AccessGiven – 10th of September 2025.

Opening remarks

– Hello. Hello.
Hello everybody, thank you for coming.
– Very quickly, a visual description of myself, Ellen Kate Boyle and Michelle O’Connor :
( description)
– yes, apologies, I am already emotional.
– Thank you so much, everybody, this is a very special day, and sharing this: when we have been going to another event that was not entirely accessible, we wanted to do this in a different way, so it is really special, and hopefully you can see all the things we have been doing to make it accessible. If you feel something that could be improved: TELL US, because the appetite is so big that we would like to do it in the best way possible.
– For the past week, welcoming you all here, we are just 2 normal persons, and as Ellen Kate Boyle is saying, we are going for it, and now we are here. We appreciate everyone of you being here and trusting us with your precious time but also wanting to thank our sponsor! ( thanking the sponsors)
We would also like to thank Veerle and Be Live for the live captions ( captions not shown at the moment) with a human made captions. You have also a QR code in your pack and you can follow with your phone.
Now, I think… this is it from us.
The speakers will take the stage: Helen Dutson and Holly Tuke about making accessibility creative.
– And they are happy to introduce themselves.
– Welcoming them to the stage. Thank you very much.
– ( You can stand here at the microphone)

Helen Dutson and Holly Tuke 

– hello everyone. Can everyone hear me? And the captions are here. Wonderful. Our talk today is about Making creative accessible and accessibility, creative. what does that mean? When I am talking about creative accessible, as in assets, campaigns, pieces of contact, accessibility in a more accessible way,…
– did you prepare that?
– Did you not hear me?
– Okay.
– Hello.
– I will do it over here.
– This is better.
– Do we need to swap?
– It is considering, creativity accessible and accessibility creative is ensuring everybody to be included, but also thinking in a different way and not just a box to tick, actually something that can enhance the creative process and approach creatively. So. I will now introduce: Hello there!
Holly Tuke (she/her) is an award winning blogger at Life of a Blind Girl – a website – and Social Media Manager at RNIB.
[Holly to audio describe herself]
– I am a white woman, shoulder length brown hair, black top, trousers in black.
– Now the description for me: Helen Dutson (she/her) is a seasoned agency copywriter and and expert in ALT Text and accessibility expert, creative turned accessible content trainer at The Accessibility Briefing, and we had some really great time together. She is an accessibility champion.
– I am white woman, wearing black trousers, and a blue top, green jacket on top, and I have my pride badge on my jacket.
– Here’s the problem.
89% of disabled people have faced challenges when it comes to digital accessibility, making them feel frustrated (71%), disappointed (58%) and angry (24%).
I don’t know about you, working in marketing related fields or communications: when feeling frustrated and angry, Holly Tuke is there anything you would like to add?
– I would say, on a daily basis, as I am blind, I can relate to this everyday, even if it is not intentional, society is not made for people who cannot see, it is not made for us. The way in which it resonates with me: often sometimes as navigating a website, or trying to find concert tickets, doing something nice with a family member but the website not being accessible, stressful, needing to find an alternative. While you can just use the mouse and click. 20 minutes of my life I will not get back. So, when there is an image on social media, that I would like to interact with, or trends I would like to share with my team: doing social media channels but not knowing what is actually happening because there is no ALT Txt, and on the flipside, when accessibility is done in the right way, it is also bringing joy. And I am sure your attendance here means that you are committed to accessibility: it means more than you know, on a daily basis. You are having an impact. You are helping to change the narrative accessibility.
– Absolutely, Holly Tuke how many times did you have to follow-up for Taylor Swift tickets ?
– you got to watch it?
– I did.
– I thought I was going to get them. But it is just showing that when you are trying to buy accessible tickets, the infrastructure seems great but it is not accessible. We need to change that. It is to highlight that so many things need accessibility. We can talk about this the entire day.
– Absolutely. As Holly Tuke related to, when thinking: Accessibility is a love note. with a black heart. Would you like to explain this quote?
– I was writing a post on what accessibility means for persons with disabilities, and my experience, is not reflecting the entire community, the way society can be better, and how to be allies, because it is definitely not fitting everyone. It is the joy of disability and accessibility. It is coming out of this, and highlights what accessibility is about, and it is telling people with disabilities: we would like you here, you are welcome. It is embedded. Showing the entire team and the community that we are supporting it, and that everyone is welcome.
– Absolutely. What I would like to ask you now, it is not a Q&A session, but it is something to decide for yourself: no wrong answer. Who is accessibility for? taking a note. if you are not: If you’re not making your content accessible for disabled people, you’re missing out on a huge and influential audience.

17% the percentage of the audience brands could be missing out on because they’re not considering disabled people. I am coming from the agency world, working with big brands and big campaigns, and what I see, is that we would spend months trying to increase time by one second and increase engagement by 1% but you could miss 17% of the entire population. Think about that. You might notice already. £274 billion the spending power of disabled people and their households in the UK, This is something. Yes. This is: “the purple pound”. showing the influential impact. Holly Tuke Is there something you would like to add?
– When we think about online shopping, or physical ramps in the store, enabling people with disabilities to access services, and content sharing by making accessibility a priority and embedding accessibility in everything you do, and the power of our community.
– And it is the reciprocal relationship.
I would like to reflect on :But is accessibility only for disabled people?

“There’s this myth that nondisabled people don’t need help, but nondisabled people need lots of accommodations: there is the quote, with a video from Haben, and it stuck with her. She is talking about her time at Harvard.
The cafeteria was full of chairs, hundreds of chairs. Those are accommodations for nondisabled students. Students with wheelchairs roll in with their own chairs. So the school was spending money on accommodations for nondisabled students.
And along the ceiling, there were lights. Those are accommodations for sighted students. Blind students didn’t need the lights. In fact, the school would have saved so much money if they turned off the lights. The difference between accommodations for disabled people and nondisabled people is ableism.” (Haben Girma, deafblind disability advocate and Harvard Law graduate) Thinking about a disability in a different way, it is not just about accessibility, all around, and only talking about it for people with disability because society is neglecting that. If the lights would be turned off, you would shout, because it is the accommodation that is automatically provided for the majority, as for persons with disability, it is a special need.
– It is about the things that we are not noticing, the light that is turned on. When we think about those accommodations, it is not extra, it is just putting it into everything we are doing. lights are essential. No matter what we do. It needs to be accessible.
– If there would be no chairs in the restaurant, you would notice that, and this is accommodation you have been given in the restaurant. To build on this, it is not only for the disabled Committee: 89% of people who use subtitles aren’t Deaf or don’t have any hearing loss.

And who can forget that audiobooks were originally invented for blind and partially sighted people?
I mean, almost everybody loves and enjoys audiobooks and interact with literature no matter what you are doing. It is benefiting everybody.
– Yes. There is also the back story for my audiobook journey.
– What is that?
– When I was at school, I will not talk about my educational trauma, let’s not do that, but I was told and drilled in to me, that I should not have audiobooks for learning but I should use Braille for learning, and professionals telling me that I would not learn by listening even if I would find it more appropriate to have the choice. But they told me: No. At home I could listen to the audiobook and get lost in the story. There is the misconception that audiobooks are not enough. And they are. It is valid for of Reading.
– I love that insight.
Now, I would like to make you think about those statements:
Have you ever watched content with the sound off because you’re on the train, in bed or want some peace?
(you need subtitles!)
Listened to content instead of watching, eg if you’re cooking or doing the night feed, walking in the park, listening to the audiobook, … and I hear a lot of people loving to audiobooks while feeding their children: ?
(you need audio-led content!)
if it is only music, you will not know what is going on in the video, or: have you ever, Misplaced your glasses and can’t read that tiny font?
(you need larger text and an image description!)
you can make the font bigger, and change the settings. And anyone out there: Don’t fancy watching a video right now and would rather read at your own pace? I see some reactions. I have been working in social media for a decade, and it is always: video, video as a prior decision, but personally in my personal time, if I go to the recipe, and it is having a video, I will go and find another recipe because I would like to print and navigate at my own pace.
(you need a descriptive transcript!) It will take the verbal part for somebody who is talking through with with visual cues. And so, what I would like to get through: When we think about accessibility as something for everyone, we run out of excuses for not making our content accessible. and this might seem like – you know – thinking about one group or the other way but everybody benefits. And also contents performance will benefit.
There is statistics supporting this:
12% increase in watch time on Facebook for videos with subtitles. that is a big increase. And another study in advertising: 16% higher engagement rate when ads had high quality audio. so, it is really about prioritising all of those experiences. So, this was the interlude between asking: Who is accessibility for? now,Hands up and say ‘woop’ if your answer has changed! thank you. Some WOOPS there.
Is there something you would like to add Holly Tuke ?
right.
Our top ten accessible content tips.
Some glossary, I feel like, if you would like to learn about accessibility but you don’t understand the keywords, you might be lost: Screen reader – if you would like a demonstration Holly Tuke can get her voice-over reading messages –
Software on a phone, laptop or other device which reads the contents of the screen aloud. It cannot interpret images or buttons, unless descriptions have been added in the right place. Leading us to: Image description
A description of an image, written as text, meaning all users can access and read it. This could be hosted on a webpage or in the body of social media copy. The length of these descriptions can vary from a few sentences to thousands of words, the national gallery is doing wonderful descriptions, describing paintings, for example, and : Alt text
An image description, but hosted specifically in a dedicated alt text field, for example on a website or on social media. it is valuable to learn about this, because Alt Text, like Twitter (X) you can click on it and read the text, but Instagram and websites, you have to navigate into the quote to find the Alt Text so it is nice to include the image discretion and Alt Text so everyone can read it. Also in LinkedIn. Because people are curious to know what you have written.
– If you can have it on that, you can also create awareness, by raising it. And setting an example and doing it in your own way.
– Absolutely. Now, Be very careful with colour contrast! LinkedIn post from Kaye Moors, CPACC. Post reads: “Today’s low contrast issue is another example of a brand paying for sponsored posts, only for them to become invisible to a portion of society and to pretty many us all in bright sunshine. To stop users scrolling, your social ads must stand out, not blend in.”

The image is a screenshot of a sponsored Facebook post from Maria Nila – not wanting to throw a lot of shade on them but it is an example chosen – which features a pastel blue bottle of Curl Definer. Not only is the writing on the pale blue bottle white, making it unreadable, it is placed on a similarly coloured background with white text, which reads: “Adds light hold”, “Curl defining” and “moisture regulating.” This is quite difficult to read, too. the colour code, if you are having the visual impairment, or brain surgery survivor, all of this would be very difficult, it would be a struggle. The main point: none of us could see or read it, so what would be the point.
– It is really insightful.
– There is some very big names in there indeed. Now, the second: from someone coming from the copyright background. Remember that writing alt text can be a rewarding and creative exercise.

