Captions That Follow the Speaker
- Tim Scannell
- Jul 6
- 5 min read
Why the Future of Accessible Communication Is About People, Not Screens
By Tim Scannell, Accessibility Consultant | Deaf Educator | AI Accessibility Researcher
What if captions didn't sit at the bottom of a screen?What if they stayed with the person who was speaking?
For over 50 years, captions have transformed access to television, education, workplaces, live sport and public events. They remain one of the greatest accessibility innovations ever developed.
Yet despite remarkable advances in technology, one thing has barely changed.
Captions still appear in a fixed location, usually along the bottom of the screen.
Every few seconds, we look away from the speaker.
We lose facial expressions.
We miss body language.
We interrupt natural eye contact.
As artificial intelligence, augmented reality and spatial computing continue to evolve, perhaps it is time to rethink not only captions but accessible communication itself.
"Captions That Follow the Speaker"
Figure 1: A vision of future accessible communication where captions follow the speaker, accessibility preferences travel with the individual, and information adapts to people rather than devices.
We've Been Designing Around Technology
Accessibility has traditionally been designed around the limitations of technology.
Television captions had to sit at the bottom of the screen.
Video conferencing software places captions in a single panel.
Presentation software overlays text in one fixed position.
Those solutions made perfect sense when technology offered few alternatives.
Today, AI can recognise speech.
Computer vision can identify who is speaking.
Augmented reality can place digital information into the physical world.
Instead of asking:
"Where should captions appear?"
Perhaps we should ask:
"Where would communication feel most natural?"
Imagine Watching a Basketball Match
Imagine watching the final minutes of a close basketball game.
The coach shouts instructions from the sideline.
Instead of looking down at the bottom of the screen, the words appear beside the coach.
A player replies.
Their response appears beside them.
Another player calls across the court.
Again, the caption follows that individual.
The referee makes an announcement.
His words appear beside the referee.
Rather than moving your eyes between the game and a caption bar, the communication stays with the conversation.
You remain immersed in the match.
You maintain eye contact with the action.
You continue reading facial expressions and body language.
Accessibility supports the experience instead of interrupting it.
Communication Should Stay With the Conversation
Traditional captions tell us what was said.
Future accessibility could also show who said it.
Imagine captions automatically identifying:
Coach
Referee
Player 1
Player 2
Commentator
Each person's words appear beside them in real time.
The conversation becomes easier to follow, particularly in busy environments where multiple people are speaking at once.
Accessibility becomes contextual rather than static.
What Happens If There Is No Audio?
Most speech recognition systems depend on clear audio.
But communication does not always happen under perfect conditions.
Imagine:
A theatre performance where microphones fail.
A sports arena where crowd noise overwhelms speech.
A classroom where someone is speaking across the room.
A noisy railway station.
An emergency announcement with poor audio quality.
Should accessibility simply stop?
Perhaps not.
Future accessibility systems could combine multiple sources of information instead of relying on audio alone.
These might include:
Speech recognition
Visible lip movements
Facial expressions
Speaker identification
Gesture recognition
Context from the surrounding conversation
If audio becomes unavailable, AI could assist by analysing visible lip movements alongside contextual information to improve communication support.
This would not replace professional lipreaders or guarantee perfect accuracy. Lipreading is inherently complex because many speech sounds look identical on the lips, and understanding depends heavily on context.
Instead, it becomes another layer of accessibility.
The future is likely to be multimodal, combining several inputs to provide the most reliable communication possible.
Accessibility Should Travel With You
The real innovation is not smarter captions.
It is personalised accessibility.
Imagine having your own accessibility profile.
Not your identity.
Not your personal information.
Simply the accessibility preferences you choose to share.
Your profile could include:
Real-time captions
Captions that follow the speaker
British Sign Language interpretation
Audio description
Speaker identification
Preferred language
High-contrast display
Larger text
Simplified language
Whether you use smart glasses, a television, a transparent display, a heads-up display or future wearable technology, your preferences remain consistent.
The device changes.
Your accessibility does not.
Accessibility at the Touch of a Button
Accessibility should never be hidden inside complicated menus.
It should be available whenever you need it.
Imagine sitting down to watch live sport, a film, a television programme or a theatre performance streamed into your home.
You simply press the Accessibility button on your remote control, television, smart glasses or wearable device.
A personalised accessibility menu appears.
You choose exactly what works for you.
Perhaps you select:
✓ Captions
✓ Captions that follow the speaker
✓ British Sign Language interpretation
✓ Audio description
✓ Speaker identification
✓ Preferred language
✓ High-contrast display
✓ Larger text
Another person may choose only audio description.
Someone else may choose sign language without captions.
Another may combine captions, speaker identification and multilingual translation.
There is no single correct solution.
Accessibility is personal.
Once selected, those preferences automatically apply to compatible services and devices.
No searching through complicated settings.
No repeatedly turning features on and off.
Simply press one button.
Choose your accessibility.
Enjoy the experience.
This Isn't Just About Deaf People
Although my work focuses on Deaf communities and British Sign Language, this vision extends much further.
Imagine:
A Deaf person choosing captions and BSL.
A blind viewer selecting audio description.
Someone with dyslexia can use simplified language.
A person with low vision increases text size and contrast.
An international visitor receives information in their preferred language.
An older adult choosing larger text for easier reading.
The information remains the same.
Only its presentation changes.
That is inclusive design.
Where Could This Work?
This concept extends far beyond basketball.
Imagine accessible communication in:
Live sporting events
Football stadiums
Concerts
Theatre performances
Business meetings
Universities
Schools
Healthcare appointments
Public transport
Airports
Museums
Conferences
Government services
Everyday conversations
Accessibility becomes part of the environment rather than something people have to configure every time.
Technology Is Only Half the Challenge
Technology is advancing rapidly.
The bigger challenge is trust.
Questions remain.
Who stores accessibility preferences?
Who controls them?
How does consent work?
Can users choose exactly what they share?
Can organisations adopt common accessibility standards?
How do we ensure people without the latest technology are not excluded?
These questions are every bit as important as AI itself.
From Universal Design to Personal Accessibility
Universal Design encouraged us to create environments that work for as many people as possible.
The next evolution may be something even more personal.
Rather than designing one experience for everyone, systems could adapt information to each individual while respecting privacy, dignity and choice.
Accessibility becomes:
Personal
Portable
Intelligent
Context-aware
Device-independent
It belongs to people - not products.
Looking Ahead
This article is not a prediction about one company or one device.
It is a proposal for a design principle.
As technology evolves beyond smartphones into smart glasses, transparent displays, spatial computing and technologies we have yet to imagine, accessibility should evolve with it.
Communication should stay with the speaker.
Information should adapt to the individual.
Accessibility should not depend on a single screen, a single device or a single source of information.
Whether communication comes from speech, sign language, visible lip movements, audio description, gesture recognition or future technologies we have not yet imagined, people should receive information in the way that works best for them.
The future of accessibility is not about creating one solution for everyone.
It is about giving everyone the freedom to choose their own.
One profile.
Your accessibility.
Your choice.
Any device.
Anywhere.
About the Author
Tim Scannell is an Accessibility Consultant, Deaf Educator and AI Accessibility Researcher. His work explores the future of inclusive communication, accessibility governance and artificial intelligence, with a focus on creating technology that adapts to people rather than expecting people to adapt to technology.
Comments