Translation giants like Google and Microsoft are faced with a powerful but small competitor.
The startup from Cologne, DeepL, was founded by a former Google employee. It provides a translation tool that produces more naturally worded text than any of its big competitors.
The service uses so called deep learning technology based on artificial neural networks. But how exactly does it work?
From dictionary to translator
Before the company launched the new translation tool in August 2017 it had been in the "language business" for quite some time. In 2009, they started a a dictionary called Linguee. What was special about it? Apart from various translation options, it offered samples of already translated pieces of text in chosen language pairs.
It gathered these text samples from the Internet using special Internet bots that help web indexing called web crawlers. Many text samples came from the documents of multilateral institutions like the European Union (EU).
A human-trained machine-learning algorithm then computed the structure and meaning of the gathered text and learned how certain ideas are being expressed in another language. Furthermore, the users can leave feedback and rate the translations manually. Over nine years Linguee.com has improved by internalizing an ever growing number of high quality translations. It has thus became the base for a new translation tool.
DeepL stands for 'deep learning'
Deep learning is an aspect of artificial intelligence (AI) that wants to replicate human learning. In the past, programmers used to write individual commands for every task a computer had to do. Now they are constructing an artificial neural network that can be trained by showing the computer existing examples. The more examples the better the computer can learn. Big data further amplified these opportunities for machine learning.
On the technological side, deep learning is an algorithm that uses multiple layers of processing information. "Deep" refers to the number of layers through which the data is transformed. Today, this technology operates self-driving cars, virtual reality headsets and face recognition software. And now it also helps us with translation.
Read more: Germany launches digital strategy to become artificial intelligence leader
It's more than just deep learning
DeepL isn't the only one to use deep learning technology for the past two years. Internet giants like Google, Microsoft and Yandex are also using it. But what makes DeepL different?
First of all, the performance of the neural network is based on the quality of the input material used for training. And here the "big brother" Linguee comes into play, as it provided billions of high-quality translation samples.
"The Linguee data is very high-quality training material," said company spokesman Lee Turner Kodak.
It is not only the quality of the data that made DeepL possible.
"Our researchers are well-versed in the latest developments in the field and have put together a unique neural network architecture," he stresses.
Exactly how the technology is put together remains a secret, however. He does not want to lose his market advantage.
Three times more powerful than Google Translate
To control the translation quality, DeepL runs regular blind tests to make sure the program keeps its high standards. The company backs its claim of being "the world’s best translation machine" with the results of a August 2017 blind test published on its website. The blind test compared DeepL's results with those of translators by their rivals: Google, Microsoft and Facebook.
The machines were given 100 sentences to translate from English into German, French and Spanish and from these three languages into English. After the experiment, professional translators looked at the produced text and rated the quality of the translation. DeepL was chosen three times more often than Google because its translation sounded more natural.
Read more: Artificial intelligence in medicine: The computer knows what you need
More dangerous than the atomic bomb?
Silicon Valley icon Elon Musk, founder of SpaceX and Tesla Motors, is famous for his warnings with regard to Artificial Intelligence. Last summer he declared that AI was the greatest conceivable threat to our existence. Stephen Hawking isn't new to the discussion, either, calling it famously the "worst mistake ever made."
Aren't AI robots more helpful than harmful like in the recent Hollywood film Chappie? A reprogramming gives the robot feelings and thoughts, and he helps humanity against an aggressive robotic police force.
High-speed drone flop
Recent occurrences have shown, however, that not every instance of AI is without fault. All it took was nine minutes for the Falcon HTV-2 to sink in the ocean on a test flight in the summer of 2011. The US military drone was unmanned.
Not really all that new
Despite the resurgence, AI in military systems is a foregone conclusion. For over two decades, machines and robotic components have been advancing military systems. One prime example - the Eurofighter.
Sci-Fi meets reality
The intelligent machines are getting more and more advanced - in many cases operational. The four-legged robot BigDog can haul cargo on offroad terrain, ice and snow. The robotics developer Boston Dynamics was bought by Google.
