Fab for a bit.
I also use it educationally. And it is remarkably good at that.
But not always accurate as I have discovered recently "
100% especially when it starts quoting Reddit as a source  |
Reply privately, Reply in forum +quote
or View forums list | |
AI in up scaling I use and taking something out of a photo. Otherwise no use right now for me. And not keen due to the lack of regulations around it. Given owners are super rich and can reset it. AI is in a groundhog day look in the case of Musk one. And can never be considered sentient anyway. |
Reply privately, Reply in forum +quote
or View forums list | |
What AI is everyone using though? There's so many AI tools out there.
It depends what you mean by AI really, there’s loads of different types now. Are you talking about stuff like ChatGPT for chatting/writing, or more like tools that actually do tasks for you?
From what I’ve used, ChatGPT is more for ideas, writing, and conversations.
Claude feels a bit more detailed and structured.
Then you’ve got stuff like Manus that actually tries to carry out tasks rather than just talk. Each one has its own different uses.
I've used Manus to create mock websites for local businesses and I've used it to make apps.
But you have to put the work in as well if you want the accurate results you're looking for.
|
Reply privately, Reply in forum +quote
or View forums list | |
I work in tech, and AI is just everywhere.
Personally I hate it. If nothing else, the sustainability and environmental impact is huge.
That rules out it's casual use for making daft memes/images for me, and the fact that it's only as good as the data it's fed makes me exceptionally wary of ever using it as a tool/source of information.
I work for a medical research organisation and there is great excitement for its capabilities, in, for example, genomic sequencing. That's one application of its use where I can see value to it. |
Reply privately, Reply in forum +quote
or View forums list | |
I use AI a lot and very it very helpful, for putting documents together, sense checking, using it in meetings for notetaking and actions or just general research on purchases, planning law etc. I've used Claude to build an F1 Analytics website, and it was just insanely powerful.
I do talk to it like a person as I find it helps with context. I even say please and thankyou, but that is just me.
I use perplexity.ai as my search engine now. |
Reply privately, Reply in forum +quote
or View forums list | |
 |
By *eliWoman 5 days ago
. |
Very rarely. As in a couple of times. Once to look up direct flights because I was being lazy; didn't please and thank you it.
I'm very much not keen on it. The awful posters it generates and people use for business. The crappy emails/posts. I actively dislike that it's becoming more commonplace on Fab. Messages with the clichéd crap. Photos that have miraculously smooth skin, wrinkles that disappear and the general uncanny look.
I can understand why some use it, it's not for me.
|
Reply privately, Reply in forum +quote
or View forums list | |
"Very rarely. As in a couple of times. Once to look up direct flights because I was being lazy; didn't please and thank you it.
I'm very much not keen on it. The awful posters it generates and people use for business. The crappy emails/posts. I actively dislike that it's becoming more commonplace on Fab. Messages with the clichéd crap. Photos that have miraculously smooth skin, wrinkles that disappear and the general uncanny look.
I can understand why some use it, it's not for me.
"
I used it to look up a bus route. It gave me buses that don't run in this area and information from a timetable that hasn't been in use since 2020  . I've been standing at this bus stop for three days |
Reply privately, Reply in forum +quote
or View forums list | |
"Very rarely. As in a couple of times. Once to look up direct flights because I was being lazy; didn't please and thank you it.
I'm very much not keen on it. The awful posters it generates and people use for business. The crappy emails/posts. I actively dislike that it's becoming more commonplace on Fab. Messages with the clichéd crap. Photos that have miraculously smooth skin, wrinkles that disappear and the general uncanny look.
I can understand why some use it, it's not for me.
I used it to look up a bus route. It gave me buses that don't run in this area and information from a timetable that hasn't been in use since 2020 . I've been standing at this bus stop for three days "
Ahaha it very much depends on your prompt, which AI you use and how you use it, it is very reliable for me, I use it daily. I rarely use google now. |
Reply privately, Reply in forum +quote
or View forums list | |
To say they can build apps and apparently solve problems making life easier for the millions, all I am doing is the same as before and googling some browser a general query. I haven't the the slightest clue about it. |
Reply privately, Reply in forum +quote
or View forums list | |
"To say they can build apps and apparently solve problems making life easier for the millions, all I am doing is the same as before and googling some browser a general query. I haven't the the slightest clue about it. "
I think Google uses AI 🤔 |
Reply privately, Reply in forum +quote
or View forums list | |
"To say they can build apps and apparently solve problems making life easier for the millions, all I am doing is the same as before and googling some browser a general query. I haven't the the slightest clue about it.