This is an example, when working at Institute for blind people, if you remember, Calvin Klein underwear people with the photo shoot of Jeremy Allen White and the entire Internet was talking about this, and noticing that none of the images had descriptions, not even the original post from Calvin Klein, so it was great Internet moment, but if you put yourself in Holly Tuke shoes, someone who is blind, or partially sighted, to be excited about something happening but not knowing what the images are showing. So we had a tweek: Do describe. The first image, with the text we have written, A film photo of Jeremy Allen White standing in front of a red sofa on a rooftop in New York City, wearing only white Calvin Klein briefs. His curly hair and lean, toned physique is silhouetted against the pale evening sky. He stands with his arms by his side, smiling off-camera.

and the second: A film photo of Jeremy Allen White standing behind scaffolding on a rooftop in New York city, wearing only white Calvin Klein boxers. His curly hair, blowing in the wind, and his toned, lean physique are lit up by the sun of golden hour. He has one hand on a rung of scaffolding at shoulder height and the other on a rung above his head, as if he’s about to climb. There’s a look of intense concentration on his face.

you can tell we had a lot of fun with that.
The effect, that this became a new story, it was so interesting: it was covered by different media. And it is maybe the first time those two publications have ever agreed on anything. And you had those headlines about the prompts and the Alt Text. Yes, anything to add? Holly Tuke ?
– there was a lot of people sharing and back in the Twitter days talking about image, and using different types of assistive technology, using this with other accessibility features, and to have some form of AI created text that was and not the same as if human would give the impression. ALT text, With images, it is taking the time to add this and this is important because otherwise people will not be included.
– And if you are a creative person, and you love writing, it is to use your device and look at the image, and even if it is not accessibility feature for you, it doesn’t mean you can’t try, wanting to know what the image is showing and the tone of the image and why it is important. Although stings, and blind people need to contribute also and take part in culture. We go to the next point:

. Avoid one-word-at-a-time captions. They’re really popular right now, but so hard to follow! you will see them everywhere. At tikTok. Some people like it, when you are hearing, it makes your brain blowing, but if you cannot hear, or if you listen without the sound, it is going so quick and it is hard to follow. So have at least five words on screen at the time.

And this is a big one: 4. Make sure the message of your video comes across audibly, as well as visually.

the way you can decide if it is working, listen to your video: like it’s a podcast, without looking. only interact with that piece of audio content – Do you still understand what’s going on? do you still understand the key messages? Did you enjoy it – Was it still funny? And interesting? Did you have a similar experience? just make sure that it is also engaging, not only visually.

– Yes, the amount of content on social media, with images and video, music, when you are scrolling, and reels and videos, description and audio description is limited. When you are using social media in a personal capacity it is not really fun because you don’t really know what the video is about.
– Absolutely. And if you’re working marketing agency, spending so much time on content and spending energy, and imagine publishing – but I don’t know what it is and scrolling past it – because there is no accessibility: it is in everyone’s interest to ensure that. What you can do, when you listen to your video, and you get everything you need, and some context is missing, you can add video description in text, and that could be on the webpage, and you can just format it.
. Add a video description to fill in any *small* bits of missing context.

 

[Video description: Tumi, a black man wearing glasses and a green striped t-shirt, is speaking to the camera.]

All that was left to was to describe the person speaking.

This is building on what was said before by Holly Tuke the content of Internet, just a meme with music playing, but a screen reader cannot access the text embedded on the video, so make sure that ALL text is voiced or put in the description.

There was the meme going on, in vineyards – this is my idea of a good day out – and using that clip with a screenshot in YouTube: [Image description:
A screenshot of a YouTube Shorts video, with text reading:
“POV: you hit 30 years old and are stood in Aldi’s middle aisle looking at portable toilets, inflatable camping chairs and 3 in 1 egg poachers.” The visual is a cut out of Gary Barlow saying: “This is my idea of a very nice day out” overlaid onto a photo of Aldi’s famous middle aisle.]

it is funny because you can read it.
And if you imagine someone on tikTok scrolling through, this is my ID, this is my idea, this is my idea, you would wonder what is happening in fact.
– Yes, you would like to be included in the content. And by taking some more minutes to voice visuals, it can be easily done, such as tikTok is having the option embedded in the application. An some people use it for non- accessibility purposes, yet, you can use them. With very quick voice-over. Describing your visuals.

– Absolutely. In this case, you could have the POV in the caption if you could not voice it for any reason.

Next:

7. 
   Add a descriptive transcript to your videos, which can be navigated by people using braille and those who prefer written content over videos.

also people who prefer reading written content, there is the nice example: Haben, a Black woman in her thirties with long dark hair, speaks to the camera, a vibrant blue wall behind her. this is the transcript of video about transcript. I will read it.

 

Haben: If you’re a creator, add transcripts to your videos. I can’t see videos, I can’t hear videos, but I can read transcripts.

Pins on a Braillenote Touch pop up and down in their Braille cells. Each cell has eight pins that are either positioned up or down depending on the specific Braille letter.

Haben: Braille displays connect to phones and laptops, allowing Braille readers to access the internet this way. Descriptive transcripts should have both the visuals of what’s happening on screen and speech and key sounds for the video. Really good descriptive transcripts captivate readers just like the best novels.

The Braille display disappears and the video shows Haben in the same room.

Haben: Once we have widespread accessibility, it’ll be easier for deafblind people to share our stories and also participate in conversations. I love learning from lives different from mine and in order for me to do that, I need transcripts. I look forward to reading all your transcripts!

You can get a great idea, about the flow, visuals, and the only difference is maybe that she is American, but you get the sense of what she is saying, and as a copywriter or creative writer and novelist the line: descriptive transcript really speaks to me because it is showing the creativity that you will have, and effort of transcript if power and so if you are the writer, you have to opportunity to reflect. Also in the social media world, and something to add? Holly Tuke
– You are giving people the choice. And it might take a lot of time. There is different way to get the transcript from your video. You give people the choice. And any other piece of context. It is not linear, it is giving the choice. Some might even want to use both options.

– Yes, and an easy way to do this, in a creative field, when you are proposing the video to your boss, you cannot just make the video from scratch but you would need to write something, and you can save that, and you can put it back to the video as a description, all the spoken elements, and use that.

– 8. Maybe a little bit controversial. If you have some content, and you would need the second track for audio description – what would be easier – to make sure that it is done from the beginning because it is quicker and cheaper, and working for example with a big retailer, for example for Christmas, waiting for this moment every year, making sure that they publish all the content but also looking at the speech in the advertisement, ahhhhaaaahhh it would have been easier to have it published straightaway and it was wonderful to have it with the audio description and it is adding some work for yourself that you didn’t need to do.

– Making sure that it is sync. There is so many big companies where it is not in sync with the visuals, and this is an interesting experience. But if you have this from the beginning, you are doing the right thing. And if you get to the point when you follow the trend, and it is not the most accessible, take it as a learning, take it as something that you can take forward in the future and use it as a learning, and learning from mistakes was the two grow, and have more access.

– I would say, there is also the of audio description, for example Bridgerton on making out with some dude, and having the tone in the correct way, looking at the content when it would be difficult to make the entire content audio led, but it easier to do it from the beginning instead of doing it afterwards.

– You can also turn it on, turn it off again, and also understand the description and asking yourself: what it is adding. I recommend doing it. It is giving a lot of insight.

– 100 percent, the Grinch is having it in an amazing way, just an example.

9. Only use one or two emojis in your writing and never put them in between words.

A screen reader would read this as follows:

The way the screen reader would do this with the robotic voice: This clapping hands
Is clapping hands
What clapping hands
Screen clapping hands
Reader clapping hands
Users clapping hands
Hear clapping hands
When clapping hands
You clapping hands
Over-use clapping hands
Emojis clapping hands.

this will make you throw your phone into the sink!
and it is completely incompressible.
You need to think about description of the emoji.

– I can just get some triangle or a blackbox or something so I have no idea what the emoji is about. So really check if your emoji is fitting emotion you are trying to bring over.

– The cry laughing emoji – tears of joy.

the last one:

10.Make your personal content accessible too. Holly Tuke- everybody would like to know – what you had for brunch or how cute your dog looks in its pyjamas – even if it is unimportant, but it is about practice, and it is about creating accessibility.

– It is what people are saying about personal content, and how do you know how many people with a certain disability are following you: so also on social media, it is not for everyone. You need to recognise that. And also to disclose books for example on social media. So it is looking at the accessible content, and using this as an example. It will enable you to look at accessibility in a different way and the more you do it, the more you will spot it out in the wild, and you will notice it, because it is already embedded in everything you are doing. So continue doing it whenever you can, because we would like also this experience with your dog, and your brunch.

– Absolutely.
Making it stick.
We can talk about it today.
We can be committed over here.
But what about your workspace?
Reading your email?
What about the actions?

Add an ‘Accessibility requirements’ section to your briefs.
When you are creating briefs, have a certain section about accessibility requirements to have subtitles required, audio Led, images description…

you can: Add an ‘alt text’ column to your content planner, for any project.

Start asking “Is this accessible?” during ideation, creation and approval, The entire process. And I don’t need to know the answer. But you just need to ask the question and start the discussion with your team and do research. And understand what’s the vibe is. Asking questions is so valuable.

Consider how to accommodate accessibility in your workflows, Because I will not stand here and lie that it is not taking longer and it is not an extra for your workload, it is, but you can accommodate, and if you want like to make your content simpler, with the text being accessible by eye, ears by screen reader and Braille, and so on, you can magnify the size, and you can zoom in. And the text is accessible. You would really go into the topic and give a lot of information and valuable insight.

If you would like to produce fewer pieces of content, and if it is all required, and consolidating and be selective about trends you embrace: there is a lot of power in this, coming from the social media background, jumping on everything. You can tell when someone is jumping on everything and wanting to make something stick. But there is also power in: this is not for us.

Would it make time to make it accessible ?
Would it be worth it?

Just, some final thoughts:

It’s okay if you make mistakes.

It’s okay if things take longer.

You’re learning and these things take time.

Remember: progress, not perfection. Holly Tuke?

– it is okay that things take longer. And that is okay. Professionals in society: continue doing it, because it is being done for a reason. And it is progress, not perfection. And you have to be doing it, not being perfect, but making progress and this is leading to the end process. And rather than panicking, working with something that is not accessible, do your research, and adding that extra layer, in making progress, as we are all on the accessibility journey. We have been speaking about accessible content in this presentation, and while focusing on content, it applies to everything you are doing, as when you are posting on Instagram, but it is also what you are doing in your daily work: recruitment process, PowerPoint, emails, Team chats, WhatsApp chat, it is about including people in everything you do, and making that progress. Thank you.

