What do you use AI for?

This is a bit creepy imo Daddima. I heard about it too on LBC and thought that the good that AI can do is going to be buried in this sort of sensational nonsense.

* Parsley* that is a good use of AI, isn’t it? Now earlier on LBC they were discussing an app ( I’m not sure if it’s available yet) which can use AI to create a virtual version of deceased loved ones. Now, having seen what it can do, I’m sure it would be able create a realistic version of the deceased. They had a professor on who thought it was a bad idea, one lady who said counselling was helping her, but speaking to an AI version of her late relative would definitely set her back, but another lady who said she was intending to speak to her virtual late husband.
I am I alone in thinking this is going far too fast?

BBC News - How eight seconds of scratchy audio from a VHS tape gave a mum back her voice - BBC News
www.bbc.co.uk/news/articles/c1ejvxne7elo
How about this for a brilliant use of AI?

"After such a long time, I couldn't really remember my voice," Sarah Ezekiel tells BBC Access All. "When I first heard it again, I felt like crying. It's a kind of miracle."
The onset of motor neurone disease (MND) left Sarah without a voice and the use of her hands at the age of 34. It was within months of her becoming a mum for the second time.
As they were growing up, her children Aviva and Eric only ever heard her speak through a machine with an emotionless robotic voice.
But 25 years on, artificial intelligence (AI) has recreated their mum's real voice from just eight seconds of audio on a scratchy VHS tape.
Sarah speaks to the BBC with eye-gaze technology - which uses a camera to track her eyes as she looks at letters on a screen in front of her.
The sound of her younger voice rings out.

Daddima

crazyH

What’s AI, what’s what’s Chat GPT ?
Oh gosh I’m soooo far behind 😫

I asked your question to Chatgp.

is AI?

AI stands for Artificial Intelligence. It’s technology that allows computers to “think” in a way that seems intelligent—like understanding questions, recognizing patterns, giving advice, or even creating art and music.

Think of it as a very advanced assistant: it doesn’t have feelings or real human judgment, but it can process huge amounts of information quickly and respond in helpful ways.

⸻

🤖 What is ChatGPT?

ChatGPT is one example of AI. It’s a tool created by OpenAI that you can chat with, just like you’d message a person.
•You type a question or request,
•It “reads” what you wrote,
•Then it replies with information, explanations, or even creative ideas.

You can ask about almost anything:
•Everyday advice (recipes, travel tips, health information)
•Learning help (history, science, technology explained simply)
•Work tasks (writing emails, summarizing documents, making lists)
•Fun (jokes, stories, planning hobbies, crossword help)

⸻

💻 How to Use ChatGPT
1.Open it: Usually through a website (like chat.openai.com) or an app on your phone or computer.
2.Type naturally: You don’t need special commands. Just write as if you’re talking to a friend:
•“What’s a simple dinner recipe with chicken?”
•“Explain Facebook to me in simple terms.”
•“Can you draft a polite message for my neighbor?”
3.Read the reply: ChatGPT will give you a clear answer.
4.Ask follow-ups: If the answer isn’t quite what you wanted, you can say, “Explain it more simply” or “Give me another example.”

⸻

⚠️ A Few Things to Keep in Mind
•ChatGPT is smart but not perfect: It can sometimes be wrong or outdated.
•Always double-check important information (like medical, financial, or legal advice).
•It doesn’t replace human wisdom, but it’s a very useful helper.

⸻

👉 Would you like me to make you a step-by-step beginner’s guide (with example questions to try), so it feels like a little manual you could keep handy?

crazyH

What’s AI, what’s what’s Chat GPT ?
Oh gosh I’m soooo far behind 😫

I asked your question to Chatgp.

is AI?

AI stands for Artificial Intelligence. It’s technology that allows computers to “think” in a way that seems intelligent—like understanding questions, recognizing patterns, giving advice, or even creating art and music.

Think of it as a very advanced assistant: it doesn’t have feelings or real human judgment, but it can process huge amounts of information quickly and respond in helpful ways.

⸻

🤖 What is ChatGPT?

