Has anyone used AI to help choose outfits and personal style?

[MakemineaGandT]

I did this for the first time today and uploaded some photos of myself in different outfits. The responses were surprisingly insightful - and the outfits it picked as my ‘strongest’ are actually the ones I feel best in.

How have you used AI to help with style?

Well I wouldn’t have gone to an actual person previously for this. So my use of Ai has had literally zero impact on the local personal stylists.

But if some are being impacted then they’ll just have to do what millions of others whose jobs are impacted by AI are doing. Retrain. Find something else to do.

And the environmental impact of all your Vinted sales sent through the post?

Use of energy is everywhere. The answer is that it needs to be cleaner and regenerative. Which is not something an individual consumer can control.

if the same people complaining on here aren’t also lobbying the government, their MPs, their local
councils, their energy providers then it’s virtue signalling to help them justify their own ‘better’ use and demean others.

I criticise all sorts of things (inside my own head mostly, not publicly). Like people who drive journeys they could have walked, or people who fly multiple times a year, people who buy loads of plastic stuff they don’t need… I worry about all of it, so I’m not just singling out LLM.

Like I said, I just wish people would make more considered decisions re need vs want and consider the impacts of their purchases and behaviours; the AI threads I’ve seen in the last week on MN have shown me that a large proportion of people don’t do this. There was a thread the other day about people just having conversations with AI throughout the day. That seems like a disproportionate use of resources to me.

OP you’re using a hammer to crack a nut. I’m old enough to know what suits me and what I like. Every now and again I try something new. Sometimes it works, other times it doesn’t. I’d be surprised if in 5-10 years’ time I’ll be surrounded by people dressed with amazing style because they used AI. I’ve used AI to work out some pension stuff as I’m looking to retire early. For me, that was justified.

No, I mean singling out LLM usage specifically when it comes to data consumption. You don't NEED to be having this conversation on MN, consuming energy, yet here you are doing it.

Virtue signalling.

Ok @Oliveoy. Thank you for giving me something else to feel guilty about 😂 Perhaps it’s time to consign MN to the past for me. I genuinely didn’t realise it was so bad for the environment (not very good at science at all so I have to just rely on what I read/hear).

I'm not trying to justify the environmental impact of anything, or prove that using chatGPT is a good choice. All I'm doing is pointing out the hypocrisy and virtue signalling.

And I'd be willing to wager that your motivation for cancelling Netflix wasn't because you were concerned about the environmental impact of streaming video.

Re the environmental impact, it’s not great… BUT, the impact regarding water usage is actually much the same as scrolling on social media.
All those hopping on the “if you use AI you’re a bad person” trope should consider this… whilst making their points ON SOCIAL MEDIA.
No one is perfect so kids in glass houses, hey?

I don't think you should feel guilty, or that you posting on MN is terrible for the environment.

The point is that all data usage has an impact. A simple text based query to an LLM (as in asking it for style advice, not complex coding tasks) would use slightly more energy than a traditional Google search. Now that Google has its AI overview too, a single Google search is likely worse than the average LLM prompt.

I've used it, not so much for styling advice, but to use alongside Whering to show outfits I am wearing, to give recommendations on other items I already own to try with something I'm struggling with, to create holiday packing lists, to give constructive advice on what I'm looking at on Vinted- am I buying it because it's cute or is it filling an actual gap in my wardrobe?

I've also used it to give style prompts for a week to try and push me to wear different things from my wardrobe.

It's useful to give it pictures of what you're wearing,o utfits you like on the internet, stuff you don't like, and an idea what's currently in your wardrobe

Sure. Asking it one question that you can’t find an answer for anywhere else is one thing. But just having bantz with it all day long (as many posters seemed to be admitting to on a recent thread) just seems ridiculously wasteful to me.

I want to use Google without the AI summary and I think I read that there is a way to do that, I’ll look it up.

They should do one of those ‘save water ‘ ads with the kids telling adults how to save water (don’t use the hose or let taps drip etc etc) and have one where they are telling people not to use AI for such ridiculous things as this .

Its not that easy for some people. I have brown hair and brown eyes and look quite soft overall so I would often wear mauves and taupes as I thought they were blending harmoniously. But in truth I often felt a bit mousey and off in them.

When I eventually had money I thought buying nicer fabrics was the answer so spent ££ on nice merino from White Stuff and Fat Face. I couldn't work out I didn't feel right in a nice teal or duck egg cardigan when these are supposedly expensive looking colours. So I went down exploring tones etc. and still could not work it out.

Had my colours done at cost of £300 and turned out I am a paintbox spring - I look fabulous in bright turquoise, bright leaf green, bright blue, violet. I also look great in true poppy red, tangerine, all browns, except the lighter ones should be worn in lesser amounts. Absolutely hideous in cream...yet was always trying to buy cream knits and lace. My most amazing colours were bright navy and chocolate, which I did not have any of. I would never have worked all of that out. I looked like I should have been an autumn, now I am trying to wear lots of lovely neutrals instead with pops of bright.

😂 I’m going to say, she didn’t. But it’s going to be much less than possibly buying several outfits and returning most of them.

I have used ChatGPT to help me with outfits and it’s been so useful. It’s also told me my colour palette from photos. Something my colleague paid someone over £100 to do.

I am in my early 50s & have a v defined sense of my own style & i'm not particulaly interested in what other people think of how i dress, though it's nice to receive a compliment, i won't lie.

