Mark Zuckerberg announced new developments Meta’s working on for its metaverse initiative — including a universal speech translator. The AI powering it would give us, “The ability to communicate with anyone in any language,” — in real time, just by speaking.
That’s tech magic from about 20 different sci-fi franchises.
It makes busting language barriers so easy that we would easily trade good conversations, genuine connection, and our 2nd favorite child just to use it.
We’ve done it before with tech. We’ll do it again.
But maybe it could be a lesson in the right way and the wrong way to use disruptive tech to improve our lives.
If a new technology can do 80% of a job well, we’ll happily forget about the other 20.
Texting beat out calling. Video calls beat in-person meetings.
We might lose the option of forming long thoughts by texting someone. We might lose the option to throw someone out a window when we’re not there in person. But we trade away that 20% for convenience.
Even if we lose something crucial by switching to new, more convenient technologies, it’s a good bet that we’ll make that trade.
What is that 20% if we use a universal translator instead of learning a language?
It’s all the important stuff.
All of the local expressions, the sarcasm, the ways you can describe someone’s mom when you’re verbally assaulting each other.
It’s also the culture and subtlety behind that.
The AI that Meta is working on is a data-eater like any other. It pulls from mountains of data to learn how to hear and speak a language. From that, it learns common phrases and even basic topics quickly.
But it likely won’t ever be able to access enough data at once to keep up with how language evolves.
New expressions pop up all the time in every modern language. A lot of the time, their meanings have nothing to do with the literal words. It would take an AI ages to pull together enough context to understand.
And then there’s all of the really tricky stuff.
The nuance that machines can’t ever understand. The way we express sincerity and emotions. If any AI can help you crack a joke in a language you don’t understand at all, I’ll eat my words with a shovel. But robots just aren’t that funny.
A universal translator would be a miracle. What’s the right way to use a miracle?
We’re all members of the church of Data. So a universal translator will become an addiction.
It can crunch data to translate so much faster than we can. It can build connections between anyone, and the benefit is instant. Plus it doesn’t rot your teeth half as much as most addictions (looking at you, crystal meth).
But clearly if we use it without ever learning a language, we would be trading away deep connections and understanding for convenience.
You avoid that by taking the same approach with any disruptive tech.
Instead of letting it completely take you over on its own terms, you identify its role.
A universal translator can be a highway to actually learning languages. With it, you should be comfortable going anywhere in the world without the worry that your knowledge isn’t enough. You always have a backup plan.
At the very least, you can use it to cut out 80% of the studying you would need to do. Let it handle everything easy, and use your time to learn the way people in that culture talk. Then mold your native tongue and the expressions you use to their norms.
You can enjoy to full benefits of any technology, limit the drawbacks, and end up only enhancing yourself.
That’s not a given — it’s an active choice.
