Our Analysts' Insights

After we published our recent Q3 2024 update on market sizing for the language sector, which was also covered in a public webinar, this blog addresses some of the questions we have received from clients, prospects, and investors. This quarterly update for Q3 2024 is noteworthy because it includes our final annual market sizing numbers for 2023, which are based on a representative sample of LSPs and produced only after careful examination of vetted revenue data following the close of the year. Th...

In October 2023, we argued that the future of AI would be in “focused large language models” (FLLMs). These are purpose-built language models that target a specific industry, set of languages, or task and that are correspondingly smaller than the large language models (LLMs) being created by OpenAI, Google, Meta, and others. Those massive models – GPT-4o has over 175 billion parameters – are like Swiss Army knives: They are prepared to handle almost any task, from creating a haiku to drawing...

It’s the end of 2023 and, rather than contemplating a bright and happy new year, many enterprise localization groups are looking at another year of austerity measures. Through most of 2023 they told us they were either holding steady on spending or cutting it. Here’s the rub though: By saving a bit on language services now, they may block substantive contributions to revenue in order to save comparatively small expenses. In this blog we review why this penny-wise approach could be pound foolis...

A common worry about generative AI (GenAI) is that the content that it creates may be subject to copyright claims. Our recent survey of freelance linguists reflected this concern: Copyright issues are their second most important concern with GenAI, with 74% viewing the technology negatively or strongly negatively in this regard. However, an examination of claims about copyright and how GenAI works reveals a different picture. This blog post covers details about how the technology works with its ...

It is incredible to think that, less than eight years after the first publicly available neural machine translation (NMT) systems appeared on the scene, some media coverage already sees NMT as so 2015. As generative AI (GenAI) really exploded into public view in 2022, it wasn’t surprising that an overactive tech press’ imagination would see it as the be-all and end-all of technology. Our recent survey with freelance linguists certainly reflects this view, with many language workers expressing ...

The explosion of interest in generative AI technology like ChatGPT has led to general calls for the regulation of artificial intelligence – these will also affect the language industry. Although today it is unclear who would be responsible for a critical mistranslation stemming from MT on a company’s website, CSA Research predicts that within a year some unlucky company will become the textbook case that decides this matter. This post explores how current and proposed regulation of artificial ...

Large language models have been in the news a lot in November and December and the coverage has been mixed, to put it mildly. Meta posted its Galactica model on November 15 but took it down just three days later in the face of intense criticism. By contrast, when OpenAI released ChatGPT two weeks later, on November 30, the response was much more positive. Examining why the reactions were so different provides insight into the potential and limitations of machine translation (MT) as well as cauti...

It’s planning time once again, but this task is especially fraught this year as companies are facing another year of turmoil just as large-scale pandemic woes seemed to end. But now, looking at hyperinflation levels not seen in the US and western Europe since the 1970s, companies are naturally cautious. Localization groups – usually treated as a necessary evil in the best of times – may find themselves the target of bean counters eager to save money. So what are you to do when executives and ...

Technology developments tend to follow a typical pattern of improvement over time, known as an S-curve. Although it is a familiar pattern, it is worth unpacking its five phases and considering how they apply to language technology and forecasts about it. Examining how they have played out with successive generations of machine translation points to a future in which other advanced natural language processing technologies have tremendous potential to deliver useful and innovative capabilities.

Last week, the Washington Post published an article about Blake Lemoine’s claim that his employer Google’s LaMDA language model/chatbot system had achieved sentience and had a “soul.” Lemoine, an engineer in the company’s responsible AI group, based his assertion on a dialogue in which LaMDA expressed human-seeming sentiments and concepts. Google placed Lemoine on leave, thereby sparking renewed discussion about what machine sentience is and what it means. What can the experience of the lan...