The ChatGPT revolution is here, but is the AI chatbot a viable option for your translation projects?
ChatGPT generates human-like responses to a wide range of inputs by using natural language processing algorithms. And much like Google Translate, you can also use it for translation.
AI translation is already a prevalent tool among language professionals. It helps them complete jobs faster and to create more cost-effective translation processes and workflows.
ChatGPT might not yet be an absolute gamechanger, but it can certainly disrupt the translation industry.
But, as we’ll discuss below, human intervention remains key, especially in the verification stages.
Power of ChatGPT for the Translation Market
ChatGPT generates human-like responses to a wide range of inputs by using natural language processing algorithms.
You can also use it for translation much like Google Translate. However, Google Translate’s mobile app can handle live speech and text in images through its Lens technology, which is something ChatGPT can’t yet accomplish. That lack of image incorporation spells the end of localization capabilities at present as well.
At present, it supports almost 100 languages, and it’s also capable of handling over 25,000 words of text. This allows for use cases like long form content creation and translation.
ChatGPT Translation in Action
Because of its conversational nature, ChatGPT’s translation flow is much more contextual.
We did a simple text to show how it works by translating the Russian phrase “Любовь зла, полюбишь и козла.”
Here’s the translation using Google Translate:
ChatGPT, however, offers the following response for a request to “translate Любовь зла, полюбишь и козла to English”:
The phrase “Любовь зла, полюбишь и козла” literally translates to “Love is evil, you’ll even love a goat.” However, the more accurate and commonly used English equivalent of this Russian proverb is “Love is blind, and lovers cannot see” or “Love is blind, love can’t be blamed.” This expression conveys the idea that when someone is in love, they may overlook the flaws or negative aspects of their partner or the relationship, and make poor decisions as a result.
So, with Google, you get a straight up, simple translation. There’s no context, and it’s maybe not literal, but it works. ChatGPT offers a much more thoughtful response.
Additionally, you can adapt translations to suit your specific needs and provide feedback on adjustments you want to make.
ChatGPT can consider cultural connotations and regional differences in the meaning of words, something purpose-built translation tools like Google Translate can’t do.
That’s but one example, though. Let’s take a look at some feedback coming in from other sources.
How accurate is ChatGPT for translation?
For starters, here’s ChatGPT’s own judgment of its translation capabilities via Slator.
“Yes, you can use ChatGPT for translation…However, it’s important to note that ChatGPT’s translation ability may not be as accurate or precise as a specialized translation tool or a professional translator.”
Additionally, “ChatGPT may not be able to translate certain technical or domain-specific terms accurately. Nonetheless, ChatGPT can be a useful tool for basic language translation tasks.”
And don’t forget, ChatGPT can produce incorrect answers to user queries. It does not always guarantee correct information. Plus, the major drawback is that its data comes from 2021, with no knowledge of events – or changes in language – after that point.
Next, Chinese tech company Tencent found that ChatGPT fell short of Google Translate and DeepL when translating Reddit comments.
Finally, Microsoft has a huge stake in ChatGPT, recently investing $10 billion to compete with Google and use ChatGPT in Bing searches and to be embedded in Word, Excel, PowerPoint, and Outlook.
They also recently compared GPT language models to both research and commercial MT engines in 18 high- and low-resource language pairs. They used publicly available datasets and tested at both sentence- and document-level. Multiple metrics as well as human evaluation were employed.
The conclusion was GPT models produce a “very competitive translation quality for high resource languages. However, ChatGPT still has “limited capabilities for low resource languages.”
For reference, English, Chinese, Spanish, French, Japanese and more of the European and Western languages are high resource.
To be more successful with low resource languages, text data collection and annotation is necessary. If you’re translating into these languages, then, ChatGPT won’t be enough to get the job done.
Importance of Post-Edited Machine Translation
Make of all that what you will, but here’s our two cents.
There’s no one way to approach translation projects. Truly, we’re quite firm on flexibility around here, and offer the specialized services to back it up.
It’s a sliding scale between pure human translation and handing everything over to the machine.
Human translation is slow but solid, and it comes at a cost. The quality, though, is premium.
Machine translation – like ChatGPT – is quick and simple, and you can get it done inexpensively. The quality, as we’ve seen, isn’t outstanding, and it’s better as a translation evaluator at the moment.
But with post-edited machine translation, you can achieve some semblance of balance.
That’s the process of revising and refining your content after it has gone through machine translation. A human translator checks the content to make sure it’s not only accurate, but also has the right voice and tone and is culturally appropriate.
We still need human review for more in-depth localization and translation projects for the highest levels of accuracy. And that’s especially true for low-resource languages.
Machine translation with human touchpoints is where you’ll find a balance between quality and cost efficiency.
Let Us Help You Find the Right Translation Balance
At Summa Linguae, we consider all the factors before rolling out a custom price.
First and foremost, you must know what exactly you need. Make a list of the languages into which or from which you need to translate and decide on the level of language services your company needs.
Will you need translation or localization? And is it a one-time need, or will there be ongoing projects?
The more detailed information you provide to the person quoting for your translation needs, the greater the chance that you will be offered the most attractive price. And if you don’t know exactly what you need, we can help figure it out.
Contact us today to get started.
Pending No More: Move Ahead with Patent Translation Services
Patent translation services require a balance of language and technical expertise, and a flare for legalese.
Quality Insurance: Leveraging AI for the Best Machine Translation
Run it through the machine translation process, test accuracy with AI, then comb it over with post-editing...