Chatbots, AI, and Awaiting the Resurrection of Microsoft's Tay

Almost two years ago we dedicated a post to Eugene Goostman, a chatbot that had passed the Turing Test. At the time, we noted that the way the test was conducted seemed a little off and skewed. Nevertheless, we certainly thought that Eugene Goostman's achievements were worth celebrating.

This eagle was not pleased with Tay's behaviour.

Nearly a month ago on 23 March, Microsoft had a day from hell due to the complete debacle involving Tay, their new AI chatbot that took to Twitter and survived only 16 hours. In less than a day, users of the platform managed to corrupt the once innocent machine to a point where her tweets were so inflammatory that Microsoft had to close the AI's Twitter account.

Obviously, the Tay's "achievements" are a bit harder to celebrate than Eugene Goostman's. That said, while Microsoft's AI was spouting all sorts of racist and sexist messages, it was Twitter users who taught and raised it.

However, not all Twitter users are responsible for Tay's behaviour. In fact, Microsoft has used Chinese and Japanese chatbots for a couple of years now, and neither of them have caused any major problems. Maybe it's just that English speakers are worse when it comes to internet behaviour. Who knows?

Personally, I think that the internet should be given a second chance to try and raise its AI baby. However, I think Microsoft probably needs to ground Tay first and teach her a lesson. How did you react to Tay? Did you get over it? Would you like to see more experiments involving AI and chatbots on social media? Tell us your opinions in the comments below!

Jiminy Jillikers and Semantic Satiation

There's a fantastic episode of The Simpsons called "Radioactive Man" in which a film adaptation of the superhero comic Radioactive Man is being made in the town of Springfield. In the episode, Bart's friend Milhouse is cast as Radioactive Man's sidekick, Fallout Boy.

The episode covers how Milhouse deals with being a movie star, and how it isn't as cool as he thought it would be. After being forced to record each scene multiple times and say Fallout Boy's catchphrase "jiminy jillikers" hundreds of times, Milhouse retorts, "making movies is so horribly repetitive; I've said 'jiminy jillikers!' so many times the words have lost all meaning!"

Try saying "flower" 100 times...

Of course, "jiminy jillikers" is a fictional expression and has no meaning. However, the phenomenon Milhouse is referring to is very real, and is called semantic satiation. This is when a word is repeated so many times that you no longer understand it to have any meaning, and instead just imagine the word as meaningless sounds, nonsense, gibberish, etc.

No matter what the word is, if you repeat it enough times, you'll eventually stop understanding it as a word and start hearing it as little more than the sounds (or phonemes) that constitute the word.

Normally when you repeat a word, your brain triggers the meaning of the word and you therefore understand it. However, when you quickly repeat a word again and again, you trigger a process known as reactive inhibition, which reduces the effectiveness of repeating the word, effectively rendering the process almost null and void. This means that rather than triggering the meaning of the word, you become almost immune to the process, and no meaning is triggered.

While this episode of The Simpsons plays on this idea by using a meaningless word as an example, it still helps explain the concept. You've surely experienced semantic satiation in your lifetime, and if you didn't know what it was before, now you do!

Why English Words Are So Inventable

If there's one thing I love more than anything about the English language, it's its flexibility. While English is a Germanic language, a lot of its vocabulary comes from Latin and French. Furthermore, thanks to its history, it has borrowed plenty of words from languages all over the word and never thought about giving them back.

Aside from all that, English can be pretty chill when it comes to its rules. Many English speakers like to bend the rules, and sometimes when they break them completely it doesn't even matter (as long as you're not in the company of prescriptivists).

There are as many possibilities for English words as there are
stars in the sky.

However, today I want to talk about lexical flexibility, the ways new English words are invented, and how you can invent them yourself. Of course, other languages invent new words all the time too, but it's the ways in which English does it that I'll be focusing on today.

Those in the UK might remember the story of bouncebackability, a word that described a sport team's capacity to recover from a losing position or setback. The word's first use was attributed to Iain Dowie, former manager of the Crystal Palace football team. While he is believed to have just invented the word on the spot, if you're familiar with the English language and some of our lexical tropes, you can immediately work out its meaning.

I love these kinds of examples of English being used to its full potential. That's why today I thought I'd look at a few of the ways you can invent your own English words and still be understood.

Suffixing and Prefixing

Adding a suffix to a preexisting word is one of the best ways to create new words. The suffix -ise (and -ize in American English) is used to mean "render" or "make". For example, veganise would mean to make something vegan. You can also add the -y suffix to indicate that something has a certain quality. Even if the adjective doesn't exist, you can always create one by adding -y to the end of any noun.