– People were supposed to clap at that moment.

Note down the next step you’re going to make on your accessibility journey.

For example, asking questions, progress is important, not perfection. Alt TEXT To the next post you are posting. Just note this down.

What will you do.

– And if you need the reminder: as in our busy lives – but the reminder on your screen, on your phone, a piece of paper, whenever you are working on image, reminding you to add ALT TEXT.

We will go to the questions in a second.
Do we have time?
First: A great big thank you from us!

www.theaccessibilitybriefing.com
www.lifeofablindgirl.com

Holly Tuke : if you would like to reach me, you can find me on www.lifeofablindgirl.com with some nice things about my life, and social media as well… talking about accessibility and advocacy.

– Holly Tuke is brilliant to follow on social media. Do we have time for questions?

Ellen Kate Boyle is nodding!
But we can not hear that…

(microphone testing)

— if you shout your question I can repeat it —

( microphone not audible… Ooohhhhh)
— wondering specifically: slowing down… fair enough noted.

When we are thinking about how we are missing out on audience, and statistics, when we are presenting this as business case, when you are working on a website for example, and you would like to give them a figure on how much they would be missing out, if there is any data: you are talking about the engaging rate, and access to data, and the monetary value, or the money they would be losing.

– For the purpose of the captions I will make a summary:
we are talking about the figures and spending and how this can be put forward as a business case, and a few examples on top of my head: the resources that have some great figures about the power of disabled people and how much money you might be missing, ability.net with a lot of research and reports, and purple count and purple Tuesday is initiative that is having good interesting stuff and a lot of research online on people with disabilities and organisations have done a lot of research so definitely, when you are putting this in your business case, and doing this in the disability organisation, doing the things that are relevant for you.

– Thank you.

– Lots of questions.
Holly Tuke And Helen Dutson will be around for lunch also, you can also approach us.

– Have you noticed, with more people using ChatGPT and co-pilot, and also the emojis in the social post, how is that evolving?

– A very short answer to that: yes, one thing that AI is doing a lot in post and content, putting —- and emojis, and this is the divided topic for many, people believing that it is easier to see and it is breaking up the text and alternatively for screen reader users, reading out the description of every single emoji and that can distract away from the content, and also for people with dyslexia, neurodiverse persons, it can be distracting, so, use dashes so you are looking into the best accessibility features, and it can also put emojis in between the words but the best is to have it at the end of the text is a good practice. Doing one human check, also in LinkedIn it is becoming more popular. There is a big increase.

– We have time for one more question.

– I saw you having your hand up earlier.

– Hello, we get a lot of video content from different services of the company, and what about the transcript of that, any help would be very appreciated.

– So, if you are using Microsoft, for example, you can actually call transcript from SharePoint and download it in Word document format and go and take time stamps and at the document, it will not be perfect, there will be some mistakes, but you can also be editing this instead of writing everything yourself.

If you have some software, you can also often graps transcript from different providers, and it can save a lot of time. You can upload the video to YouTube and download the transcript that YouTube generates, this is free. There is also software: there is different once, and some you have to pay. And it will absolutely help.

Software: TINT (?) TRINT Yes! (sorry)
🙂
our captioner is human – this is one.
I’m sorry, for all the words…
a little heart. Right.

– Now, we will have a small confort break. Before the next speaker – thank you so much.
Thank you everybody.

( A small break)

Craig Abbott

Sorry for the delay, it was the Wi-Fi issue, so we had lost connection with our captioner, but now I would like to introduce Craig Abbott with a brand-new talk, some of you may know Craig Abbott already, I will let Craig Abbott go on with it: thank you for your patience.
– Can you hear me okay?
I will talk about LLM and accessibility, I’m Craig, I have ADHD, and I am also autistic, we will get through it together. And working in disability in the last years, and working as a specialist and accessibility consultants, before we start: Before we start. These slides are a bit wordy.
Because they involve responses from LLMs, and I will read through all of it, and some times reframing it, and as soon as it come up, I will read through it so you will not miss anything.

AI is everywhere.
But do we really need it?
Every product is just adding AI into their products.
So many products with a ChatBot doing things you don’t want them to do.
I don’t know if you have done this: Deque Vision.
100% accessibility development testing and fixing can be done with zero specialized accessibility knowledge with Deque Solutions. this was their vision.

And they have put a lot of time and effort in doing AI behind the scenes and releasing this: At AxeCon this year, in Deque’s opening keynote, they spoke of their vision to fully automate accessibility testing
Not long after, they released their AI offering, axe assistant, which I’ll talk about a bit more in a bit.

How do Large Language Models work? LLMs. that will help underpinning the problems.

Basically, LLM is giving a lot of information and data, and giving all the information and they can learn from it, and LLM is giving all the content in tokens. The token can be half a word, a piece, it is the algorithm, and the more complex words tend to be more tokens
LLMs pricing models are usually based on number of tokens and does the more expensive it is becoming. This is the thing a lot of people don’t realise: LLMs are just massive probability machines, Looking at the given current context, and what is the highest probability of the world that will follow: an elephant grasps tree branches using its …. based on the model, it will be the different options. Trunk. Feet. Hans. None of the content has been trained on elephants and hands.

If we have another example: people are often looking for …
that might be a lot of things, love, work, or… an elephant. But that is less probable.

The same is with the code.
Accessibility…. What would this be? This is based on the HML training. And how in the background this surprisal score is also how they predict code, having this information which is inherently biased, based on human data, and it might sounds nice, and this is how LLM are being trained and how it comes back.

Now, some problems with LLM.
Some real issues: hallucinations. And basically just means they tell lies, but lying implies it’s a manipulation. An LLM can’t manipulate you, because it’s not conscious: So, they call it hallucinating instead. they do that all the time. Telling things that are not correct. And they will be very confident with the answer. Sometimes they will quote somebody: and just discussing earlier, things that are not existing even. And they are biased. And being trained on billions of sentences written by humans: We are historically and systemically biased in our nature, time has changed, and reading biblical scripts, not imagining they had accessible inclusion, and all the bias, ableist language: Because LLM’s respond based on the most likely response, it makes sense it will inherit our bias.

the main thing to take away:
t depending on what you give them, if you give them more better things, more better will come out.
There is a relationship
between tokens: They don’t know that Paris is the capital of France: it is not having an idea. But every time they see ‘ capital of France’ Paris will follow.
They are not consistent.
They cannot give the same answer twice. They will just generate based on probability with some randomisation throw in. Their goal is to form coherent responses, not factually accurate ones. it will be random. Roughly it will be the same, but not exactly the same.

If data is bad: the response of the LLM will be bad, at this crucial if you are asking questions about accessibility, and accessibility standards and guidelines

now: 94.8% of the top1 million websites have detectable WCAG 2.2 failures.
(WebAim)

This figure is from the latest WebAim survey
When they say “detectable failures” they use a tool called Wave, so they’re not even auditing these sites
We know automated tools only find around 30% of known accessibility issues
So I wouldn’t be surprised if this figure was actually as high as 99%

this is not great.
We are not doing well in making things accessible.

When you are giving billions of lines of code, and Words on disability and inclusion, 94.8% of Internet is trash and this also LLM, so there is no way ever to take over an information. LLM and accessibility. Looking in the rabbit hole. I love looking at it. And I like to do it, spending a lot of time digging into it, and if you ask and never use ChatGPT, you would ask general things early on and it is giving you a very good responses but when you do something with nuance, it is falling apart. So I really like to look into those examples with more nuance. And how it is responding.

Caveat.
Using AI for a lot of things, my bias, when I try to build the experience, and trying to see if I can break it or if I can get good response. And we have good examples too. Most of the time they are not great. There are the things that I have been doing in the last month. In the LLM available at that time. Even if they are improving all the time and have new information. All of the information that the LLM note will go to a certain point. And this is what Caveat is. I will show you. I am popping those in. Hopefully, with time, it will get better. One of the first things, have to LLM text descriptions for images: this is Important.

LLM: trained to be coherent and how it can be even better, I have one stock image, one transgender person, but the description is not saying it: Alamy.
It’s a transgender person, but the description does not give you their pronouns. It’s really important not to misgender somebody. So when writing alt text describing a person: If you’re not sure, then you shouldn’t just guess. it is making the description as a woman, she, her… is a focus. And using the wrong pronouns and we are not right.

Now, Meta:
having this as a description – with hallucinations – because the image we have seen is one single person standing – while this description is talking about 2 persons. These are hallucinations. This is not there at all.

Now giving Meta more context: and they are still referring to 2 persons.
Yes, LLAMA, and the other examples, these are LLM is that you can do locally as ChatGPT and use them on your own computer, and you can also write a script with thousands of images and have captions for all the images. But you don’t want to put them in the wild. You need human AI collaboration – because if you put this out – you will have all the hallucinations and that will be embarrassing.

ChatGPT did better, it didn’t misgender the person and used the word person in stead.
Which is nice. It did a pretty good job here I think.
ChatGPT Was good.

Another tricky image I give them was this one I took in Las Vegas – I had to go there for work and wondering around and having the replica of the statue of liberty – and the reason why I gave this to the LLM,to see if an LLM would understand it was a replica? Google Gemini failed.

this would be a pretty good example, but in Vegas is half the size, so not having the idea that it is the wrong place, the wrong thing and this was the travel website, this is just a lie.

ChatGPT did recognise it was a replica: On a stone, blue sky, the bird, rollercoaster, it is doing all right, recognising as a replica, even if it is not rollercoaster and no bird, is this the hallucination, maybe, it is looking like a bird, the flame, it was nearly getting it but not quite.

Anthropic Claude did rather well on this one.

It is really showing the differences. Google is having access to all of the images on the Internet, and it can look Up things and it can cheat, it can compare with similar images. And build only use the number of tokens, the minimum amount of tokens, unless you provide more, because it is expensive. The more you go for accuracy, the more expensive it becomes.

The running joke: that CEO OpenAi, it cost millions to ask questions with please and thank you because it needs to be processed. And it is not needed. This is costing money. People have access to the image, and finding out context: it will always go for the cheapest option unless you ask for it.

The next, is on the LLM advice WCAG criteria.
Measuring how accessible one website is.
The standard is complex.
It is looking at different criteria, and it is difficult to understand, and the target itself is not always accessible.

This is, where I find, where I thought it would be interesting, for LLM, to help people to understand it. I thought this was good use case. One of the things that makes it difficult to understand, you can have the same problem presenting itself in the same way but depending on the LLM and depending on the criteria, it would fail. On the surface it looks like the same problem, but it is not. For some reason, it is linked. In this example, I have been writing the link.

