Article by Sarah Hickey.
It’s a new year, a fresh start. While some may make resolutions, we at Nimdzi make predictions. Five predictions, to be exact, about what we can expect to see in the upcoming year 2023.
Open AI’s ChatGPT is ruffling feathers
OpenAI’s latest GPT-3 (stands for Generative Pre-trained Transformer 3) variant, called ChatGPT, is all the rage these days. And for good reason — the technology has come a long way and some are calling this a tipping point for artificial intelligence (AI). For those not in the loop, here is a brief introduction: GPT-3 is trained on existing content and applies machine learning to perform a number of on-demand tasks with only minimal input, including:
- Translation of common languages
- Summarizing text
- Writing articles, poems, stories, scientific papers, etc.
- Answering questions
ChatGPT is the latest variant of GPT-3. In addition to the tasks listed above, ChatGPT can also:
- Write and correct code
- Be used in chatbots and virtual assistants
- Generate conversational text
- Explain complex subject matters
- Be fine-tuned/customized for a variety of language tasks
How does that make it a potential game-changer as opposed to what’s already out there? There’s two main ways: 1) the quality of the results, which sound remarkably human, and 2) the potential use-cases because it is open to everybody! Although not without its limitations and by no means a de facto source of truth, ChatGPT has already changed the game because it enables anybody to produce more work at a much faster rate. To what extent the technology can be used (e.g. how much editing and fact-checking is required), and whether it will truly become “the Google killer” that some hype it to be, remains to be seen.
In the language industry people are talking about ChatGPT and trying to figure out how it will impact them. It is too early to tell how big the impact will be, although use cases are starting to crystallize. Already, it has proven to be a useful tool to unlock writer’s block in the content creation process or to fast-track scripts for presentations and speeches. While not entirely monolingual, the technology works best in English for now.
For our prediction, we are applying Amara's Law, stating that we tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run. What does that mean? Right now, there is a lot of hype and ChatGPT has already become a game-changer no matter which way you look at it. However, whether this means that the technology will be a true disruptor in the long run in the sense that it will impact our everyday lives as well as our workflows in the language industry is rather doubtful but it remains too early to say.
5G and the Metaverse: No impact yet
In previous years we predicted that 5G and the metaverse would trigger a lot more content being created (in different formats, attracting new end users) and that this would have a knock on effect in the language services industry (e.g. higher volumes, new workflows). While we still expect this to happen eventually, it has become clear that mass rollout and adoption are still a long way off. This is why we predict that 2023 will not yet be the year that we will see the real impact of 5G and the metaverse.
Consolidation slowdown
Over the last few years, the language services industry has been flooded with announcements of mergers and acquisitions (M&A) and investment from private-equity and venture capital firms. While we do not expect the M&A and investment wave to be over any time soon, we predict a slowdown until the global economy has stabilized. At the time of writing, in January 2023, the world is experiencing inflation, currency fluctuations, and an energy crisis. While this will not necessarily stop companies from growing, it does create uncertainty for investors.
The slowdown can already be noticed if we compare the figures from 2022 to those from 2021. While there were 62 recorded M&A transactions in 2021, only 35 were recorded in 2022.
More use cases for synthetic voices
2023 will be the year we see more use cases for synthetic voices. The quality of synthetic voices has come a long way. Not only do some voices sound so remarkably human that in some cases it can be hard to tell whether or not the voice is synthetic, but also the latest developments involve technology that is able to mimic the original speaker's voice in the translated, synthetic version.
In the conference and meeting space, machine interpreting is making inroads. Although not fit for entertainment purposes (yet), AI dubbing is already being used for international broadcasts, as well as for voiceovers for documentaries and corporate videos.
No new lang tech unicorn
With the amount of outside investment that has been going around for the last few years, one can easily get the impression that the language technology space is booming. And only as recently as November 2022, DeepL became the first business in the language technology market to receive a unicorn valuation (i.e. a privately held startup company valued at over USD 1 billion). However, with the changing leadership at Smartling, layoffs at KUDO and Lokalise (23% laid off), the rebranding of Phrase (formerly Memsource), and TRADOS again becoming the top brand in the RWS suite, we don’t think that 2023 is going to be the year that we will see another language technology unicorn.
Although nobody can predict the future with absolute certainty, based on our expertise as industry analysts, we feel strongly about the predictions made above and will continue to monitor developments throughout the year.
This article was prepared by Nimdzi's VP of Research, Sarah Hickey. If you wish to learn more, reach out to Sarah at [email protected].