This week's letter has made clear that the entire AI community is seeking ethical guidelines, and even political regulation, to ensure that standards are set for how machines can be programmed. This is the only way to prevent abuse of Artificial Intelligence - and to put a stop to it getting out of control.
Investment from the Silicon Valley
On December 2, DeepL announced that the Cologne-based startup had won a new investor — the Silicon Valley venture capital firm Benchmark. With a team of 25 employees, DeepL now plans to double its research and development within a year, says Kodak.
"This partnership represents not only money, but connections. We will be able to learn from the business intelligence of the investor and other startups that are further along in the development process," he said.
100 languages against nine
On December 5, DeepL launched new translation services in Russian and Portuguese. With these two languages, DeepL now offers nine European languages and 72 language combinations in total. Google Translate knows more than 100 languages. But the startup does not want to catch up too soon. Quality is more important than quantity, the founders believe.
"We want to have the best machine translation system available in each language. So rather than bringing out a couple of hundred languages just to have them available, we train our networks to deliver the same high quality that we have become known for," Kodak said.
Read more: Artificial intelligence, or the end of the world as we know it
AI: 'Third revolution in warfare'
Over 100 AI experts have written to the UN asking them to ban lethal autonomous weapons — those that use AI to act independently without any human input. No "killer robots" currently exist, but advances in artificial intelligence have made them a real possibility. The experts said these weapons could be "the third revolution in warfare," after gunpowder and nuclear arms.
The "first revolution in warfare" was invented by the Chinese, who started using the black substance between the 10th and 12th centuries to propel projectiles in simple guns. It gradually spread to the Middle East and Europe in the following two centuries. Once perfected, firearms using gunpowder proved to be far more lethal than the traditional bow and arrow.
The invention of gunpowder also introduced artillery pieces to the battlefield. Armies started using basic cannons in the 16th century to fire heavy metal balls at opposing infantrymen and breach defensive walls around cities and fortresses. Far more destructive field guns were invented in the 19th century and went on to wreak havoc in the battlefields of World War I.
Guns that fire multiple rounds in rapid succession were invented in the late 19th century and immediately transformed the battlefield. Machine guns, as they came to be known, allowed soldiers to mow down the enemy from a protected position. The weapon's grisly effectiveness became all too clear in WWI as both sides used machine guns to wipe out soldiers charging across no man's land.
Military thinkers did not ignore the invention of the first airplane in 1903. Six years later, the US military bought the first unarmed military aircraft, the 1909 Wright Military Flyer. Inventors experimented with more advanced fighter and bomber aircraft in the following years. Both became standard features in many of the national air forces established by the end of WWI.
Armies had traditionally used soldiers and horses to fight and transport military equipment. But around WWI, they started using more machines such as tanks and armored vehicles. Faster and more destructive armies were the result. Nazi Germany put this new form of "mechanized warfare" to destructive effect in WWII using an attack strategy known as "Blitzkrieg" ("lightning war").
Although artillery was effective, it had a relatively limited range. The missile's invention in WWII suddenly allowed an army to strike a target hundreds of kilometers away. The first missile — the German V-2 — was relatively primitive, but it laid the foundation for the development of guided cruise missiles and intercontinental ballistic missiles (ICBM) capable of carrying nuclear warheads.
Jet aircraft first saw action alongside traditional propeller airplanes at the end of WWII. Jet engines dramatically increased an aircraft's speed, allowing it to reach a target quicker and making it far harder for an adversary to shoot it down. After WWII, military reconnaissance planes were developed that could fly higher than 25 kilometers (15.5 miles) and faster than the speed of sound.
The "second revolution in warfare" announced its horrific arrival on August 6, 1945 when the US dropped the first nuclear bomb — "Little Boy" — on the city of Hiroshima in Japan, killing between 60,000 and 80,000 people instantly. In the Cold War that followed, the US and Soviet Union developed thousands of even more destructive warheads and raised the specter of a devastating nuclear war.
Recent decades have witnessed the ever more prevalent use of computers to conduct war. The devices made military communication quicker and easier and radically improved the precision and efficiency of many weapons. Armed forces have recently focused on developing cyber warfare capabilities to defend national infrastructure and attack foreign adversaries in cyberspace.
What you need to know