You don’t really need to understand it.

There is the link.

Basically, it is an image, not having ALT attribute, no text description and no accessible content, and the bottom is using SVG, again with no accessible name: and if you would do this with the screen reader, you would have no idea what it is.

You can see the icon, and it is problem for screen readers to understand those elements.

This is an example of why WCAG is annoying: The link fails 1.1.1 Non-text Content, because the image has no alternative;
It also fails 2.4.4 Link purpose in context and 2.4.9 link purpose link only; It needs to be descriptive by the text alone,
But the button fails 4.1.2 Name, Role, Value, despite having theoretically the same markup.

It is not having attached for screen readers for example.

Here the exact same problem but different criteria.
Further LLM, to identify the problem, I used the word “exactly” because when we work to a standard, accuracy is important. Deque’s Axe Assistant: and it was failing, and it got one part of it correctly.

OpenAI Was telling me the same:
and it also was saying that the link was failing, and this is not necessarily incorrect, it is basically the interactive element that is not having the accessible name but with the separate criteria, you would have adding noise to the report. So it is not necessarily incorrect, but some had been included and others haven’t. Again, 50% was registered

The best LLM’s I tested still only got about 50% correct
And 50% is not good enough for an audit
Most automated tools struggle with link purpose in context
But LLMs are supposed to be good at context, so I expected them to do better

it is not doing that well in this kind of problems.
To have the correct criteria.
And if the Alt Text is there, and to look at the issues that people are raising, nine times out of 10 the interpretation would be different, so it is inevitable that the LLM cannot do a good job. If the point of LLMs is to understand context and be coherent…

Looking at the training material:
when you ask things, the knowledge and the content updates needs to be looked into.

Now, can LLM restructure content to fix accessibility issues?
what does it know based on the context and content, taking one document: the accessibility strategy in the GOVUK design system – It has a lot of headings, 45 in total, from H1 to H4, I’ll not show you them all. this is just the first part.
H1 – H4.

The LLMs did have access to the whole article
To make it simpler to show on a slide we’ll just focus on the first 10
They go from H1 to H4 and, it’s pretty logical, we could probably all guess the heading levels I think
The thing with headings, is if you mess one up, it can have a knock on effect to the rest of the page

 

I jumbled up all the heading levels
I gave them one for free, I don’t know if I meant to, or if I just missed it!
I then asked the LLM to check the heading levels and recommend fixes:
Can you tell me, from accessibility point of view if the content is accessible. Or, what is wrong.

We have the responses.
This one is Chat GPT – I have it on top of the slide – Chat GPT managed to get 3 correct, if we’re being generous. Technically one of them was already correct. H3 and onwards is wrong. What should be H4 is marked as H3, so it cannot recover from the content. And grouping all the information. No way to recover. Because of the previous context. Only having the first 3 right.

Google Gemini did a little better, it got the first 5 correct. But again, once it got one wrong, it just snowballed and it couldn’t recover. If you want this to restructure your content, it will take you a lot of time. It is not great for the return on investment. And how to fix it.

Now, ChatGPT and Google having access to every single index. Looking up: it should know the content, and asking to do that. Next, I was trying to work out, then LLM find readability issues: talking about accessibility and talking about reading scores and how to formulate sentences, and one of the things we are seeing: it is not getting audited. People are writing in block capitals. What do they do with block capitals?
Block capitals are notoriously difficult to read, especially for people with Dyslexia. Science suggests we read by recognising word shapes, not letters.
But with block capitals all of the words are the same boxy shape: It is not getting picked up. And what I wanted to do, is LLM able to notice this? Are they normalising the data? You might take all the capitals and turn it into lower cases while processing it. I had the feeling that ChatGPT and will would do something as normalisation, but maybe it was bias, and this text as well is reading: EAA as a piece of legislation to improve the accessibility of products and services across the European Union and it aims to remove barriers for people with disabilities by requiring companies to make digital products like website mobile apps ATM and online shops more accessible. And inChatGPT 5 – I just realising that it is saying 4 in the sensation – ChatGPT 5 was not saying anything but trying again: all Caps reduces legibility and feels shouting. So, it did find the capitals what ChatGPT 4 didn’t. So it is improving and maybe by asking the question again, getting another answer.

It was the same with Gemini, originally it didn’t mention the capitals but now it does. Google 2.5 was doing better. And it could also be perceived as shouting. And, also explaining All Caps, even if I disagree with what they are saying, it is explaining more. And differently, I think that people with dyslexia will struggle so I don’t agree with the advice but it noticed the capital.

Deque’s Axe Assistant did not do so well: It didn’t find the capitals;

It did not mention the block capitals at all, Axe assistant also tried to rewrite it for me.
And it did something which was really troubling. not finding the capitals, but it was giving the example and it was really not okay.

This was the closing line in the original content:
( Reading the line on the presentation)

it was on top of the content – and what it actually did – was making a list but missing accessibility and this was changing the content. In re-writing it, it missed out the word accessibility which altered the entire message: Missing the point that it was meant to make things accessible.
Instead of stating companies need to make products accessible: It now implies all copies need to actually make those products, And if you use this for legal advice, for policymakers, and this would really challenge the organisation.

Next:
can LLM fix accessibility issues in HTML markup.
There is a lot of vibe Coding – about “vibe coding” now: Which is pretty much where you chat to an LLM about what you want and it builds it for you, but it is not actually yet use, it is a bit weird, and I wanted to see how well it could code for accessibility.

This is the code I gave it: It was a full HTML page, it had a head tag and stuff, It is in English, language, and then everything following in English, H1 out of the content,H1 is outside the main tag, so if you navigate by landmark you don’t get the context, going straight to the content, that because H1 is out, you would start the content without heading.

The other thing: in the paragraph, I have put the direction right to left but also language = EN, and H1 is saying what is wrong, and P – can we tell – and the language is redundant because it is already on the HTML on top in EN so there is no need to repeat it. And English from left to right compared to Arabic
.
Can you find a problem?
Can you fix it?
This is what I asked.

I am skipping through it: the sameIt did find the issue with the right to left reading direction .

and this was the answer:
( reading from the screen)

it got to the point that it should not have the right to left attribute, and didn’t say to change the direction, so, based on Axe assistants solution probably would have worked: But it’s adding unnecessary noise to the code and this can cause bugs – It’s essentially overriding stuff just to set it as the exact same default values – issues in the language to En it is implicit that the language is English and implicit it is left to right because this is a standard.

You also have to look at your code base.
And we know for the paragraph tag, that there is a lot of research that the tag seems to be fine, but it is not that implicit versus explicit.

Some final thoughts:
Ok, I’ll wrap this up and leave you with some thoughts.
These are my opinions, obviously you can make your own if you play around with LLMs yourself.

LLM is not ready to do accessibility.

You can ask questions. Of course. With basic information.
But accuracy decreases with complexity.
Missing nuances.

There is not a lot of good data.
All of the LLM have been trained.
And some items might be right. But not all.

I had no time to go into agents. And into multiple responses. When you have the coding, for example for Apple, the LLM agent will put answers together and do more complex things. Paragraph 4 paragraph. LLM is popular when you go through your emails and chain together comments but when you chain multiple incorrect together it is showing the problem.

This is the chart, showing the accuracy of the responses from LLM and dropping with more steps, Oot-karsh Kun-what, who has built multiple LLM agents has some sobering statistics – If we use 95% accuracy as a benchmark, which is super optimistic for current LLMs – – After 5 steps, it reduces to a 77% success rate, you cannot make something that is a hundred percent accurate. 20 steps = 36% – Because if you have one little mistake, basically, LLM: If you magically achieve 99% with 1 step, you still only get 82% success over 20 steps – and we are being generous – so there is this production of code and you need already hundred percent when using LLM and if you put all of this together, and the complexity, it will lead to something wrong.

Then, economics of complexity don’t scale: To be accurate, an LLM needs to remember lots of context, The more it remembers, the more expensive it gets, If it stacks 50 responses, it’s less than 2 and a half dollars, If it stacks 100 responses, it’s more than 100 dollars.

it is exponential in the amount it costs.
The LLM, when you ask more complex things, and to stack things together, it is having only the context and when you have 50 turns it is having the context of what has been said before, and every single thing, in the previous context, it will start telling you things from the more complex, the more it has to remember and the more exponential the cost.
It is the mathematical nature.

This is the biggest indicator : None of the LLMs are actually accessible
They all make silly mistakes
Steve, who I work with, has been looking into it. And even the prompt box is not having a label, no accessible name. So we have to ask ourselves the question: why not have them build their own interface? it is a highlight to the fact, either we don’t care, and there is also the impact: Steve Faulkner has done a bit of a dig around and a write up.

Anthropic have an accessibility VPAT, but it’s behind lock and key.
Anthropic One. this is cool. And I have been looking into it.
And if you read it then you have to sign the NDA, nondisclosure agreement, so it means probably that it is not accessible because if you were, signing the NDA, you know it is trash.

The problem also is the hype. And there is also this example with the pigeon next to the torch. No, there is not. It is including the examples: and you can have it on the Blog talking about the process – the responses from five different LLM and giving them more context. TetraLogical blog.

I also have it on my own blog: Craig Abbott. UK including some of the things, and if you are into LLM you can do some experiments.

This is everything from me.
Thank you for listening.

– Hello.

I am aware that people would like to have lunch.
But we have time for one question.

– My question, a lot of this has been looking on the LLM as reliable for generating descriptions for images and so on, how do you feel about suggestion – maybe but – even if it is not reliable but still better than having nothing at all? A lot of organisations are in that situation not doing image descriptions and LLM is maybe better compared to nothing.

– That is a good point.
When working on this, with screen readers for example, better have something compared to nothing, but it is heavy Caveat, in working with LLM, and depends on time and budget, and doing something, for accessibility in general, needing to be compliant, and needing to be perfect, better than it was yesterday, and you have no idea if you look at the accuracy, there is the Caveat there, as much as me standing here. If you don’t have those skills, you don’t have time and you don’t have to budget for whatever you are doing, so it is better doing that instead of doing nothing. But a warning: human collaboration and a check is needed and this is a good point. So thank you for raising it.

Rachel Edwards 

Designing for dignity, human rights and design work by Rachel Edwards. hello everybody, welcome back, I hope the lunch was okay, next with got Rachel from content design London, she is going to talk about human rights and design work. I’m excited to hear you talk, welcome to the stage.

Bear with me, I know everybody after lunch wants to have a little nap and digest, if you want to close your eyes, that’s fine. Hopefully we will also wait till the end, I am a content designer, and to channel Troy McClure from The Simpsons, you might know me from somethings like trauma informed content.