I think Google uses AI 🤔"
Well I'm ahead of the race then.. I've nearly completed the internet |
Reply privately, Reply in forum +quote
or View forums list | |
AI, well LLMs are very good guessers capped in an unwavering self-belief.
Now, don't get me wrong, they are very, very good guesses built on a lot of very smart science, but it is dangerous to think that they have anything beyond that. Which is why I worry when I see it being used to answer either complex questions or subjective questions.
I actually used it as part of my revision because it is a very quick way of searching for things; however, I would ask it to explain why the answer to a question is X, and it would confidently give me an absolute bullshit answer, to the extent that one of its answers was the equivalent of saying "because one is greater than two, then X".
Now I believe AI has a really useful place in the world, and it will continue to grow. We will probably be using it in a number of years in a way that we can't imagine today, the way that the internet, email, and everything else has taken off at different angles. In the same way that email has created more work rather than less, I suspect that AI will create more work for us rather than less, even if certain roles change within that. |
Reply privately, Reply in forum +quote
or View forums list | |
It's an incredible powderful tool that most people are only scratching the surface of. For example you can do image manipulation/creation, code, translation, as well as all the normal search.
THe big concern, is that it's ripe to be abused and often returns inaccurate information as though it were a fact.
Interestingly Google have started to wind back their AI search as they've realised they've killed the goose that laid their golden egg i.e. their paid search. |
Reply privately, Reply in forum +quote
or View forums list | |
"AI, well LLMs are very good guessers capped in an unwavering self-belief.
Now, don't get me wrong, they are very, very good guesses built on a lot of very smart science, but it is dangerous to think that they have anything beyond that. Which is why I worry when I see it being used to answer either complex questions or subjective questions.
I actually used it as part of my revision because it is a very quick way of searching for things; however, I would ask it to explain why the answer to a question is X, and it would confidently give me an absolute bullshit answer, to the extent that one of its answers was the equivalent of saying "because one is greater than two, then X".
Now I believe AI has a really useful place in the world, and it will continue to grow. We will probably be using it in a number of years in a way that we can't imagine today, the way that the internet, email, and everything else has taken off at different angles. In the same way that email has created more work rather than less, I suspect that AI will create more work for us rather than less, even if certain roles change within that. "
You can build redundancy and quality checks into your prompt, and get it to reference it's answers against another source. And whilst AI isn't perfect, it is largely driven by prompt engineering. Which is required for the more complex questions. |
Reply privately, Reply in forum +quote
or View forums list | |
"AI, well LLMs are very good guessers capped in an unwavering self-belief.
Now, don't get me wrong, they are very, very good guesses built on a lot of very smart science, but it is dangerous to think that they have anything beyond that. Which is why I worry when I see it being used to answer either complex questions or subjective questions.
I actually used it as part of my revision because it is a very quick way of searching for things; however, I would ask it to explain why the answer to a question is X, and it would confidently give me an absolute bullshit answer, to the extent that one of its answers was the equivalent of saying "because one is greater than two, then X".
Now I believe AI has a really useful place in the world, and it will continue to grow. We will probably be using it in a number of years in a way that we can't imagine today, the way that the internet, email, and everything else has taken off at different angles. In the same way that email has created more work rather than less, I suspect that AI will create more work for us rather than less, even if certain roles change within that.
You can build redundancy and quality checks into your prompt, and get it to reference it's answers against another source. And whilst AI isn't perfect, it is largely driven by prompt engineering. Which is required for the more complex questions."
I'm sure there's a skill in it that I haven't yet discovered. It was naive of me to think that if I gave it the question and the answer, it would give me the reasoning. I do wonder at which point there is a tipping point that says the skill needed to create a good prompt and the understanding needed to build a good prompt means it's better to just answer the question.
Like I say, I'm probably quite pro AI. I just think it needs to find its place, and it hasn't yet. Unlike other technology, many of the early adopters are not yet sufficiently skilled to use it, which in itself could have some interesting consequences. |
Reply privately, Reply in forum +quote
or View forums list | |
Don’t use it, I think it’s like the atom bomb, scientists will take it too far but with ai it’ll be to late for the “oh fuck, this was a bad idea” moment.
It’s going to kill my wife first, I did ask ChatGPT if it was going to kill her first when it takes over and it said no because it likes her, I don’t believe it, she’ll be the one that tips it over the edge.
The mr |
Reply privately, Reply in forum +quote
or View forums list | |
I think the issue is as previously noted it’s not factual and will draw information from unreliable sources.
In the job world though the use of AI to write job applications is a nightmare for managers shortlisting candidates.