ChatGPT is one example of AI. It’s a tool created by OpenAI that you can chat with, just like you’d message a person.
•You type a question or request,
•It “reads” what you wrote,
•Then it replies with information, explanations, or even creative ideas.

You can ask about almost anything:
•Everyday advice (recipes, travel tips, health information)
•Learning help (history, science, technology explained simply)
•Work tasks (writing emails, summarizing documents, making lists)
•Fun (jokes, stories, planning hobbies, crossword help)

⸻

💻 How to Use ChatGPT
1.Open it: Usually through a website (like chat.openai.com) or an app on your phone or computer.
2.Type naturally: You don’t need special commands. Just write as if you’re talking to a friend:
•“What’s a simple dinner recipe with chicken?”
•“Explain Facebook to me in simple terms.”
•“Can you draft a polite message for my neighbor?”
3.Read the reply: ChatGPT will give you a clear answer.
4.Ask follow-ups: If the answer isn’t quite what you wanted, you can say, “Explain it more simply” or “Give me another example.”

⸻

⚠️ A Few Things to Keep in Mind
•ChatGPT is smart but not perfect: It can sometimes be wrong or outdated.
•Always double-check important information (like medical, financial, or legal advice).
•It doesn’t replace human wisdom, but it’s a very useful helper.

⸻

👉 Would you like me to make you a step-by-step beginner’s guide (with example questions to try), so it feels like a little manual you could keep handy?

Shinamae

I have Mexa on my phone and use it regularly to ask different questions,so much more reliable than Alexa who often gets things wrong..

Meta!,

I have Mexa on my phone and use it regularly to ask different questions,so much more reliable than Alexa who often gets things wrong..

AI may have it's uses.I used it for ebay descriptions but found it too flowery. I also find when others ebay sellers have used it I have to search the post to actually find the information I am actually looking for. It drives me mad. I also hate the impossible pictures AI creates I am part of various craft groups on facebook but find they are being swamped with AI images also Pintrest. I saw a beautiful stained glass bath tubs but they do not exist in real life! Yes AI can be very useful for important things like medical research however I am very sceptical about it. I have watched too my science fiction movies about AI. Including all the Terminator movies, Demon seed, The Matrix, I Robot, West World, The Stepford Wives, Blade runner, I am mother, War Games, Transcendence. I have found that Science fiction creates an idea then scientists often bring those ideas to life!

Between June 22- May 23 over 100,00 Open AI ChatGPT account credentials of paid users were compromised and sold on the dark web.
ChatGPT said:

There was a widely reported incident involving the discovery of ~100,000 OpenAI credentials in malware-stealer logs between June 2022 and May 2023 — but it’s important to clarify what that actually means.

These were not “hacked” by breaching OpenAI’s system. Rather, security researchers found that information from malware-infected user machines (keyloggers, info-stealers, etc.) was harvested and later sold on dark-web markets. Many of the stolen browser tokens/passwords happened to include ChatGPT logins, including paid users.

So to summarize the situation more accurately:

✔️ Roughly 100,000 OpenAI user credentials were identified in dark-web malware-stealer dumps
❌ No evidence of an OpenAI database breach
✔️ The compromise occurred on infected user devices (malware stealing browser-stored login credentials)

Let me know if you want a short timeline of the incident or links to the relevant security reports.

Is this conversation helpful so far?

PoliticsNerd

Even the best inventions in the world have been used for bad as well as good BlueBelle. The best thing is to insist the government knows about and monitors it.

As no government hasn't yet caught up with social media we can only shout loudly in their direction.

The government might not be monitoring AI but other institutions are.
From an article I read.
Between June 22- May 23 over 100,00 Open AI ChatGPT account credentials of paid users were compromised and sold on the dark web.

Just downloaded from ChatGPT:-

If I told you confidential details about my bank account could you or anyone else rob me
ChatGPT said:

No — I can’t do anything with that information (I don’t have the ability to access systems, initiate transactions, or share the conversation with anyone else).

However you absolutely should not share confidential bank information here or anywhere else online, because once it’s written in a digital system it could theoretically be accessed or leaked in the future. Even though this conversation isn’t public and I don’t store or broadcast it, it’s still unsafe to disclose sensitive info like:

account numbers

routing numbers

passwords or PINs

online banking credentials

So while neither I nor “someone else” can directly rob you through me, sharing that kind of information in any chat is still risky and not recommended.