But i dress for myself. I have worked in a creative sector my whole life & come from a tailoring background so i've always had a deep interest in clothes & shoes etc

I spend a lot of time thinking about, researching & finding beautuful (to me) things to wear.

Earlier this year i was looking for something in particular & i randomly opened a chat with chatgpt. I was clear in my instructions & it came back with surprisingly useful info which allowed me to find what i needed.

The conversation expanded & was pretty insightful. I gave info on the brands & some specific items i wear & it deciphered very well what was connecting or driving my decisions - proportion & texture etc

Then it could suggest other european brands - lots i didn't know & that has expanded my pool significantly

I sell on vinted but don't buy a lot but if i see something that catches my eye i screen grab and ask it to give me the info - year issued, original price, how it was styled in its season look book, and twice it has checked if i am def considering it & has pointed out how it doesn't suit / fit with my other clothes

I have uploaded photos of me (minus face) so we can discuss combinations

Of course it has infinite patience to talk about minutae of style whereas i don't have that in real life.

I think its v helpful if you know what to ask but a random 'i'm 52 & don't know what to wear - dress me' is inevitably going to fail..

I don't feel guilty about using it - i don't drive & eat seasonally with as much local produce as i can find

Tbh, I do. It's a worry. But when I'm putting the recycling out in the rain, I think about that Kardashian flying in her private jet from somewhere in USA to Paris to buy her boyfriend some macarons, because he'd never had a macaron, aw. Haters gonna hate, for sure but if 30 mins with Claude enables us to be better consumers, it's probably OK in moderation.

Streaming one film on Netflix would be roughly equivalent in usage to sending several hundred prompts to an LLM, which would likely far exceed even somebody chatting crap all day. I use Netflix as an example because nobody ever says that's wasteful. And whether you think that's a more deserving use of energy is subjective really.

To get rid of the AI overview, add -AI at the end of your search.

It's useful to give it pictures of what you're wearing,o utfits you like on the internet, stuff you don't like, and an idea what's currently in your wardrobe

Apart from my test question of "what could I spend £1,000 on" , to which it responded with laughably hideous results, I asked it to put a skirt and cardigan I'd ordered from Palava together just to see what they looked like. And the same with a dress from Casey Casey and a Dries van Noten jacket. I didn't want advice. I had no doubts about them. I was just being impatient waiting for them to arrive.

The results were horrendously bad- dreadful, completely incredible (in the literal meaning of that word) AI "people".

The Palava outfit was turned into a simpering Disney character. The Casey Casey dress and Dries van Noten jacket were even worse. It reshaped both and managed to turn them into something from one a Mother of Bride site.

To get rid of the AI overview, add -AI at the end of your search.

Thank you. I really dislike the AI overview - not least because it's more often than not wrong when you drill down into its results.

Thank you, I will use that on Google.

I watch Netlix rarely- half a dozen times a year maybe. There’s not much on that interests me. But I don’t think it’s an either/or thing. Lots of people will be watching loads of Netflix and YouTube and posting on MN frequently AND bantering with AI on a frequent basis. I just think people should think about it a bit more and be willing to try and moderate some of it, but I do realise that isn’t going to happen.

This is ridiculous. Using generative AI consumes vastly greater resources than a web forum like Mumsnet. They're hardly even comparable.

That's a completely nonsensical statement. Determining which one would consume greater resources would depend upon your usage of each. Obviously if you are on MN for 5 minutes but are spending 3 hours getting Claude to code for you, the AI usage would use more resources. However if you are looking at something like 5 text based prompts to an LLM vs an hour scrolling MN, the scrolling would use far more.

And on a worldwide scale, general web activity uses far more resources than AI does.

How is it nonsensical? If one compares equal time spent on each (which would be the only fair comparison), then generative AI consumes magnitudes more energy than a static web page like Mumsnet.

In fact, the need for generative AI data centres is a massive shift and increase in the impact of server farms – AI requires much more computational power than general internet usage.

"While traditional data centers are built primarily to support CPU-centric workloads with standard requirements for security, reliability, and energy efficiency, generative AI demands a fundamentally different infrastructure approach. AI data centers require high-density GPU clusters and advanced networking, storage, power, and cooling capabilities to handle the extraordinary computational intensity of modern AI workloads."

Also, as this article points out, "Unlike when you stream video or load a large web page, with generative AI most energy is used at their end, while processing your query."

Obviously, video streaming uses energy too, but not with the same intensified data centres that have much more (an ever increasing) impact on the environment and consumers. Besides, you were talking about Mumsnet usage – that's a drop in the bucket compared to the impact of using generative AI on a daily basis.

This Wired article is interesting too, regarding the impacts of data centres especially.

Equal time spent on each wouldn't deliver a fair comparison. It depends upon WHAT you're doing on each. If on AI you're coding or reasoning, that's going to use far more resources than just chatting. And if you're chatting, how much time of that are you spending on typing in your own prompts, which isn't using resources. On MN, what are you doing in that same time? Very slowly reading one long thread, or jumping around between pages? How much data is being transferred in terms of new pages, adverts, analytics etc?

I think you honestly might be surprised to see how MN usage could easily outstrip that of using AI for the same time period.

No. This is just plain wrong. Just making text queries and 'chatting' is still going to have more impact than scrolling on forums and static web pages (especially for people who use ad-blockers), particularly in regards to data centre impacts.

I suggest you read up on the issues, as you don't seem very well-informed on the massive energy suck of AI, how much worse it's going to get as AI gets more advanced and more widely used, and the effects of the far more resource intense data centres.