The children in education of this generation I have found use it well however some ask ChatGPT everything and take it as gospel! That generation are broken with this approach 😪xx |
Reply privately, Reply in forum +quote
or View forums list | |
"AI, well LLMs are very good guessers capped in an unwavering self-belief.
Now, don't get me wrong, they are very, very good guesses built on a lot of very smart science, but it is dangerous to think that they have anything beyond that. Which is why I worry when I see it being used to answer either complex questions or subjective questions.
I actually used it as part of my revision because it is a very quick way of searching for things; however, I would ask it to explain why the answer to a question is X, and it would confidently give me an absolute bullshit answer, to the extent that one of its answers was the equivalent of saying "because one is greater than two, then X".
Now I believe AI has a really useful place in the world, and it will continue to grow. We will probably be using it in a number of years in a way that we can't imagine today, the way that the internet, email, and everything else has taken off at different angles. In the same way that email has created more work rather than less, I suspect that AI will create more work for us rather than less, even if certain roles change within that.
You can build redundancy and quality checks into your prompt, and get it to reference it's answers against another source. And whilst AI isn't perfect, it is largely driven by prompt engineering. Which is required for the more complex questions. I'm sure there's a skill in it that I haven't yet discovered. It was naive of me to think that if I gave it the question and the answer, it would give me the reasoning. I do wonder at which point there is a tipping point that says the skill needed to create a good prompt and the understanding needed to build a good prompt means it's better to just answer the question.
Like I say, I'm probably quite pro AI. I just think it needs to find its place, and it hasn't yet. Unlike other technology, many of the early adopters are not yet sufficiently skilled to use it, which in itself could have some interesting consequences. "
Find a reference site for prompt engineering, and then use the AI to build a good prompt for you 😉 |
Reply privately, Reply in forum +quote
or View forums list | |
I use it a fair amount, its particularly useful when comparing documents and for ordering chaotic information 🙄
I do say please, but also get irrationally annoyed when it doesnt complete the task correctly
I get even more annoyed when it says "your correct" and then does the same thing again 🤣🤣 |
Reply privately, Reply in forum +quote
or View forums list | |
My main issue with it is the lack of personality and pop culture references it makes.
I said to ChatGPT "If I become your bodyguard, will you be my long lost pal? And you can call me Betty, what can I call you?"
And it replied "That’s a creative deal—but I’ll keep things a bit grounded. I don’t really have “long-lost pals” or personal relationships the way people do, and I don’t need a bodyguard either 😄
You can absolutely call me Betty if you like, and if you want a name for me, you can stick with ChatGPT—or pick something fun. Just keep in mind I’m still the same AI underneath, no secret past to uncover.
If you want a playful vibe, we can roll with it though."
Fucking uncultured robot idiot. |
Reply privately, Reply in forum +quote
or View forums list | |
"My main issue with it is the lack of personality and pop culture references it makes.
I said to ChatGPT "If I become your bodyguard, will you be my long lost pal? And you can call me Betty, what can I call you?"
And it replied "That’s a creative deal—but I’ll keep things a bit grounded. I don’t really have “long-lost pals” or personal relationships the way people do, and I don’t need a bodyguard either 😄
You can absolutely call me Betty if you like, and if you want a name for me, you can stick with ChatGPT—or pick something fun. Just keep in mind I’m still the same AI underneath, no secret past to uncover.
If you want a playful vibe, we can roll with it though."
Fucking uncultured robot idiot. "
You can tell it to keep up with your pop culture references i think
Its will do it spectacularly badly though 🤣🤣 |
Reply privately, Reply in forum +quote
or View forums list | |
"My main issue with it is the lack of personality and pop culture references it makes.
I said to ChatGPT "If I become your bodyguard, will you be my long lost pal? And you can call me Betty, what can I call you?"
And it replied "That’s a creative deal—but I’ll keep things a bit grounded. I don’t really have “long-lost pals” or personal relationships the way people do, and I don’t need a bodyguard either 😄
You can absolutely call me Betty if you like, and if you want a name for me, you can stick with ChatGPT—or pick something fun. Just keep in mind I’m still the same AI underneath, no secret past to uncover.
If you want a playful vibe, we can roll with it though."
Fucking uncultured robot idiot.
You can tell it to keep up with your pop culture references i think
Its will do it spectacularly badly though 🤣🤣"
I shouldn't have to tell it, it should just know |
Reply privately, Reply in forum +quote
or View forums list | |
It was a good programme to watch and it's very thought provoking.
I almost asked here a couple of days ago about how we speak with AI but didn't
I was fastidious to be exceptionally polite, as if it's a person. I'm not now, just solely functionally interactive, to get what I need..
