Anyone else routinely using AI in their day-to-day life now?

[mybestchildismycat]

I started playing about with chatgpt a couple of months ago, but it's only really been in the last couple of weeks I've really got the hang of prompting it and I'm just blown away.

I'm using it daily for work: probably about half the work I do can benefit from chatgpt in some shape or form. But I'm also using it personally too - in just the last few days its helped me to turn some random ingredients into an edible meal, plan a city break against a specific set of requirements, recommend a new card game that our kids would enjoy, diagnose why my plant is losing it's leaves... the list goes on and on.

What made me really stop and think today was that I googled something and found myself thinking how... underwhelming it was. Just pages and pages of links that I then have to figure out myself. It shocked me to realise how quickly my own expectations and behaviours have shifted.

Anyone else finding chatgpt (and other similar technologies) are creeping into their everyday lives? In what ways?

The risk then is an even greater disparity between the haves (of knowledge and power) and have nots

You say that like it's a bad thing.

I've had a few plays around with it. I follow the subreddit for chatgpt and it's fascinating to see what people are using it for and how life changing it has been for some for specific uses. Especially disabilities and neuro diversities. There's a few other LLMs, one website uses one as like a chat therapist which I could see being handy for some.

I couldn't log in to chatgpt last time I tried, so haven't used it for a while, but I love seeing how other people are using it.
My job (practice nurse) isn't one you could currently apply it to, but in the future I can see it making some complex work easier (like working out which vaccinations people need for travel as its nice having someone else to check your plan with but I don't usually have that).

"AI Wrangler" is a term I quite like.

You still need to be able to think critically and check facts, or you'll not only end up believing things that are absolute bollocks, but you'll lose those abilities completely.

In a country where people literally believed a bus, that isn't going to mean much.

As someone who has no idea what it is, can someone give me a quick breakdown please?

Not allowed to use ChatGPT etc in work because the whole thing is a data sponge so any proprietary material or material that could be considered intellectual property isn't allowed to be entered into generative AI tools. It would be considered gross misconduct as expressively forbidden in our IS policy. Frankly I can't say I'm surprised - the company sack people for emailing themselves presentations, there's no way they'd consent to information being fed into a database they have no control over.

On a personal level, I have no interest in AI beyond the Google algorithms when I search for something. I don't want to be spoon fed information, I want to read what's available and make up my own mind about what's right/wrong workable/unworkable.

The internet has made the need for critical thinking even more important - AI would happily kill that off and create a generation of people who believe whatever a computer programme wants them to believe.

On a personal level, I have no interest in AI beyond the Google algorithms when I search for something.

That really isn't what Large Language Learning is about.

The internet has made the need for critical thinking even more important - AI would happily kill that off and create a generation of people who believe whatever a computer programme wants them to believe.

There really isn't any evidence that this isn't what is happening anyway. Did you know that the only reason we believe the earth isn't flat is because NASA are paying to put globes in schools and houses and on tv, and as the argument goes "Why would you spend that much money if it's not propaganda ?"

Sorry, are you actually a flat earther or was that sarcastic?

That really isn't what Large Language Learning is about.

Fair enough, but internet search algorithms were the gateway to AI and neural networks. On a personal level, that's where I'm happy to interact with AI.

There really isn't any evidence that this isn't what is happening anyway. Did you know that the only reason we believe the earth isn't flat is because NASA are paying to put globes in schools and houses and on tv, and as the argument goes "Why would you spend that much money if it's not propaganda ?"

The first part I agree with, which is why I'm teaching my children how to exercise critical thinking skills when taking in information. Too many people believe what they are comfortable with believing and dismiss any evidence to the contrary...which brings me to your second point about NASA - I'm sorry but that's just batshit. People believed the world was a globe thanks to Magellan hundreds of years before NASA was even conceived, let alone put a satellite into space. You lost all credibility with that one plus it was the Russians who were first into space

I sometimes used apps like Jasper in the past, but yeah, ChatGPT knocks it out of the park... I think I haven't written a single email without AI help at least since December 22'... (I've been helping my buddy with his dropshipping site on the customer experience end, mostly trying to craft personalized messages based on past purchases). AI tools are super-helpful in that regard!

Fair enough, but internet search algorithms were the gateway to AI and neural networks.

hmmm, I don't believe so. Internet search algorithms exist to ensure you get to see what advertisers are paying google for you to see. They didn't start that way, admittedly. But that's all they do nowdays. Depending on what you are looking for (or, more precisely what you are not looking for) Google has become useless.

LLMs are a completely different beast. They genuinely have surpassed the clever fast pattern matching that was previous "AI", and are capable of developing patterns in real time that are unique as they feed back into the model.

Do you think that the people you send the email to, or give the speech to, would be able to tell? Because if they can I wouldn't like it.

Thanks for this, I will be seeking out more Young Turks discussions. I’m also enjoying Marianna Spring on the BBC.

That Kandiss woman is insane, glad you are not! Funny but slightly terrifying.

If you narrow your idea of the Turing test down a bit, I'd say AI has already passed. With flying colours.

@SerendipityJane doesn’t believe the “Big Globe Conspiracy” @LittleLegsKeepGoing She was trying to make a point.

That Kandiss woman is insane, glad you are not!

I know it's snippy, and not in the spirit of MN, but the spelling of the name rather gave the game away.

Didn’t it just!

Also loving the name of her podcast “Jesus, Guns and babies”.

Also loving the name of her podcast “Jesus, Guns and babies”.

You've got it wrong. It's:

JESUS GUNS BABIES

since apparently commas are a communist plot - and if they isn't then why does that word sound like "commie" when y'all say it in proper Inglish

But coming back to the original question, is elected politician Kandiss not a clear example of how some Human Intelligence is so lacking that perhaps we’re in safer hands with the artificial type? It all depends on whose voices are the loudest as the LLMs gather their data I suppose.

When I've experimented with it for work it's given very convincing and detailed answers but they're not actually correct 😀 i.e. it's suggested using a particular feature of a program that's only actually an idea someone has posted on the web, it's not actually a functional feature.

A colleague is using the paid for version and is finding it very useful for drafting clauses for contracts. But he says it has involved a lot of patience training it.

My son used it to generate a research proposal - more of a checklist really, which gave him a framework to create one.

I think the key is remaining boss! When I've used it (the open, less advanced version) it was a bit rubbish. I have also used another AI application to summarise court judgments and that was a bit rubbish too.

The problem for me is having to check everything, I might as well just start from scratch myself.

Interesting that someone says they are using Co-pilot. I want to use that but we can't get access yet.

She is a clear example that intelligence and learning are not synonyms. And frankly ChatGPT has already overtaken her grasp of her own language.

Never. I've never even thought about it or looked at it - I don't even know how to use it! I have no intention of seeking it out, I'm quite happy without it.