The Shallowness of Douglas Hofstadter.

I am not in the least eager to see human translators replaced by inanimate machines. Indeed, the idea frightens and revolts me. To my mind, translation is an incredibly subtle art that draws constantly on one’s many years of experience in life, and on one’s creative imagination. If, some “fine” day, human translators were to become relics of the past, my respect for the human mind would be profoundly shaken, and the shock would leave me reeling with terrible confusion and immense, permanent sadness.

If (or more likely, "When") computers translate better than human, respect for the human mind should increase. It's quite a tough problem to solve, way harder than manual translation which thousands of people have been doing for centuries.

I believe Hofstadter is more worried about people prematurely adopting this technology en masse simply because, by it's inherently impressive nature and mysterious inner-workings, it generates its own terrific PR which clouds the perception of its actual efficacy at a deeply cognitive task.

Kevin Kelly in his book What Technology Wants speaks to the same willingness for human societies to so eagerly jump into new, unproven or potentially dangerous technologies without ever stopping to consider the effects of the adoption and log-term usage on their societies. A whole chapter is about how the Amish culture is not necessarily anti-technology so much as they are slow, disciplined adopters of technology. They hold their people's culture above all other values including productivity or connectivity and so to them it's a clear decision to at first be skeptical of new technologies and to deliberate the adoption of these technologies until the elders of the community can come to a consensus. Even then they usually allow small "pilots" or testing of the technology with a select few individuals at first. Kelly doesn't suggest we adopt the Amish culture, simply that there is something to be learned here.

I think A.I would need to become a fully robust form of consciousness capable of generating its own novel ideas spontaneously and to form a sort of gestalt from a mix of information, logic, patterns but also "feelings" (or whatever term would be used for "feelings" in a conscious AI) in order to produce the same type of translation a human can. Granted, even without consciousness it will likely come extremely close over time with more training and improvements to the algorithms but as Hofstadter explains it takes more than a sort of algorithmic proficiency:

"..To me, the word “translation” exudes a mysterious and evocative aura. It denotes a profoundly human art form that graciously carries clear ideas in Language A into clear ideas in Language B, and the bridging act not only should maintain clarity, but also should give a sense for the flavor, quirks, and idiosyncrasies of the writing style of the original author. Whenever I translate, I first read the original text carefully and internalize the ideas as clearly as I can, letting them slosh back and forth in my mind. It’s not that the words of the original are sloshing back and forth; it’s the ideas that are triggering all sorts of related ideas, creating a rich halo of related scenarios in my mind. Needless to say, most of this halo is unconscious. Only when the halo has been evoked sufficiently in my mind do I start to try to express it—to “press it out”—in the second language. I try to say in Language B what strikes me as a natural B-ish way to talk about the kinds of situations that constitute the halo of meaning in question."

I predict we will prematurely adopt machine-learning powered translation at the cost of loosing depth, clarity and the richness of the human expression of ideas but for the majority of us the impact may be minimal especially if great human literatures are still translated by humans.

I think it's just as much that Hofstadter doesn't want people making overstated claims giving a bad name for AI research (again!) when people finally realize how limited the current approaches are.

my longstanding belief that it’s important to combat exaggerated claims about artificial intelligence.

I think Google translate is pretty amazing, but yeah, there's no need to declare game over. We are only beginning.

Overestimating progress goes back to the early days of computers. Seeing impressive results, people think the computer must possess something close to human intelligence on some level, because a human would have to be pretty smart to perform that well.

But no, with computers with have idiot savants.

Yeah, with all the people making ridiculously overstated claims about what deep learning can do I'm pretty sure we've got another AI winter coming.

And it's a shame. I really want to see more genuine progress in AI research; I really want to understand _what consciousness is_. But this boom/bust cycle that happens every time there's a tiny bit of real progress is a painfully inefficient way of getting there.