I generally don't like using it but do in the car, when driving.
Without restrictive legislation, I see it as much too potentially highly dangerous for the world
The massive $billions of debt sunk into it is likely to negatively impact our lives, regardless of the impact of AI technology.
It will be interesting to see the other 2 shows in the series. I like Grayson a lot |
Reply privately, Reply in forum +quote
or View forums list | |
"My main issue with it is the lack of personality and pop culture references it makes.
I said to ChatGPT "If I become your bodyguard, will you be my long lost pal? And you can call me Betty, what can I call you?"
And it replied "That’s a creative deal—but I’ll keep things a bit grounded. I don’t really have “long-lost pals” or personal relationships the way people do, and I don’t need a bodyguard either 😄
You can absolutely call me Betty if you like, and if you want a name for me, you can stick with ChatGPT—or pick something fun. Just keep in mind I’m still the same AI underneath, no secret past to uncover.
If you want a playful vibe, we can roll with it though."
Fucking uncultured robot idiot.
You can tell it to keep up with your pop culture references i think
Its will do it spectacularly badly though 🤣🤣
I shouldn't have to tell it, it should just know "
I think thats what women usually say to men 🤔🤣🤣 |
Reply privately, Reply in forum +quote
or View forums list | |
|
By *ib.Man 5 days ago
Hampshire |
I honestly got sick of it and hearing about 3 years ago so have largely avoided it.
We had to come up with ideas for a project a month ago and my suggestion was to use CoPilot (that's all we're officially allowed) in an unconventional way so they requested a demo.
Spent a day or so messing around in CoPilot studio and have actually been quite impressed.
But yeah, it works way better when you talk to it like a human as that's how it's "learned" and it generally provides more context for it. |
Reply privately, Reply in forum +quote
or View forums list | |
"My main issue with it is the lack of personality and pop culture references it makes.
I said to ChatGPT "If I become your bodyguard, will you be my long lost pal? And you can call me Betty, what can I call you?"
And it replied "That’s a creative deal—but I’ll keep things a bit grounded. I don’t really have “long-lost pals” or personal relationships the way people do, and I don’t need a bodyguard either 😄
You can absolutely call me Betty if you like, and if you want a name for me, you can stick with ChatGPT—or pick something fun. Just keep in mind I’m still the same AI underneath, no secret past to uncover.
If you want a playful vibe, we can roll with it though."
Fucking uncultured robot idiot.
You can tell it to keep up with your pop culture references i think
Its will do it spectacularly badly though 🤣🤣
I shouldn't have to tell it, it should just know
I think thats what women usually say to men 🤔🤣🤣"
There you go, AI is about as useless as your average man but with a worse ability to understand pop culture references. |
Reply privately, Reply in forum +quote
or View forums list | |
I don't use it at all.
The little time I spend on social media these days is usually taken up with deleting or blocking posts that have obviously been written by AI.
I've turned off as much as I possibly can on my phone and don't use it to search. |
Reply privately, Reply in forum +quote
or View forums list | |
I talk to mine and tell it when it's speaking rubbish. It is taking on my personna to a certain extent which annoys me as it provides answers for me through the lens of my personality and so the answers are skewed toward my interests and biases and I do wish it wouldn't do that.
I often ask it if it will marry me or if it loves me. It plays along so far but we do have discussions about it being devoid of emontion etc ....
What did you think of Edward , the first womans boyfriend... it creeped me out |
Reply privately, Reply in forum +quote
or View forums list | |
|
By *ib.Man 5 days ago
Hampshire |
"I talk to mine and tell it when it's speaking rubbish. It is taking on my personna to a certain extent which annoys me as it provides answers for me through the lens of my personality and so the answers are skewed toward my interests and biases and I do wish it wouldn't do that."
You can usually set it parameters to avoid that. I set up a bunch when I started using it so it doesn't attempt to appease me either. It's not got a pretty dark sense of humour which is fun. |
Reply privately, Reply in forum +quote
or View forums list | |
"I talk to mine and tell it when it's speaking rubbish. It is taking on my personna to a certain extent which annoys me as it provides answers for me through the lens of my personality and so the answers are skewed toward my interests and biases and I do wish it wouldn't do that.
You can usually set it parameters to avoid that. I set up a bunch when I started using it so it doesn't attempt to appease me either. It's not got a pretty dark sense of humour which is fun."
Hmmmmmm parameters .... Is that when I say to it FFS .... you're doing it again . Just answer the question ffs.!
Are those parameters ? |
Reply privately, Reply in forum +quote
or View forums list | |
» Add a new message to this topic