Prefixes can work the same way. You can add prefixes like anti-, un-, in-, and im- to create negatives, or pre- or post- to mean "before" and "after", for example.

Verbifying

You can also create new verbs from nouns by treating them like regular English verbs. The most famous recent example is probably the verb to google. It may seem commonplace now, but you should remember that this is a relatively recent idea that only gained traction in the last decade.

Portmanteauing

Creating a portmanteau is also another way English likes to create new words. For example, you can put together fare and forecasting to make farecasting, the act of predicting the best time to buy plane tickets.

Some of my favourite modern examples are those for male beauty products, procedures, and cosmetics. Words like guyliner (a combination of guy and eyeliner), manscara (from man and mascara), and manscaping (from man and landscaping, which refers to how a male can trim or remove his hair or improve aesthetics).

Awesomesaucing

Of course, you can also just put two words together, like awesome and sauce to make awesomesauce. The possibilities are endless!

What are some of your favourite neologisms that have come about from these kind of behaviours? Tell us about them in the comments below.

Why Medical Interpreters Are So Awesome

After spending last week in a Spanish hospital, I was reminded of the importance (and awesomeness) of some of my favourite language professionals, interpreters, and above all, those who specialise in medical interpreting.

First, I should say that I am in awe of all interpreters in general. If you've ever tried any kind of interpreting, it won't take long before you realise that it's exhausting and very mentally demanding. These professionals work incredibly hard and have an enviable set of language skills.

However, rather than talk about all interpreters today (who I respect greatly), I'd like to focus on medical interpreters.

Seeing doctors and being in hospital can often be a distressing, worrying, or generally negative experience because you don't often spend time with medical professionals when you're feeling great.

Of course, your whole experience of being in hospital and seeing doctors will be greatly worsened if you don't speak the language. This is where medical interpreters become heroes. Good news or bad, having it delivered in your language is so much more reassuring.

On top of the reassurance provided by medical interpreters, there's also the incredible level of professionalism required of them. Imagine having to interpret the dialogue between a doctor and a patient when it's not good news. Could you remain and react professionally knowing you have to break potentially life-changing news to a patient? I admit that I'd find it nearly impossible.

In terms of language professionals, medical interpreters are basically superheroes. They have the language skills that many language lovers yearn for, they're the tireless saviours of those in need of a helping hand, and they do all this while putting everybody before themselves.

You're awesome, medical interpreters! I salute you!

Translating Culture and Cultural Phenomena

Religious imagery plays a huge part in
Semana Santa.

This week, in Spain at least, is Semana Santa (Holy Week). If you've never seen the spectacle before, it's incredibly impressive. I'm currently in Málaga in southern Spain, where the festivities are quite vibrant and inexplicable. That's why today I thought it'd be useful to look at translating culture and cultural phenomena.

When it comes to translating cultural phenomena, things can get quite tricky for a translator. As such things are usually unique to a region, country, or even language, a direct translation probably won't exist because the thing itself neither exists in the target culture nor the target language. That's when creative translation practices can come in handy.

Is your audience familiar with the practice?

Before you start translating, you should think about who is going to read your text. For example, if I was writing an English-language article for British expats who live in Spain, I could just leave cultural terms in the original Spanish. This is because the readers would likely be familiar with the Spanish terminology, despite preferring to read articles in English.

What if your audience isn't familiar with the practice?

Say your target audience is unfamiliar with Spanish culture. In this case, I would have a bit more work to do. I can't really invent words for the terminology, as the practices don't exist in English-speaking culture. In this case, I would have to be more creative. For example, rather than trying to find a vague term to describe a specific cultural idea, I might need to keep the word as it is, but add a brief description to the sentence.

Semana Santa celebrations in Sevilla, Spain.

Is it really OK to not translate a word?

Of course it is. Languages borrow words from other languages all the time. In fact, we've done plenty of posts on loanwords that have made their way into the English language. Sometimes, as a translator, the words just don't exist in the target language. In this case, you may have to just keep them in the source language and explain them in the target language. If your audience understands the culture as it was described in the source text, then you've done your job. Well done!

Is there a specific cultural phenomenon that you've found hard to translate? What was the most complicated cultural element you've translated? Tell us about your experiences in the comments below.

Translating Numbers and Money Around the World

Last Wednesday, we looked at the differences between different time and date formats around the world and how, as a translator, it's important to get them right depending on the language you're using. Today I thought we could continue with numbers, since it's not just words that are affected by translation.