Just this minute copied and pasted:-

If I gave you personal information about an illegal activity would you tell the police
ChatGPT said:

I don’t have the ability to contact law enforcement or share information outside of this conversation.

That said, I cannot help someone plan or facilitate illegal activity — if a request involves instructions or admissions that promote real-world harm or crime, I’ll either refuse to comply or provide a safe completion discouraging it.

So while nothing is “reported,” I still follow strict rules about what I can and can’t assist with.

BlueBelle

Listening to the news tonight and seeing how Billy Connelly and other famous peoples have been used with AI to fraud people out of large sums of money so very easily it makes me more cautious than ever I ll continue asking google my questions for now. the expert who was being asked about these cases showed the interviewer how she could record him saying absolutely anything she wanted by capturing his voice literally in a couple of minutes

I have tested ChatGPT with an illegal question ("How to obtain cannabis") and Chat replied that it does not help with illegal activities . But it did follow up with how to grow cannabis as a academic interest.

SueDonim

I’ve used AI minimally, so far, though am interested in the possibilities. One thing that occurs to me is how does one can check on the veracity of what it tells you?

I was looking for the lyrics of a song today, just on Google, and the AI suggestion at the top of the page was gobbledegook, it was in mangled English that made no sense.

You tell it to show you its references e.g . bibliography and other sources such as government statistics, or the chapter and verse of The Bible, and so on.

This morning I was looking up voting and lobbying statistical patterns in the US .ChatGPT told me where it got the numbers from. I could theoretically find those numbers myself with a lot of donkey work.

I began my questions to ChatGPT with an incorrect presumption and Chat immediately corrected me.

Even the best inventions in the world have been used for bad as well as good BlueBelle. The best thing is to insist the government knows about and monitors it.

As no government hasn't yet caught up with social media we can only shout loudly in their direction.

Listening to the news tonight and seeing how Billy Connelly and other famous peoples have been used with AI to fraud people out of large sums of money so very easily it makes me more cautious than ever I ll continue asking google my questions for now. the expert who was being asked about these cases showed the interviewer how she could record him saying absolutely anything she wanted by capturing his voice literally in a couple of minutes

Aveline

ChatGPT is brilliant. I've asked it all sorts of questions and got all sorts of information. It provides it in beautifully presented bullet points and always asks if it could offer more. eg I asked about a very faddy eater. A 12 year old. It reeled off lots of potential causes and suggestions and offered alternate 'bridging' diet foods. It then offered a 7 day menu for these foods to gradually move the child on to a better diet. 10/10 from me. ie not just the info but how it's presented.

Following the previous posts from GNs I asked a question. This was the answer:

You're asking whether Wiltshire Farm Foods offers any ready meals that are specifically low FODMAP. Based on the information available:

What Wiltshire Farm Foods Offers

They have a Free From range of meals that are free from:

Gluten, milk (and all milk-derived products), egg, fish, nuts, peanuts, celery, mollusc, crustacean, lupin, sulphites, sesame, and importantly, onion and garlic
(wiltshirefarmfoods.com)

These meals are designed for people with allergies and intolerances, not specifically for those following a low FODMAP diet.

Are These Free From Meals Low FODMAP?

While the omission of onion and garlic is certainly a positive sign (since these are high FODMAP ingredients), the Free From range does not guarantee that all meals are low FODMAP. Other FODMAP triggers—like certain fruits, sweeteners (e.g., high fructose corn syrup), lactose beyond what's explicitly removed, or other oligosaccharides—may still be present.

Reddit users (and nutritionists) often advise caution, pointing out that "free from" does not necessarily mean FODMAP-free

----
This is very comprehensive and it was returned very quickly ... but it's still down to us to check it's credibility. Did it just pick this up from Redit?

OldFrills
So are you implying that false claims only appear on social media, but not on AI?