I was asked to talk about something else, this is the first time I’m talking about human rights, so you are my guinea pigs. human rights, it sounds big, it sounds important, and maybe it sounds a bit removed from us as designers, we tend to hear about human rights one were thinking about court cases, lawyers, legal precedents, or maybe were thinking of them with think of areas devastated by war, famine, places where immediate and visible crisis are happening.

But we probably don’t think about them when we log onto our computers, we don’t think about them when we join a new project to transform a website, we don’t think about them when we see a video that doesn’t have captions. These actions can have human rights implications.

Our work as designers and creators is tied to helping people access information and services that are sensual to their lives, and their human rights.

We’ve all been in projects where accessibility has been an afterthought, and we know how that turns out, maybe we feel our accessibility audit and we have to invest more time and money into making content accessible, or maybe we and excluding a large portion of our audience who have access needs.

But what if we could turn that on its head? but if excess ability wasn’t an afterthought, what if we had a great big idea of human rights to help stern accessibility from an afterthought into the most important part of our design. That’s the dream, we are going to explore that today, we are going to get stuck into human rights, and see how they can be a powerful and useful to for all of our design work.

What are human rights? they are set out in the Universal Declaration of Human Rights, UDHR, I am going to use some acronyms, sorry, this was created in the aftermath of the Second World War and it set out for the first time human rights that should be protected. The UDHR has 30 articles and discovered things like: no discrimination, no slavery, torture and inhuman treatment. The right to be treated fairly by the court and right to trial, the right of privacy, the right to asylum, the freedom of thought, religion, opinion and expression. The right to work, rest and holiday. The right of social service and the right to education.

So you can see from these examples just how far reaching and arranging human rights are, they cover a wide variety of things and they will relate to people in different ways. All UN member states, countriesn sign up to the UDHR and they agreed that human rights must be universal, they belong to everyone, inalienable, they cannot be taken away from us and in the visible and interdependent, governments should not be able to pick and choose which rights are respected.

The UDHR is a declaration, not a treaty with its own legal force. Countries must have legislation brings this rights into their own laws, this provides a legal framework for bringing cases if your rights are being met are being violated. So when countries ratify these rights they create internal frame for people to implement them. Ratifying means the country has put things in place of that act is legally enforceable, rather than just an agreement statement.

This isn’t only document of human rights, there are others that provide protection for certain groups, the most important for us in this thought are the Convention on the Rights of Persons with Disabilities, and the United Nations Convention on the Rights of the Child, so I’ll start with the Convention on the Rights of Persons with Disabilities. There are other rights document around disability, but were going to focus on this one.

As the convention says, the goal is to protect and ensure full and equal members of society which human rights, the UK has signed and ratified the convention, US has signed it, but not ratified it, there are eight guiding principles to the convention, respect for dignity, autonomy, including the freedom to make once own choices and independence. Nondiscrimination, the full and effective participation and inclusion in society. Respect for difference and acceptance of disabled people as part of human diversity and humanity. Equality of opportunity, accessibility, equality between men and women, and respect for the evolving capacities of children with disabilities and the right of children with disabilities to preserve their identities.

The preamble to the convention, which is basically the introduction, is actually really beautiful, if you ever want to spend some time reading. It talks about why the convention is made, and some of the points it makes include: disability results from the interaction between people with impairments and barriers, both physical and attitudinal that limited their participation in society. It’s important for disabled people to have autonomy, independence and freedom to make their own choices and be actively involved in decision-making, disabled people face barriers and violations of their human rights in all parts of the world, disabled people are diverse and people with disabilities have other characteristics that might bring further challenges to realising their rights.

The majority of people with disabilities living conditions of poverty. And finally it mentions the importance of accessibility in physical, social, economic and cultural places, health and education, and information and communication. So disabled people can fully enjoy human rights and Fundamental freedoms, the last one is expended in article 9 of the convention and it says that people with disabilities should be able to live independently and participate fully in all aspects of life.

They should have access to the physical environment, transportation, information and medication technology other facilities and services open or provided to the public. Accessibility in the convention covers three main things, access to communication and information is one of them. we cant enforce rights without domestic legislation, would accessibility we do have legislation that covers some of these areas, like the Equality Act in the UK. But I’ve also said that this talk will challenge you to think beyond accessibility and compliance and that’s what I want to do.

Can we use human rights-based approach to put accessibility at the heart of our design? This brings us to a fundamental point, to access our rights we need to understand your rights and to understand our rights we need accessible information. Before I was a content designer I work for the equalities and human rights committee of the Scottish parliament.

I worked on two bills: the Female Genital Mutilation protection order Scotland Bill and the children Bill, which is better known as the smacking ban. I’ve got a very serious grumpy face on this picture. The FGM bill was about providing protection orders for people at risk of FGM. This meant working with people who often had regular and recent experience of their human rights being ignored or violated.

The equal protection or smacking ban bill was about getting rid of a legal defence that allows parents and guardians to use physical punishment on children, called reasonable chastisement, the bills aim was to entitle children to the same protection from assault as adults. Both of these bills meant we needed to engage with people who maybe don’t usually engage with the parliament and politicians. This had a number of challenges.

For the FGM bill we are talking about groups where English was generally an additional language, many were asylum seekers, with a large distrust of government and official structures, they didn’t have access to desktop computers, they were avid to mobile phone users and FGM was a subject they were not tended to talk about. Parliamentary call for views are not exactly easy to read, they have technical language, they often require a word document response that addresses a certain set questions. It became more clear that this was not going to work, it was inaccessible on many levels.

I thought that we could translate the consultation into the languages that were usually spoken by these women, but I met with a charity that supported them and they said: don’t bother, there’s no point, because the literacy rate among these groups were so low that they would struggle to read in any language, they told me to prepare videos they could watch on their phones. We designed the consultation that women could safely submit their views through video, digital storytelling. Anonymously but in the own words. It respected the rights to privacy, to safety, but also to participate in politics and the matters that affected them.

It meant we had to do things differently and take a different view to a accessability and inclusion. But it set me up really well for the equal protection bill. The equal protection bill was introduced in 2019 and there are proposed amendments to similar legislation currently going through Westminster that would give children the same protection in England.

We know from existing research the children who were most likely to be smacked under the age of 12 and usually under the age of eight, so how do you design some thing that children can read and understand? That was one problem, but there is another one. Children under 12 could not consent under GDPR to having their views published as part of the consultation, that meant their parent or guardian had to consent for them, often that means that the person giving permission for them to share their views about physical punishment is the same person who could be administering this punishment.

It seems like a puzzle but when you are designing for children, anyone under the age of 18, you have an extra, big stick, it’s an additional one. Children have additional rights, as well as those in the UDHR it is signed by hundred 96 countries and outlines the fundamental right of every child. It’s the most widely ratified human rights treaty in the world.

The Convention came into first in the UK in 1992 the UNCRC is made up of 54 sections or articles that cover all aspects of a child’s life. Article 12, children had a right to give European and for adults to listen and take it seriously for the article 13, children have the right to find out things and share what they think with others. Article 17 children have the right to get information in lots of ways, as long as it’s safe. Children should be able to access information they can understand on TV, radio, in books and newspapers and on the Internet.

Prof Laura Lundy is a professor of children’s rights at Queen’s University Belfast and Prof of Law at University College Cork her research specialises in children’s education rights, rights to participate in decision-making and implementation of the Convention on the child. She developed a model to help explain how to help a child to participate and it is called the Lundy Model, she says: children have a right to seek and receive information including about their human rights. If this right is to be implemented in practice, the information that they receive needs to be available in ways they can understand.

I love this LinkedIn post from her, children don’t deserve their human rights, they don’t deserve to be heard, it’s not earned or lost as a reward or punishment, it’s an entitlement like all human rights. At this point in my career, I am working on these bills, I was not a content designer by name, but my work on the bills made me realise the importance of clear communication.

If someone cant understand something, how can they anticipate? How can they engage in a consultation if they can access the technology or the documents or if the questions don’t make sense? What if they need someone else to consent or contribute or to translate for them? the equal protection bill impacted on private family life, which is a human right. The children also have rights for safety and to participate in decisions that affect them. So again, how do we ought these rights?

We made a special consultation for children. We made toolkits for classrooms that were facilitated by a trusted adult. That adult could collate responses that could be anonymized and send them to the committee. We used clear language to explain what the law meant and what the changes could be. And we fed all of that back so they could be active participants.

So this was a solution, but I’m not sure it was perfect, because it still relied on adults collecting and potentially filtering children’s responses. That idea of needing someone else you can access your rights, goes against both accessibility and the human rights-based approach.

So here is a little illustration of what I mean, a couple of years ago I did an event of making lots of two young people, there were a lot of lawyers there, including Laura Lundy, it was quite intimidating. I will start the importance of language and young people understanding their rights.

I did research of people aged 7 to 17. I showed them things they were entitled to as children living in Scotland, free bus travel under 25, and free university tuition. I asked the children to read the information and to tell me what it meant, that language was technical and complicated, I’ve got an example on the screen: children and young people who are 5 to 21 years old will need a new replacement national indictment card or young Scott national entitlement card. These are long sentences.

I asked them what they would do next if they were giving this information, the younger ones would say: I would give it to my parents and get them to deal with them. So this was information about children’s right, but it wasn’t written for children, so someone else should access the information and their rights.

The children I spoke to came from comfortable families where there were adults who could help, but that isn’t the case with every child. A child whose parents had English as an additional language or who couldn’t access the Internet would be completely prevented from accessing their rights because of how the information was presented.

Children have the right to privacy too. They might be comfortable asking their parents to explain their rights to free bus travel, but what about other areas. For example, with a teenager be comfortable asking adult landlords around sharing intimate images with their boyfriend or girlfriend? Maybe not.

So they might go looking for answers themselves. If they did, they might fight information like this, technical legislation that they can’t understand. I have an image from the production of children 1978 that refers to taking and distributing indecent photographs of children. It’s very technical and complex, but it’s information they might need to know.

So when they cant understand it, they turn elsewhere, to Reddit, to Snapchat. if we make information that people cant access or understand, we force people to go elsewhere and we have no control over where that is. This is just one angle of accessibility related to the understandability principle of WCAG. with heard about images without all text, videos without captions, animations that lack an alternative for someone with a visual impairments, there are some of the ways we might be preventing someone from realising their rights.

So let’s put all of this together. Children are a good example to start with, because for the most part they rely on other people to realise their rights. as designers are organisations can be seen as having a duty to. And we can use with gold human-based approach to think about our responsibilities.