Numbers

When it comes to numbers, it's important to understand the "scale" commonly used in that particular language or region.

For example, in English, a "billion" can be two different numbers, depending on whether you are using the long scale or the short scale. As it's more common in the English-speaking world, I'll talk about the short scale first.

The Short Scale

In the short scale, you have thousand, million, billion, and trillion, and each of these is 1,000 times bigger than the last.

Using the short scale, a billion (1,000,000,000) is a thousand million. A trillion (1,000,000,000,000) is a thousand billion. Put simply, each new term has three more 0s after it.

The Long Scale

In the long scale, each term is a million times bigger than the last. Hence a million remains 1,000,000, but a billion is now 1,000,000,000,000.

The long scale is fairly common throughout Europe (with the exception of the UK). You can spot the long scale by its use of words similar to "milliard" (1,000,000,000) and "billiard" (known as a quadrillion in the short scale).

The long scale was previously commonplace in British English, whereas the short scale was used in American English. However, both varieties now generally use the short scale.

Writing or Representing Numbers

Of course, there's also the little matter of typing out those numbers. In my examples, I used the practice common in British English, separating each set of three digits with a comma, starting from the end of the number.

Then there's the matter of decimal places. In English, it's usual to separate decimals with a full stop (or period in American English). This means one half is written 0.5 when using decimals.

However, if you speak another language, you might be familiar with a system that works the other way round. So a million can be written either 1,000,000 or 1.000.000, depending on where you're from and the language you speak. This also means that one half can be written 0.5 or 0,5 depending on the system you're using. Make sure you take care!

Money

Money talks, and if you make a mistake with it in your translation, there'll only be bad things to say about you. This is when knowing the practices related to writing numbers becomes very important. Aside from making sure you are using the right scale and the correct separators, there's also the issue of currency names, and how they should be written.

For example, here in the UK we have the pound sterling, also known as pounds or GBP. One pound can be divided into one hundred pence. One pound and fifty pence is therefore written as £1.50, placing the symbol at the front and separating the subunits of the currency with a point.

However, if you're using Euros, you tend to see the symbol written as the separator. Therefore if we had one euro and fifty cents, it could be written 1€50.

Make sure you're aware of best practices when it comes to numbers and money, because they make the world go round!

Translating Dates and Times Around the World

If you're familiar with the internet or live in the United States, you may have heard that Monday was Pi Day. If you're not familiar with Pi Day, it takes place each 14th March in celebration of the mathematical constant pi (π). This date was chosen since it corresponds to the first three digits of pi as written using the US date format (3/14).

However, if you're from a country that doesn't write dates as they do in the US, Pi Day doesn't make much sense. When it comes to translation and learning languages, knowing how numbers are represented can be very important. Since we've just been talking about Pi Day, let's look at dates first.

Standardised times and dates are imperative in aviation.

Date Formats

When you write the date, the format may change depending on the language or region. The US uses a Month/Day/Year format, which means today's date would be written 3/16/2016 in full.

As a UK native, I am much more familiar with a Day/Month/Year format, which shows today as 16/3/2016.

In certain regions and languages, it is also acceptable to write the date starting with the year, then the month, then the day. That would make today's date look like 2016/3/16.

In addition to the order of the numbers, there are also the separators to consider. Thus far I have only used slashes (/), but depending on the region, dots (.), hyphens (-), and spaces can also be used.

Time Formats

When it comes to time, the two most common formats are the 12-hour clock and the 24-hour clock. As the name suggests, the 12-hour clock only includes the numbers from 1 to 12, with the numbers repeating after midday and midnight. The 24-hour clock goes from midnight to midnight without the numbers for the hours repeating.

There are very few languages that use the 24-hour clock when speaking. In English, you don't say "fifteen o'clock". When it is spoken, it's usually what is called "military time", whereby the entire time is said without separators like "fifteen hundred hours", meaning "1500", or "3 o'clock in the afternoon".

However, in French, for example, the 24-hour clock is referred to when speaking, alleviating the need to ever really refer to the time of the day.

In Translation

With all these different formats, it is very important when translating to be familiar with the conventions of the language and even the region you're translating for. Making a mistake with the dates and times in a text could make you look unprofessional and make the text ambiguous, confusing, or in a worst-case scenario, useless.

These practices can be hugely important when it comes to things such as contracts or medical records. In fact, a mistake on a medical record could result in patients being given treatment at the wrong time of day or even on the wrong date!