By 2040, will there be audio translation tech that's a full substitute for knowing a foreign language?
11
56
250
2040
89%
chance

This question will resolve to YES if there's some sort of an audio translation technology that gives the user a comparable experience to being a fluent speaker in a foreign language. The tech should meet the following criteria:

  • Can be reasonably worn by an adult person, so <0.5 kg in weight

  • Works in areas with poor or no internet connectivity

  • Works in situations with lots of external noise being present, such as a busy restaurant or a ferry ride, not just in perfect lab conditions

  • The tech can use just microphones alone or a combination of audio and video

  • The translation quality should be so good that one could reasonably go study in a foreign country or conduct negotiations in a foreign language by just relying on it.

  • The translation should work in both directions (speaking and listening)

  • The device should work even if only the primary user is wearing/using it, it cannot require everyone present to own said device

  • Works for real-world, imperfect speech, not just for fully articulated speech without any mistakes. If a fluent speaker can make sense of it, so should this tech.

  • Its sufficient if it works for just one language pair (i.e. English-Spanish or English-Mandarin)

The final experience should somewhat resemble getting an instant audio-feed from a professional interpreter at the UN, for both speaking and listening.

Question inspired by this tweet:

Get Ṁ200 play money
Sort by:

theres too many benefits to learning multiple languages. being able to think in multiple languages allows a lot especially when there isnt a clear connection btwn concepts/words in the language.

it also makes swearing much easier

Note: ChatGPT can do this currently if you provide it with an accurate transcript, but not in real time. No tech can do this in real time for speech recorded in noisy conditions and spoken by the average person.