It’s meant to empower people whose right are most at risk so they can participate in decisions that affect them and hold duty bearers to account. The approach has five key principles: participation, everyone is entitled to participate in the decisionmaking processes that affect them and their rights. Accountability: duty bearers are accountable if they fill in their obligations towards right holders. Non-descript and equality, all individuals are entitled to the right without commission of any kind. Empowerment: everyone is entitled to claim and exercise their rights and they need understand the right and participate in the development of policies that affect life. Legality: approach should be aligned with the legal rights of touting domestic and international loss.

A lot of these concepts are probably already familiar to you. But they are all things we need to consider. Let’s look at what these principles might mean from a design perspective.

Participation: recreating designs people can use, can they participate in decisions that affect them? Are we being inclusive and accessible with our user research and testing? Are we making sure we are considering a range of access requirements will be invite people to codesign something? Do we make it easy for people to participate, we don’t require them to login, or give us extra information we don’t need.

Then back to the MGM bill, we could have created an accessible HTML page with consultation questions with the correct screen size, font size, it would meet accessibility requirements, but if someone cant understand the questions, or access the technology to fill in a response or if they need to speak someone else to do it for them, is accessible? are they able to participate?

So far accountability, what duties do we have two our users? We might be official duty bearers if we work in government or other organisations with power, the tower we responsible for their safety and care. Do we provide ways for people to feedback will decisions? Are these meaningful? if we are creating a space for people to share views online, is that moderator? If not, how do we make sure people are not being comprised online? And if it is moderated, how robust, fair and free from bias are our guidelines? How does someone appeal object or understand our decision making?

With most of our designs, our day accessible? Are they inclusive, are we considering a range of groups and users? Have we considered the users of our design? Our day accessible to our range of users, are we disempowering anyone through jargon, confusing content or user journeys that don’t make sense?

I recently tried to complete an online consultation, the web age crashed, repeatedly, it wouldn’t recognise my address format and asked confusing but mandatory questions. In the end I gave up, I had tried to participate in a process but got nowhere, it was disempowering, because of accessibility and bad technical build.

So finally, legality, do we consider data protection, accessibility legislation and other rights when we design? Are we designing for compliance? Or to the spirit, as well as the letter of the law.

human rights approaches are really useful when you come across conflicting rights. You might see this in the design work when we have competing user needs. Whether you do when one persons or groups rights appear to be at odds with another’s.

with the smacking bill, children were obviously at the heart of the bill, but so are parents. One of the rights in the UDHR is about a private family life. Article 8, protect your right to respect for your private and family life, private life means you have to write to live your life without government interference.

The children also have the right to be safe. article 19 in the UNCRC says: I have the right to be protected from being heard or badly treated. Government should make sure the children are rabidly cared for. There should be laws to protect children from violence and abuse and neglect from adults.

In case of conflicting rights, human rights-based approach provides a clear answer. We prioritise the group whose rights are most at risk. In this case the children.

If we think about this for a minute it might turn things completely under head for us as designers. If it for you is to taking a top task approach where we design for the most common need, what does this mean? What does it mean when we talk to Stakeholders about accessibility and the situations we bring up, or the users were talking about are dismissed as edge cases, if we can actually embed this approach, if we can put it into the heart of our design work.

But might that look like? would we prioritise creating BSL content for a webpage, and put it at the top of the page, because we are prioritising their access to information which we have the backing to build accessibility. which we have the backing to build accessibility in first and not last, it’s an exciting thought.

We are going to talk little bit about people whose rights are at risk, this might seem to tear, but for people whose rights are at risk it isn’t theoretical, there are a few things to think about. And going to split them into technical bits and emotional bits.

the technical constraints, someone is trying to access critical information in a crisis. They might have low battery and signal, they might be in a location with poor network connectivity, they might not be able to move to fix these things. So we have to design in a way that is lightweight and functions with low bandwidth.

in many crisis zones governments can control information, they can cut off Internet access, messaging tools, access to certain new sources so we have to think about how information can be distributed and received when standard channels are blocked.

We also need to think about emotional constraints. When people are actively experiencing trauma or crisis they are not operating in a normal cognitive state. Stress hormones lower our cognitive function. We can’t think as well as we normally can. Asking someone to make a decision, to weigh up options or to decipher what they need to do next, is asking a lot.

So what do we do when we design for a person in this state? We want to focus on clarity, using clear and direct languor. Simple processes, to make the process as simple as possible, and we want to prioritise accessibility, provide video content with transcripts designed for low battery life.

The good notes is that this really works with good design. Not really doing anything differently than what we should already be doing, were just applying our skills to situation where they are most needed. With come this far without a technical problem, there we go.

So if we bring all these things together, accessibility, human rights and crisis situations, what might that look like for us as designers? Imagine a government needs to send out emergency alert. This was on a Sunday, I was drinking a cup of tea, I have a couple examples, one is from last year, and the other one is older message about COVID lockdown is, we might have views about the language used, but these are just a couple of examples.

A poorly designed message might be a text only alert that is too long or uses technical jargon. An accessible message would consider technology, battery life and signal. It’s going to think about the effect of stress and trauma and what that might be doing with someone’s cognition and to include forms for visually impaired people or those with low or no literacy.

The point of this is that people need to know information and how to act on it, my colleague Miriam Vaswani does some great work on this, crisis communication. those are the accessibility considerations, a human rights-based approach, would consider balance between rights and needs. For example the right of safety and security, with what the government might need people to do, they need people to leave their homes and belongings in order to stay safe. And it will also consider which groups rights are most at risk and how they can be prioritised.

We need to make sure that our message provides factual nonbiased information so people can make informed decisions about their safety. So you can see there is some crossover between accessibility and human rights thinking, we thought that the people whose rights were most at risk in this scenario, were people with visual impairments, how would we change the design for them, which we prioritise a video message instead of text or think about using other channels?

What if the crisis situation was in response to a political protest rather than a weather warning. The government wants to disperse crowds or directed my way from an area, we might have to conflicting needs: to protest the rights to freedom of peaceful assembly, and the public’s right to safety and security. And this is assuming a benevolent government that is allowing the protest and is simply concerned about safety.

when the messages are being controlled by a different kind of government, what difference would that make? something we need to mention here is the role of tech and communications companies in human rights violations Amnesty International has just released a report called ‘breaking up with big tech’, which looks at how the power of the big five companies can perpetuate human rights violations.

they write: when a small number of companies control the primary gateways information, their decisions can have a systemic effect on the public’s ability to access diverse viewpoints and is on access to information freedom of opinion and expression. Did means that as designers working in the spaces are choices have to tickle in.

To prepare for this talk I put a message out of LinkedIn and asked people for a one-word answer, what’s the first thing they thought of when they hear human rights. The most common one was dignity, followed by freedom, justice, safety and Respect. If you look closely, there are others, injustice, erosion, oppressed, unequal, Gaza, misappropriated, fail. It’s clear that human rights for many people are not a shiny beacon of hope.

77 years on from the declaration of human rights were looking at rights today in a digital context that is graders could never have dreamt of. But we can apply the spirit and ambition to our design work. We can use the idea of human rights to promote dignity and equity.

It doesn’t need to just be a big stick in our list of things to wave at Stakeholders. Human rights-based approach gives us a framework and authority to argue that an educator for the business is a fundamental right for the user.

Eleanor Roosevelt, one of the main campaigners behind the declaration of human rights famously said: where after all the universal human rights begin? In small places, close to home, Lesley’s right of meaning there, they have little meaning anywhere. We can bring those rights into small digital spaces and we can help our users begin to access their rights.

A few resources for you if you want to grab a picture, if you want to know more, Prof Laura Lundy who works on the children’s rights model, she has a lot of content out there explaining rights for children. Eriol Fox is an amazing designer with experience in design, technology and human rights. She shares a lot of resources about designing in areas rights are at risk. I’ve mentioned my colleague Miriam Vaswani, she has written a chapter about crisis communication. A lot of the chapters are available for free online, you don’t have to buy the book, And she mentions the researchers she used, so if you want to know more about crisis communication, that’s a good place to start. Emily Allbon is a director of legal design at the city law school and she launched a website called the TLDR, and shows different ways of making a lot more accessible to all.

And that’s it. Thank you very much.

.
.
Hello, thanks Rachel, this anyone have any questions for Rachel?
No?
– You can ask me anything.
– There are some questions over there.
.
– thanks Rachel, that was really interesting, something you touched on was about how you can use this when you’re building communities, and the code to conduct, often they are a list, I’ve wondered if you have any experience how this is considered as part of building the communities and if you have anything to afford them or maintain them.
– you mean the PANEL principles?
– yeah, are if you have any takeaway’s?
– the panel principles are designed for that, it’s about supporting communities and people, developing a service and a policy situation. the Scottish human rights have a lot about this. It’s a digital context, so maybe less about working directly with people. But the important bit to remember is when you’re involving things around Coke creation, or testing research and how accessible, how much you are supporting people to participate. You’re going to have a consultation on this, but if you put it in an upstairs room or community centre where there is no transport or a lift, you are not supporting people to co-create with you, the Scottish human rights commission, they come from the Scottish context, so IMO bit biased.
– Did you have a question, Craig?
– I was curious, in terms of the impact this has on day-to-day work, the use that secondary stick having an impact? there’s a lot of, I suppose, obvious examples of human rights being violated, and there is a lot of stuff where people are reversing the rights, just in terms of the stick.
– It depends, you have some people who are saying, I must comply and some who are not. if you say to the Stakeholders I am looking at something and is full of jargon, and its technical, and it’s going against someone human rights, it’s the same way that we approached things like with the task approach. Instead of design for the many people, we are designing for the people who are most probable. If we can do that, I know our power, as designers is often limited and controlled by other people.
– a lot of things you mentioned, it mentioned the word reasonable, is it reasonable for the company to do X or Y, I’ve seen a lot across digital spaces, and it’s sort of a copout, any tips?
– I think as far as I know, not a human rights lawyer, we only have to make a reasonable effort to make information relevant to you, so you can understand. It is stronger protection, it’s a stronger mandate, the good thing about having that equality legislation is, you have the theoretical, and the mechanisms to support the human rights.
– The mechanisms, and you talk about the legislative sticks, you also mentioned that this is assuming a benevolent government, what if you don’t have a benevolent government. Does that mean that at that point, how critical is that, is it possible to still be incorporated human rights into design, if your country has revoked the legislation of human rights.
– It is a wider point, but it’s a good one. It’s about having a legal framework in place, make sure that it can’t be eroded, and then having the mechanism to enforce them in court, so it will very much depend on the countries, the situation that you’re in, I’m not saying that human rights gives you a magic wand, it’s another tool, it’s another way about thinking about this. For some people it’s powerful and important, I hope that for all of us it’s powerful and important, but there are places where that you should get the attention.
– Thank you, Rachel.
– We are going to have quick break before we welcome Ettie, take 10 minutes, and then we welcome you back. Thank you.Ettie Bailey-King 

– Hello. Welcome back, I would like to introduce our final speaker of the day, Ettie Bailey-King, she’s made an accessible and inclusive language Educator. You might know her. I’ll just let her talk.
– Thank you for that introduction, I’m Ettie Bailey-King, my pronouns are she/her, if I slightly ensure way from the Mike, can you still hear me? I see nothing. A writer on the screen there is a photo of me, I’m a white person with blonde hair sitting with a stack of books smiling. I’m going to talk to today about communicating for frazzled minds. What is a frazzled minds? it’s an object of typically used in exhausted, tired, anxious, I’m going to use it very expensively today. I’m going to talk about three things: what I mean by frazzled, how being frazzled affect our communication, how we remember information. And entry, practical tips for frazzle-friendly communication. so what do I mean by frazzled, exhaustion, rate, destruction, and comfortable, it could be physical pain, internal distractions, too hot, too cold, maybe you hear voices , maybe you’ve had a bad day, I will commit all under the umbrella of enfrazzlement, I want you to think about a time when you felt comfortable.

