AW: Breaking barriers

[Mark E. Shoulson]

On 10/22/21 12:17, Asmus Freytag via Unicode wrote:
> On 10/21/2021 3:40 PM, Mark E. Shoulson via Unicode wrote:
>> If I recall correctly, someone has proved that "fully automatic 
>> high-quality translation" is AI-hard.  Meaning that it's basically 
>> the same as making a fully aware, human-intelligence AI.  Now, that 
>> probably depends a lot on the details of "high-quality." There are 
>> probably sentences and texts one could cook up that a would-be 
>> translator would need arbitrarily good understanding of the context, 
>> situation, shared cultural memories and references, etc etc for, and 
>> I guess that would be what the "proof" was about. 
>
> Sentences that require some understanding of the meaning for a 
> successful translation, even if you only consider factual accuracy, 
> are not hard to come by: they do prop up regularly.
>

Yeah, you're right.  I was wrong to imply (or think) that it only 
mattered in rarefied corner cases.  You give some fun examples of 
languages that don't mesh because they encode different information, and 
I'm sure a lot of us could come up with more. That makes any kind of 
language-independent representation difficult or impossible—if used or 
envisioned as a translation intermediate or codes "equivalent" to some 
sentence (because sentences may not be capable of being equivalent.)  
You can use it on its own to express concepts in its own way, but at 
that point it isn't a translation intermediate, nor even 
language-independent, but is a language in its own way (see 
Blissymbolics, which, fair warning, I really hardly know anything about, 
so maybe you shouldn't see them.)

>
>
>> It seems to me that that does have farther to take us, and we'll 
>> probably see a lot more improvement, but it can only take us so far.  
>> Then again, "so far" might be far enough.  If you have a translator 
>> whose results are semantically satisfactory, say, 97% of the time, 
>> and sound only a little awkwardnessful to a native speaker in the 
>> target language... well, customers' standards may be willing to duck 
>> a little.
>
> There's a level of "quality" that equates to "a human looking at the 
> translation can guess what might have been in the original".
>

And it is over-optimistic to expect the level I expressed any time soon, 
yes.

Thanks!