December 23, 2016
On December 14, 2016, the New York Times published a piece by Gideon Lewis-Kraus about the recent improvements to the Google Translate application and the implications it has for the future of artificial intelligence. Google Translate is both a standalone application and a tool utilized in a number of other Google products like Maps and their standard search engine. It can, as you might imagine, take one word, sentence, or passage in one language and produce that same content in another language of the user’s choosing. It is the real-life version of such sci-fi concepts as Star Trek’s Universal Translator or the Hitchhiker’s Guide to the Galaxy’s Babel Fish.
The longstanding problem with the application was that it wasn’t very effective. There always seemed to be odd conjugations, missing articles, or curious phrasing, but after the recent update, that is a thing of the past.
A Side by Side Comparison of Google Translate
To showcase the advances in the application, the Ney York Times article placed two selections of text from Ernest Hemmingway’s “The Snows of Kilimanjaro.” One passage was the original English text, and another was the same copy translated to Japanese and then back into English. Lewis-Kraus challenges his readers to guess which is the translated passage, and it proved harder to discern than expected.
Passage 1
Kilimanjaro is a snow-covered mountain 19,710 feet high, and is said to be the highest mountain in Africa. Its western summit is called the Masai “Ngaje Ngai,” the House of God. Close to the western summit there is the dried and frozen carcass of a leopard. No one has explained what the leopard was seeking at that altitude.
Passage 2
Kilimanjaro is a mountain of 19,710 feet covered with snow and is said to be the highest mountain in Africa. The summit of the west is called “Ngaje Ngai” in Masai, the house of God. Near the top of the west there is a dry and frozen dead body of leopard. No one has ever explained what leopard wanted at that altitude.
Were it not for the missing article before leopard and the clunky phrasing of “Near the top of the west” in Passage 2, it could certainly pass for the original. Now compare this passage to what Google Translate could produce before the update.
Passage 3
Kilimanjaro is 19,710 feet of the mountain covered with snow, and it is said that the highest mountain in Africa. Top of the west, “Ngaje Ngai” in the Maasai language, has been referred to as the house of God. The top close to the west, there is a dry, frozen carcass of a leopard. Whether the leopard had what the demand at that altitude, there is no that nobody explained.
Clunky, awkward, and unnatural, this original translation could never fool a native speaker. It is largely incoherent and difficult to follow. The level of improvement found in the new Google Translate is astounding.
Google Translate and the Turing Test
In 1950, Alan Turing developed a test of a machine’s ability to exhibit intelligence, and it serves as a benchmark for advances in artificial intelligence. The test involves separating participants in individual rooms and asking them to engage in a text conversation with another participant. One person in each conversation will be human, the other will be a computer designed to imitate human responses to questions. If the machine can fool 70% of humans after five minutes of conversation, it has passed the test.
Since Turing developed his test, there have been a handful of programs that have passed, but there is a significant distinction between those chatterbots and the new capabilities of Google Translate. The chatterbots that have passed the Turing test use algorithms to identify keywords to create sentences relevant to the conversation and deploy stock generic answers to questions they cannot analyze. By contrast, Google Translate paves the way for applications that may actually understand the question in a much more human sense. Their ability to articulate language and organize thoughts into natural sentence structures is beyond anything we have seen before.
The All or Nothing Game of Artificial Intelligence
In a recent TED Talk, neuroscientist and author, Sam Harris described the key danger of artificial intelligence over the long haul. The end goal of developments to artificial intelligence is to eventually create a super intelligence that can learn on its own and act independent of human users. It is truly the Holy Grail of all technology, but it comes at great risk. Harris explained,
“So imagine if we just built a superintelligent AI that was no smarter than your average team of researchers at Stanford or MIT. Well, electronic circuits function about a million times faster than biochemical ones, so this machine should think about a million times faster than the minds that built it. So you set it running for a week, and it will perform 20,000 years of human-level intellectual work, week after week after week. How could we even understand, much less constrain, a mind making this sort of progress?”
While that future sounds precarious to say the least, the risk is worth it. The potential upsides are almost incomprehensible and stand to change the fabric of human society forever. Harris goes on to say,
“So imagine we hit upon a design of superintelligent AI that has no safety concerns. We have the perfect design the first time around. It’s as though we’ve been handed an oracle that behaves exactly as intended. Well, this machine would be the perfect labor-saving device. It can design the machine that can build the machine that can do any physical work, powered by sunlight, more or less for the cost of raw materials. So we’re talking about the end of human drudgery.”
Artificial intelligence over the next fifty or one hundred years is an all-or-nothing game, but even today we can see the baby steps our technology is already taking that will pave the way for the super intelligent AI of our grandchildren and great grandchildren’s generation.
Like what you read?
About Mindsight
Mindsight, a Chicago IT services provider, is an extension of your team. Our culture is built on transparency and trust, and our team is made up of extraordinary people – the kinds of people you would hire. We have one of the largest expert-level engineering teams delivering the full spectrum of IT services and solutions, from cloud to infrastructure, collaboration to contact center. Our highly-certified engineers and process-oriented excellence have certainly been key to our success. But what really sets us apart is our straightforward and honest approach to every conversation, whether it is for an emerging business or global enterprise. Our customers rely on our thought leadership, responsiveness, and dedication to solving their toughest technology challenges.
Contact us at GoMindsight.com.