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...

When friends and family hear what I’m working on these days, they typically ask: 1) won’t AI eliminate the need for translators and interpreters? and 2) won’t that eradication of the language sector do away with your job, too? The first question has taken the air out of the room in a lot of discussions over the last couple of years, especially given media coverage of humans in the loop that amounts to job descriptions that are little more than janitors cleaning up after bad MT outcomes.

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...

CSA Research’s recent survey-based examinations of machine translation deployment at language service providers, enterprises, government agencies, and among freelancers revealed an ever-widening engagement with the technology. Although it didn’t surprise us, we also found widespread skepticism of claims that MT has reached human parity with numerous calls in open-ended survey comments for “truth in advertising.” Just as significantly, we saw widespread desire for MT to be more suitable for t...

Recent advances in machine translation (MT) have been truly astounding. Even if claims that the technology has reached parity with human translators are premature and an artifact of particular methods of evaluation, it is clear that the results are much more useful than they were even a few years ago. Now researchers are pursing approaches similar to MT to facilitate communication with animals. In April, National Geographic described Project CETI, an attempt to use the same fundamental technolog...

In late 2015 most developers still treated Neural Machine Translation (NMT) as a future technology that would come someday, but by the middle of 2016 several developers had released functional neural systems that seemed to outperform even the best statistical MT (SMT) engines. Since that point, momentum has shifted decisively in favor of NMT because it delivered better fluency in its output and seemed to promise the end of incomprehensible MT output. Eventually, this rate of stellar progress slo...

TELUS International announced that it would buy Lionbridge’s artificial intelligence business unit in a deal worth US$935 million that is expected to close by December 31st. Let’s first explain what Lionbridge AI does, review the deal, analyze what it means to the language services and technology sector, and show what it represents to investors.

In 2017, CSA Research introduced the concept of “augmented translation,” a technology-centric approach to amplifying the capabilities of human translators. Although companies such as Lilt, SDL, and Unbabel had already implemented portions of this model – namely the tight integration of human and machine capabilities, we predicted that more technologies would be folded in over time to create an AI-driven platform for linguists that would make them far more efficient and capable by offloading r...

A recent examination of how computing power has changed over time calculated how much it would cost today to rent all of the computer capacity that existed at various points in the past. Up until 1968 all the worlds computers’ combined capabilities fell short of what a single server could do in 2018. It was not until 1974 that the then-current capacity (roughly equal to 0.43 teraflops) would exceed £100/month to lease in today’s costs. This exponential drop in the real cost of computing power ...

Every day, nearly three quintillion bytes of digitized data come into being. This daily wave of new content supports interactions and transactions across the entire spectrum of human and machine activity – and localizing it is essential for many international business, governmental, and humanitarian activities. There is no sign of this daily growth in content volume slowing down – and with it comes gigantic projects to transform and translate it for other purposes and markets.