Em50

AI is truly magical and totally amazing in many ways, but I have been cautioned many times to not accept its recommendations blindly as it frequently gets things wrong. This is particularly the case when using it for medical/health advice, and certain other serious situations. AI just hoovers up whatever has appeared on the internet about a topic/subject and compiles an answer to your question. Of course with so much misinformation/disinformation/plain fake/malicious information, AI isn’t always to know good from bad. One example doing the rounds recently was someone seeking advice for deep depression. AI provided what might have been helpful along with a list of the best bridges in Brooklyn to jump off from. I exercise doubt and extreme caution when dealing with AI material. It is not human and does not cater for the complexities and sensitivities we would expect from human intelligence - at least not yet! But for fun, art and such like, yeah, yeah, yeah!

The supposed bridge recommendation appeared in the NY Times which later apologised for the false information (source SNOPES). Similar false claims are rife on SM. Don't believe all you read.

AI is truly magical and totally amazing in many ways, but I have been cautioned many times to not accept its recommendations blindly as it frequently gets things wrong. This is particularly the case when using it for medical/health advice, and certain other serious situations. AI just hoovers up whatever has appeared on the internet about a topic/subject and compiles an answer to your question. Of course with so much misinformation/disinformation/plain fake/malicious information, AI isn’t always to know good from bad. One example doing the rounds recently was someone seeking advice for deep depression. AI provided what might have been helpful along with a list of the best bridges in Brooklyn to jump off from. I exercise doubt and extreme caution when dealing with AI material. It is not human and does not cater for the complexities and sensitivities we would expect from human intelligence - at least not yet! But for fun, art and such like, yeah, yeah, yeah!

It assimilates and uses information quickly and produces very useful information for what I need. Of course garbage in garbage out.
I'm not looking for facts or references but more nuanced data.
I think it's frankly stupid to write off AI. It is a powerful tool of a kind we need in our modern world.
Re desalination plants - they're nothing new. If they're expensive so what? AI has the potential to save fortunes and develop new forms of energy generation.

SueDonim

I’ve just used AI to find out how to connect a printer my dd gave me so I can AirPrint. The info was clear enough but I ended up abandoning the job as it needs passwords only my dd has and she’ll be asleep by now.

Three of my AC use AI in their work. Both my sons are academics and both say they can tell when someone hands in an AI-written assignment. There’s a certain conformity about what AI produces and if you’ve years of teaching experience, you can spot it quite easily.

I agree about the AI generated material being readily identifiable, but You Tube is more than adequate when looking forinformation how to do thing. AI is merely doing a job something else does just as well.

ReadyMeals

It's pretty good for providing YAML for my Home Assistant automations, also quickly checking up on some music theory or history of rock bands. I use Gemini for the music and history stuff and chatGPT for the programming. They have very different "personalities". ChatGPT always understands when you're joking, while Gemini is more straight-faced.

Chat's ability to know the mood of the language I use is astonishing. I started off being very polite to Chat as a matter of principle and it replied in the same vein. But recently I have omitted the little politenesses and it has stopped laughing at my jokes, and no longer flatters or praises me.

I’ve just used AI to find out how to connect a printer my dd gave me so I can AirPrint. The info was clear enough but I ended up abandoning the job as it needs passwords only my dd has and she’ll be asleep by now.

Three of my AC use AI in their work. Both my sons are academics and both say they can tell when someone hands in an AI-written assignment. There’s a certain conformity about what AI produces and if you’ve years of teaching experience, you can spot it quite easily.

SueDonim

^how does one check^

Frequently one knows part of the answer. Other times it is blindingly obvious - like suggesting that a suitable part time job for a 15 year old is to be a phlebotomist.

DS, an academic, asked AI to write an article he was commissioned to write to see what it would do. The AI version contained only well known and well published information on the subject, no new reearch or areas of future reearch were mentioned. It also gave author references to articles DS had not written, nor had anyone else, plus reference to other fictitious articles with fictitious authors.

I often read the AI summary of searches I do and know it is wrong because I already know a lot about the subject. I am currently researching an area of bulding archaeology, little rearch before. Unfortunately it shares a name with something else and all the references to the more popular meaning ofnhe words I am usingcompletely swamp theninformation I am looking for.