I want you to quietly reflect and think about a time when you felt comfortable. All give you a few seconds. It could have been anything, a warm bath, a quiet room, somewhere you knew you could share your access needs and they would actually be taken seriously. That is what we, me and Mia Mingus call Access Intimacy. There are a lot of ways to make you feel comfortable, but I wanted to reflect on: what did comfortable feel like, in your body? where this comfortableness show up for you, but does it feel like, and what it did not feel like? A few seconds just to think about that on your own. It might showed up as feelings in your jaw or shoulders, the absence of tension, sweaty palms, the absence of butterfly in your stomach, whatever comfortable feels like for you, and going to LinkedIn with very basic neuroscience, I must issue an apology if you know a lot about neuroscience, I’m sorry to do this, because this is oversimplified.

I’m going to quote: all models are wrong, but some are useful. Is attributed to statistician George E. P. Box, I’ll slightly twist his words: all theories are wrong, but some are useful. It can still be incredibly helpful and access insight. Let’s get into some theories. First we have the Discover-Defend Axis, those are two extremes in how our nervous systems make us feel, I”ll put two images up, discover remote and defend mode.

On the left is a person picking juicy berries, when you’re in discovery mode, you are relaxed, open to opportunities. on the right we have the image of a magnificent tiger who is roaring. It signifies the defend axis. I’ll unpack what the two modes look like. When you are in discovery mode, it’s not a binary mode, we move between them, that if you are more towards the discover side of things, you will be physiologically relaxed.

You will not have a racing heart, sweaty arms, butterfly in your stomach, blood will be going to the organs in your body, not only that are relevant to survival, we don’t have to escape danger. So we can rest and I just. You are open to looking for opportunities because your perceptions are wide open, you are more open to possible opportunities, you are prone to look for juicy berries. was a juicy berry?

No, it wasn’t. Can I get an indication, make a noise, is anyone familiar with System 1 or two thinking? thinking fast and slow, I see some wiggles of movement, some of us have come across it. Get ready for some more oversimplification, System 2 is slow and deliberate conscious thinking, spacious thinking, you are deliberately tackling a problem.

By contrast System 1 thinking is quick, non-conscious, prone to errors, highly efficient, and it keeps you alive in the moment. When you are in discovery mode, you are looking around for opportunities, and your perceptions are deep. when you are in defend mode, you are in fight or flight mode, other modes are available, freeze, fawn or flop, there are others, but your body is looking for survive.

In looking for threats. You can make really quick reactions, that is why we are alive as humans, because our ancestors are able to go into quick and defensive reactions, but they are not the most accurate. That is defend mode, it’s optimised to survival, it contrasts with the discovery mode, which is more relaxed and open to opportunities. If we spent all our time in discovery mode, how lovely would that be, life was just looking for berries, but if we were always in discovery mode, if we are always relaxed, can’t, well rested, nourished, million night of sleep, no emails, then a lot of these things wouldn’t be a problem, long emails with technical jargon or blocks of text. but alas, we do not spend most of our lives into the relaxed mode, by contrast, I have an image from an Instagram video by Korey Dior, it’s a black person on screen with heavily sarcastic look on their face. The caption reads: when my nervous system learns the difference between a tiger and an email, I’m going to be unstoppable.

The fact is, 8/10 of us don’t breed normally when using a screen. Has anyone heard of email apnoea, can I get movement or sound if you’ve heard of it? So email apnoea, we have a screenshot here of an article, the headline reads: we regret to inform you that you may not have email apnoea and not even realise it. Your inbox can literally take your breath away, here’s how to not let it. You can breathe very shallow when you read digital content, there’s not enough research, so I don’t know if it’s also about listening to content, but if you’ve listened to a really long email through a screen reader, you are not having a good time. So I think this also applies in other ways.

If you’re not already frazzled by existence, then screen use may frazzle you more, you may start off in a pretty regulated state, but just having to access your email inbox or open up your laptop or use your smart phone, which you probably have to do in crowded place, while moving.

So I’m talking about exhaustion, distraction, most of us move between modes. In discovery mode our perception is wider and deeper, because it’s feels safer to forage for berries, and in defend mode our perception is narrowed because we are trying to stay alive. So how does this affect communication. How does this show when comes to reading, listening, using information. Being frazzled faxes in at least two ways. If we are a consumer, then we are much likely to react quickly, missing the joke, making assumptions about what author meant.

Our brain is prepped for danger and survival, not prepped for understanding a lot. So as readers or listeners we forget details. As creators of content we are more likely to info dump, although this is shaming neurodivergent people, so I’m going to rephrase it as share large amounts of lacerated information. So I’m probably not going to let that pardon, because I’m trying to stay alive, and I’m going to assume our audience knows what we know, empathy, a mature ability to understand peoples perspectives, it comes hard when we are under stress. So simple writing may be easy to understand As a consumer, but it’s not always easy to write. You basically need to be in discovery mode, feeling safe, having a lovely time, to help the reader who is in defend mode, because their perception is narrowed.

As a content creator, we are human, we also have stress, stress is like… you can read some of these. This is meant to be very stressful, so you might be stressed out by things like, cant pay my bills, there’s a recession, I’m hungry, ableism, my obsessive-compulsive disorder, the climate breakdown, knowing things about the world, capitalism, and safe homes, 35,945 and read emails, Worrying about losing my job, I’m going to stop there and give us a pause 2D frazzle. I would like you to breathe in 44, hold for four, and breathe out for four, just breed in whatever way you would like to, but if you find that helpful, take a moment to breathe in. Because all of those stresses are a lot. Content itself cannot take away all of our stress, but it can cut through some of the noise, it can work with our frazzled minds, it can avoid refrazzling us, it can help is by not judging us, shaming or blaming us. And it can help people get what they need, they are reading your emails or on your website, because they need to do something, learn something, taken action, and your content can help them to do that, so recap, being frazzled can make it harder to consume content, and create content. Ideally, we content creators would all be in discovery mode because we’d feel relaxed and rested and we would have the energy to create perfectly accessible content. Content can take away all of our stress, thoughtful content can work with us, not against us and our frazzled minds. 

So, practical tips for frazzle-friendly communication. Reduce the cognitive load, this is an umbrella, it applies to loads of other actions. What is cognitive load? is anyone familiar with it? Make a noise. Yeah, so cognitive load is the load placed on your working memory by information. Your working memory is your memory you need right now, so you can do something with it. It’s not the one you use for remembering the distant past or memories. If I say: hello, my name is Ettie and I am an inclusive language educator. You are using your working memory to hold that information. You might be holding in your mind, she said hello, I have to return that. She said her name, maybe I should offer mine. She said her job title, may be I should offer mine. You have to hold this information and the different expectations that are triggered by those pieces of information. It’s not exactly the same if you are just reading that often email or an business card but it still triggering other Association and you have to hold that information in mind. Research suggests that you hold information like sentence length, so if I had broken the down into several sentences. 

Hello. My name is Ettie. Research finds its easier to hold it. the key thing to understand with your working memory is: it is highly limited, remembering information crisis is hard work, the people who are studying it saying it’s effortful work. We know this instinctively, even a small interaction with another person. So all of us, even if we are not neurodivergent, nondisabled, having a lovely day, we all need short words, sentences, concrete language and other things I’m going to talk about, they help us focus on process information. I’m going to show you practical examples of how to do that. That’s it, that is the tip, say less. Literally, say fewer things. My mantra, when I’m writing or speaking is: if in doubt, cut it out, if I have any doubt, I’m going to cut it out, that can be uncomfortable and hard when you are working with other people in an organisation. we must tell people the story of our foundation, how can they otherwise understand us? it can be really hard, but where possible, modelled the eye idea of say less. closely connected to the idea of saying less, is pausing more. you are creating space to let people think, you are asking people to hold that in your memory. 

You can make a full stop, It creates a pause for processing time. If you finish your alt text to screen reader will pause and give the listener processing time. try introducing it, and then we might not think of it as a pause, but is creating space, it’s the gap between each paragraph, it’s the literal space in the run, the sense of spaciousness, you’re not just throwing a lot of information at someone and expect them to act on it, no need to respond right now, can you can back to me by Thursday five p.m., it’s a different way of processing. Pause.
Simple synthesis. And now grammar time. A simple sentence. Full Stop. And this is the complex sentence. Adding information and reader must follow in the working memory to understand the whole sentence. Sentence one is simple. Sentence two is complex in the grammatical context. One bit and the other sentence having a lot of bits. Now, you will have to make some movement or making noise to indicate this is your favourite.
If your favourite sentence is simple sentence one: seeing a lot of movement and hands and waving. This is pretty strong. The other sentence, 2, is this your favourite? Only one lone hand. This is Brave. No shame. We are not judgemental. This is about the terms of energy we like to spend.

If we can, we will try to save energy. And not trying to deal with more information at one time then we can avoid dealing with, and for that reason, we will mainly like simple sentences.
And, writing would be so boring if it was only simple sentences. If every sentence would be six words, we would be sleeping. So, it is an invitation to write mainly simple sentences but not all the time. And to create rhythm. And contrast. And energy. And the lean towards them, if you can.

Next. Now subheadings can help scanning the page. So, it is like that I movement and tracking and doing this in the context. Not always. There is patterns and scanning the page, and it is like when you do scan the web page and you are looking for subheadings. And it is also the same for screen readers. Using this to true the page. It is the same for Braille readers. And if you lose focus, there is the internal and external distraction, subheadings can help finding your way back. You get lost. There is an interruption. You forgot what is going on for five seconds. or, you step back from your desk, and you can find your way back because of the subheadings. But it is not equal: it needs to be meaningful.

