A young man in a wheelchair cooks pasta at an accessible kitchen counter, illustrating PictureMaker, the new AI picture creation tool in the Photosymbols Library.

Introducing PictureMaker

Introducing PictureMaker

Every so often you search the Library for something very specific, and it simply is not there. PictureMaker is our answer to that moment. Describe the picture you need in your own words, wait about a minute, and get a Photosymbols-style picture back.

A young man in a wheelchair cooking pasta at a lowered kitchen counter, smiling

Made by PictureMaker from the words: "A young man in a wheelchair cooking pasta at a lowered kitchen counter, looking pleased with himself". This picture was generated by AI.

NEW IN THE LIBRARY

Available now to every signed-in Library subscriber.

           

Why we built it

We have spent years photographing real people in real situations, and that catalogue is still the heart of Photosymbols. But no catalogue is ever complete. You need a picture of someone using a specific piece of equipment, or a very particular room, or a combination of things we have simply never shot.

We know this happens because we can see it. Every so often a search comes back empty, and someone goes away without the picture they came for.

PictureMaker is for exactly those gaps. It is not a replacement for our photography. It is what you reach for when the photograph you want does not exist.

Where to find it

Log in to the Library and go to My Favourites. In the row of buttons at the top, next to Upload and Create, you will see a PictureMaker button with a small wand on it.

The My Favourites page, showing the PictureMaker button in the row beside Upload and Create

You will find PictureMaker on My Favourites, in the row next to Upload and Create.

There is also a PictureMaker tile on your account page, which takes you straight there.

If you cannot see the button, the most likely reason is that you are not logged in. PictureMaker is only available to signed-in subscribers.

How it works

There is one box, and that is deliberate. No style menus, no sliders, no settings to get wrong.

1Describe your picture

Type what you want to see in plain English, in the box headed "What should the picture show?"

The PictureMaker page with a description typed into the box headed What should the picture show?

One box. Describe what you want to see, in your own words.

2Press "Make my picture"

It usually takes about 60 to 90 seconds. The screen tells you where it has got to as it works.

The Making your picture screen, showing the stages and a timer counting the seconds

While it works, it shows you where it has got to – and you can press "Tell me when it's done" and carry on.

3Save it

When your picture appears, press "Save to my pictures" and it lands in the AI Created tab of My Favourites, ready to download like any other picture.

The finished picture shown on the Your picture is ready screen, with the description underneath

When it is ready, your picture appears with the words you asked for underneath.

           

Tips and tricks

The difference between a disappointing picture and a good one is almost always the description. Here is what we have learned.

Say more than you think you need to

The hint under the box reads "The more you say, the closer it gets", and it is the single most useful thing on the page. "A man cooking" gives the AI almost nothing to work with, so it will invent the rest. "A young man in a wheelchair cooking pasta at a lowered kitchen counter, looking pleased with himself" gives it a picture to aim at.

Think about who is in it, what they are doing, where they are, and how they seem to feel.

"The more you say, the closer it gets."

Use the suggestion buttons – they are your own searches

Under the box you will see a few suggested phrases. These are not generic examples. Where we can, they are drawn from what you have searched for recently, and we deliberately put the searches that found nothing first.

So those buttons are, quite literally, a list of pictures you wanted from us and did not get. They are the best possible starting point.

Describe the scene, not the style

You do not need to ask for the Photosymbols look. Words like "clear", "simple background" or "Easy Read style" are wasted effort, because the house style is applied for you behind the scenes. Spend those words on the content of the picture instead.

Press "Tell me when it's done" and carry on

A minute and a half is a long time to watch a screen. Press the bell button marked "Tell me when it's done" and you can go back to searching the Library. We will pop up a notification when your picture is ready, and it will wait for you rather than disappearing if you happen to be on another page.

One picture at a time

You can only have one picture being made at once. If you try to start another while one is running you will be asked to wait for it to finish. This is on purpose – it keeps the queue moving for everybody.

Cancel is not quite an undo

Worth knowing: pressing Cancel stops the picture being shown to you, but it cannot reach into the process and stop the picture being made. It will still be created, and it still counts towards your allowance.

So it is better to spend an extra ten seconds on your description than to start, cancel, and start again.

If your description is refused, reword it

Occasionally a description is stopped by the image safety filter. This is usually something innocent being read the wrong way, particularly around health, bodies or distress. Rephrasing in plainer, more matter-of-fact language almost always gets you through.

Try again before you give up

The same words will not give you the same picture twice. If a result is close but not right, adjust one detail and run it again rather than rewriting the whole thing from scratch.

           

The honest limits

We would rather set expectations properly than have you find these out on your own.

  • There is a monthly allowance. It is shared across all the AI picture tools in your account, and it resets on the 1st of each month. We will warn you as you get near it. For normal use you are very unlikely to reach it.
  • It is not photography. AI pictures get hands, text and complicated equipment wrong more often than you would like. Look properly at what comes back.
  • It does not know your organisation. Specific uniforms, buildings, logos or branded equipment are not something it can get right.

The message on the page says it plainly: pictures are made by AI, so always have a look before you use one. That is not a legal disclaimer we are hiding behind. It is genuine advice, and it matters most when the picture is going in front of people who may not be able to tell you it looks wrong.

How this fits with real photographs

We want to be straight about this, because we have been vocal about the value of photographing real people.

Nothing has changed there. Our photographs are of real actors, paid and consented, in real situations, and that is still what we think makes Easy Read work. AI pictures are a supplement for the gaps, not a cheaper substitute for the catalogue.

That is also why anything you make with PictureMaker is kept clearly separate. It saves into its own AI Created tab in My Favourites, so you can always tell at a glance which of your pictures were made by AI and which are photographs. And when you download one, it carries the AI markings we wrote about in our recent blog here, so it stays identifiable as AI-generated wherever it ends up.

Give it a go

Log in, open My Favourites, and press PictureMaker. Start with something you looked for last month and could not find.

Open PictureMaker

And please tell us how you get on. This is new, we are still learning what people actually want from it, and the descriptions that do not work are just as useful to us as the ones that do.