Autra—automatic content translation.
How can we encourage better communication between diverse parties using technology?
Not One the Same
We’ve all been in the scenario where we are trying to communicate something to someone. Either they don’t understand what we’re intending to say or you find yourself pausing to try and put it in their terms. While the latter is very considerate of you, it isn’t always clear.
Maybe you’re messaging a coworker in a different time zone trying to coordinate a video call. Or maybe you want to let a family member from Europe know what the forecast will be for their visit. Or let’s say you want to share a song with a friend but you both don’t have Spotify.
More and more we find ourselves embracing our individuality in our preferences of content ingestion. And this is great, we should be who we are meant to be and encourage others to do the same. So I ask myself,
“How can we encourage better communication between diverse parties using technology?”
What I imagine is the ability to communicate in your preferred way, and the recipient of your message would receive it in their preferred way.
The way I see it, there are a handful of avenues to translate as we communicate.
- Language (English, Spanish Italian, etc.)
- Time Zone (MST, EST, etc.)
- Time Format (am/pm vs 24hour)
- Date Format (MM/DD/YY vs DD/MM/YY)
- Temperature (Fahrenheit vs Celsius)
- Currency (USD, EUR, etc.)
- Measurement (English vs Metric)
- Music Links (Spotify, YouTube, Apple Music, etc.)
Examples in the Wild
Some technologies exist that help nurture this concept of automatic translation.
What all of these technologies are doing similarly, is understanding one intention, and automatically translating that intention to communicate a different message. This means users can more easily communicate in their preferred format recipients can receive the message in their preferred format.
For example, with Google Music Search, I searched for an artist name and song title and it returned a list of different streaming platforms I can choose from to listen to that song. If Google knew where I always listen to music, maybe it could have returned that as a higher level result.
An Integrated Solution
If we could build into all of our messaging platforms a seamless way to automatically translate our messages to reflect our recipients preferences, we could more easily communicate with one another. This would support the ability to receive a message, knowing it’s communicating what you need to know, allowing the sender say the message that makes the most sense for them.
What I commonly see is a coworker send a request to meet at 11am MST, or saying “1pm your time”. Or a friend send me a song link from iTunes, but I stream music through Google Play Music. So what I mean by this automatic translation technology is this coworker can just send “11am” and I’ll receive it as 13:00. Because not only am I two hours ahead of them, but I also prefer to tell time in the 24hour format. Or when my friend shares a link from iTunes, or even just sends an artist name and song title, I would receive it as a direct link to the song in Google Play Music, giving me the ability to open it directly in the native app.
Here is an example of what this could look like in a messaging format.
Using a highlight allows you to visualize what has been translated, and having the ability to see the message untranslated supports any need for clarification.
Real Life Applications
So what does this look like applied to popular forms of digital communication. Well, let’s look at a few mockups.
Here is an example of two friends sharing music with one another. The song link is translated from YouTube to Spotify.
Just like sent messages show a delivery status, translated messages from the sender display “Translated”, indicating that the bold text has been translated.
To show the original message, we utilize the pull from left functionality. When this action is activated, a color and text mask pull onto the screen from the left. The pull from left does not require the user to keep pressing down while pulling. This way they can access the original link if need be. In order to get back to the translated message, a push left is required, reflecting what happened when revealing the original message.
Here is an example of someone from the US asking a family member in Switzerland about the measurement of temperature and distance. The temperature is converted from Celsius to Fahrenheit while the distance is converted from meters to feet.
The translated message displays a “TRANSLATED” text in the bottom left corner of the text box, notifying the recipient that the message has been translated per your content preferences. To display what has been translated, there is a subtle highlight behind the text on the message.
To show the original message, a user can follow the first-time instructions next to the “TRANSLATED” text to double-tap the message box. In Whatsapp, the action of double-tapping a message box is registered but not used. Doing so would turn that box to a teal color used in Whatsapp branding displaying the original message with what was translated more clearly. Getting back to the translated message requires another double-tap.
Here is an example of an exchange of emails regarding food prices in another currency. The original message using Baht is converted to the current exchange rate of USD.
The text that has been translated is using a familiar Google style of a rounded light grey pill background with bold text to indicate this text can be interacted with. On hover, a familiar Google styled tooltip will appear displaying the original message. That will disappear as the user moves their cursor off of the highlighted text.
Here is an example of a couple coworkers setting up a video call across different time zones. Not only is the actual time of the meeting translated to the recipients’ time zone, but the recipient also prefers to read time in 24hour format so that is also converted.
Similar style and format to the “edited” text that appears next to the message after it has been edited, “translated” appears when a message has been translated. The “translated” text is actionable and can be clicked on. When clicked, the original message will be shown.
To show the translated message again, the user can click on “original message”. Another way that the preferred message will show again is if the user navigates away from this conversation and comes back.
The Adoption Process
Maybe this could work, but how?
1. AI or Machine Learning (ugh buzz words)
Build an infrastructure that enables products to integrate a technology allowing for this type of automatic translation. Instead of the user setting their preferences, a universally connected system could learn what a users preferences are from app usage, messages sent and received, and location based data.
2. Translation as a Service
Build an API that gives products wanting this technology the ability to integrate it into their systems. Users would subscribe to a third party app where they set their preferences and connect their different accounts, be it Facebook, Slack, email address, phone number, etc. The platforms that users would notice the translation occurring would be the companies who pay to use the API, being the service partners.
3. Your idea?
If you have any other ideas on how to make this technology a reality, reach out, I’d love to hear about it.
Maybe both 1 and 2 could be the same option. Whatever the right technological solution is for product and system adoption, there is a constant that remains important. That is, to build and supply a system for how each different product can seamlessly integrate these underlying functionalities into their individual interface. The goal would be to allow for this type of communication, but not sacrifice on the existing experience of each different product and operating system.