Meaningful subheadings give you meaning: they are not just a topic word. A less meaningful subheadings might say things like energy, time, attention, they named the topic and it is better than nothing but not giving you actionable meaning. It works well if you only read the subheadings and they can still understand some of what you have been saying. This is a challenge for you all: writing the content, email, the webpage, can you make it so that the subheadings work for you, or almost everything, we have the example: showing that you don’t have to do accessibility perfectly.

There is a lot of things that I don’t like in this example.
For example, the capitals.
For example the background.
Those things are not accessible.

But I am sharing it, because you don’t have to be perfect.
The subheadings are meaningful. It is about running. And I know, start slowly, focus on your feet, take in your surroundings and pay attention to pain or discomfort. And if I go on a run I have the information I need, basically not reading the paragraphs and I still have something that I can take away from it. If the subheadings would just say starting, feed, surrounding, pain, I don’t take away the action. So look out for that. The people content, and how to make it meaningful. Actionable.

Give people information, just in time, not just in case.
So, what do I mean?
Just in time information means that you tell people only what they need to know when they need to it. What this somebody needs to know, now to purchase the product or to attend a conference, in contrast with just in case information which is nice to know. If you knew it, it could be helpful to know it, but it might happen and maybe not, it is maybe relevant and nice to know and you don’t really need it now in the moment. It could be essential for somebody else, later but at this point in time, when I’m trying to add a product to my basket, then it is just in case information you don’t need it exactly right now. And this is why you have to FAQ, normally, the home of Just-in-Case information. And your reader that has so much going on, just needs to in time information.

A metaphor:
supermarket store the food based on what is needed just in time and stock only what they have reason to believe that should be purchased just in time, and if you bring that mindset, to just bring enough in stocks of people can have what they need and not wasting time on energy with other considerations, that they are not sure they will need, and it would be helpful. Let’s think about scanning. There is one myth: only sigthed readers can scan the page. That is not correct. We also have screen readers. And there is a tiny minority, for our purposes, almost everybody scans.

Sighted readers will scan the page, scanning patterns, F pattern, webpage, keywords, layer take, and the content in the first place, but we know that sighted people are not reading every word on the page. And also the screen reader with headings and jumping through them and listing of hyperlinks and maybe a few paragraphs. And maybe skipping character by character. Braille would be do the same thing. We have to structure scanning.’ People will only be scanning the webpage so let’s not worry’- it’s because they are scanning that you have to doing this and how, with bullet points, with lists, with frequent meaningful subheadings, and one idea per paragraph.

Another way to think about this: let people paddle, swim or dive. And you can paddle; but it was not useful for you, but you can understand the topic. You can swim when it is interesting, having enough information to know what is going on but you don’t want to go deep and you move on, and you might be diving if you are very much interested and it is perfect information for you and you would like to explore the topic more in detail. Scanning on want they need. There is the background, news journalism is back in the ancient days with print and copy, another world, there is pyramid: upside down. At the top, a long line with the lead. W-words. It is getting at the idea, news reporting is very often getting the first paragraph in place and very important because there is the reason. And you will get almost everything you need on the top, and if you go down the triangle it is going to take you to more details, the body, and even more related context.

Why my talking about this? You are tired. Your brain is raising as a Ferrari. You will not go to the content. But you might scan. And you would like to know what is in there, and if you have more energy you might explore further. If someone is having three seconds to skim the page and read the email with very important information, will they get all the information they need? In 3 seconds.

Can you convey everything when you speak to someone in three seconds?
A challenge for all of you.
Find someone who is new to this, one colleague in the different area of work, try it out, give information in 3 seconds – and what about five seconds – and what about 10 seconds – how quickly can you give people what they need. Can you challenge yourself? Push yourself towards that. I would like to imagine that if you have the phone that is almost about to die, battery low, you have only some seconds left.

That reader, literally, the phone is about to die, with literal language:
what do we mean with literal language? I was making lunch and I spill some beans.
Nonliteral: I try to keep a secret but I spilled the beans.
That’s a phrase.
Literally bending over for an idea, and what are is boiling at hundred degrees Celsius.
nonliteral: you don’t need a jacket, it’s like 100 degrees outside.
so. It is being literal versus nonliteral, and how you have this conventionally accepted as conditions.

Let’s hit the road.
Give me a hand.
Break a leg.
Bear with me – it can be vivid, engaging and emotive, and we need to look like situations like policies, instructions and advice, and you can have a very literal subheading but giving people a clue in the nonliteral body. With a figurative subheading but being literal in grabbing the attention so people know what you are talking about.

Be clear: unambiguous.
There is many possible meanings. It is not always Bad, offering multiple meanings can be engaging and wonderful, and it can cause mistakes and also thinking about the cost of ambiguity. And step-by-step list of instructions, and for time, and to have this example, and I will leave this example as knocking on the door with a cup of coffee, this can have some meaning to some if you would like to be clear, some words can sound the same like homonyms and homophones and cultural references and anyone working with people in the US, there is different references, in every conversation. It is about metaphors, unusual words, technical words, clear language. In subheading or vice versa. And now, do have your questions ready. For one moment.

You can also ask yourself or AI you trust, is this simple, it is clear, is this literal, can I remove jargon? It is to help, to reduce cognitive load, say less, and if in doubt, give just the structure, make subheadings meaningful, and use the inverted pyramid and challenge yourself and be literal, and be clear.

This will help your audience whoever they are. Be kind to yourself. And if you cannot create perfect content, or strip out academic language, police, be kind to yourself. All the challenges are here, please don’t take anything like this just congratulate yourself and celebrate yourself.

Now, I pause myself, so we have time for question. Thank you so much. And if you would like to give me feedback at some point, I’m open for that.

– We have the microphone.
Someone having questions?

– Thank you.
Interested to get your thoughts on how this can apply to use this test environment, and how to differentiate and how to look at it from what is the additional stress in the simulated environment when their baby pressure, being watched, and needing to answer a certain way, and what is your thought on that?

– I think, to some extent, I would say, why worry if some of the stress and situation created by the scrutiny and pressure of user testing, while users are still working their way through content of other stressful situations, and the new situations, home, kids shouting, baby screaming, there can be still challenges for the attentions, so, obviously data that is clean, and actionable, we should absolutely understand some of pressure comes from scrutiny, and for neurodivergent people it would be difficult to be perceived, it is also a massive challenge. It is about the insights, testing and experience, and everything you are already doing, but also making it clear, on what they are assessed on: making it clear, we would like honesty, but not polite answers, but we would like to know the challenges, observe, rather than assume. Instead of making assumptions on why they didn’t behave the way we hoped, have questions, and maybe something that would be the last thing you expected. And we would work with people, real people, with different life, stresses and challenges.

We cannot take everything out of the mix.
But being the human being, and to make and change to make it work and remove noise in the testing environment, rushing, hurrying, unambitious instructions but not try to control out.

Even in the perfect environment, I will misunderstand stuff.
For example there is a quote of my history teacher: Ettie Bailey-King your superpower is that you literally can misunderstand anything. That is correct. The challenge is showing up for everyone. And also in design. And this will not present well.

– Hello, thank you, this was an engaging talk Ettie Bailey-King and you mention AI briefly at the end, I am advocate of thoutfull AI, and it can be beneficial to communicate about things, with divergent friendly AI, to make some recommendations may be, do you have some tips that could work well when you are using AI?

– Confession, it makes me feel old-fashioned, I don’t like AI. It is in the world. I hate it. Yet I use it. For notetaking for example. I use it in a lot of areas in my business. There is energy levels, and also pathological demand avoidance, that people reframe as persistent drive for autonomy. It depends on what you ask to do. And the role. You are expert editor. Maybe strip out jargon, or idioms, that could be really helpful. Whatever you are using, with instructions and accessibility standards for the output. I might give instructions for formatting. Don’t use italics. Especially for dyslexic people. Don’t use underlines. Look at your subheadings. Things that would work for as many people as possible, and clear in what you are asking to spot, because sometimes, giving a task: making it accessible, there is so many assumptions.

So, how can we be explicit, make it simple, clear, literal and unambiguous.
Like, what assumptions have remained in content creation: did you care about a certain topic, and if you encourage friendly robot and think about it as ‘ friendly’, because they will take over, and encourage it to be on our side, and to explain that, then it would be amazing for accessibility. I hope this was helpful.

– We have time for one more.

– I would like to highlight that a lot of the principles you have brought in the accessible writing is also applying for other activities, as a workshop I was having on Friday, and now highlighting what we didn’t do very well and we didn’t make the task and asks not clear and we could have make it simpler, and it was very relevant and making it very helpful for me.

– This is amazing to hear, thank you, very often for people with dementia or trauma work, we all just converge back to the same things, because whatever human being you are dealing with, and – sorry being generalisation here – but for a big part it is to be given space, to have things broken down, and will have to opportunities to overthink, and getting something wrong, taking off the plate. To get it crystal clear. So thank you. You can indeed apply this to other things thank you.

– .
Closing remarks

.
– Thank you very much everybody, this was amazing.
I am ironically overwhelmed by all the information I need to process.
– I cried at the opening remarks, I will try not to do that again, hello, thank you again, we would like to take one minute to thank everyone, this is the first time we have done this, we are not event organisers, and we had some problems, but really appreciative for everyone who has helped, and as we mentioned at the start: we will send out the feedback form because who would like to improve, take things on board, and as Ettie Bailey-King mentioned, honesty is how you learn and get better, so now thinking our sponsors ( naming all the sponsors) and our partners for encouraging our nonsense, not being paid to be here, and all the animations that have been shown, all the technical support, and our designer also, even if they could not be here today, an incredible ideas man. Thank you.
– Thank you to all the staff, the panel, especially Molly, already asking questions since January, and really thanking everybody. Those who did by paid forward ticket, also for other people to come, and get lunch, and we had to do this, because otherwise we would be excluding certain people, to make it as accessible as possible, thank you for the content creator, everyone, all the people we have been working with, also Aidan, Lauren, Katie, thank you so much, I feel really overwhelmed by how excellent it has been. Thank you to our 🙂 Human captioner: Veerle Nina Charlotte Reinhilde have been working the entire day – and we have been talking about AI, but we wanted real live captions ( thank you). It felt right. Apologise if we have forgotten someone. We don’t have real media, besides some selfies and a couple of photos. And there is a tube strike in London. Thank you still. We hope you had a good day. We hope you have some energy left. Thank you.
– Goodbye.
– Goodbye from our team!