Delen:

Video Jan 23, 2025

The Future of Translation pt. 2

Hi, and welcome to three questions with MotionPoint. I'm your host, Dominic Ditabay. Today, I have the enormous pleasure of introducing our new chief proct officer, Steven Tyler. Steven, how are you doing?

I’m doing great, Dom. It’s great to be here.

How are you? I'm doing great too. I'm excited to be here. I'm excited to be on the iconic pink chairs with you introducing you to the translation and localization industry world.

And right off the bat, I'd like to talk to you about the future of it. Right? So what are some key trends and technologies that you're noticing that are gonna shape the future of industry?

I mean, it's the right of the machine. Right. And AI coming of age in this industry as it has in every industry, but particularly in this one, since sort of twenty sixteen when sort of Google really led the way into neural machine translation and has that as evolved.

We've now got to the point where, the quality being produced out of machine translation is good enough now Mhmm. For a a lot of context and a lot of jobs. So you're seeing a rise of companies in the space that that are taking a empty first approach, do a machine translation for everything, and then deal with the, deal with the exceptions, that need to be sent for human translation to take it the last mile to fix things. The machine didn't do particularly well. And then in more mature companies, you're seeing a a similar the one higher quality.

They they're still looking and asking the question, how do I get machine translation leveraged in there to bring down the cost of getting that quality so that I can stretch my budget further into other things? So the, you know, that's a big inflection point.

The last twenty years is changing dramatically versus what’s gonna come in the next five to ten, with regards to the nature of how translation service for companies like us combine machine and human capabilities smartly.

Yeah. Yeah. So with the new acceptance of machine translation in our industry, how do you really feel about how currently the translation vendors are are are in the market are aligned to that? Do you think there's a big delta between what they're doing, their business model now, to what the these these, you know, key indicators for the future are showing.

I mean, all of us are in the same boat of trying to figure out how do we do this, how do we weave into what we do the capabilities that the machines are providing, and acknowledging the fact that that is getting better and better over time.

So it’s a shifting landscape.

And that, you know, that whole thing about what was, human quality versus machine quality as being a sort of good versus evil. It's no longer a good versus evil. It's a continuum of quality, and the machines are getting better all the time. So I think all vendors are trying to figure out how they boot all their business models for to to to handle that.

It’s a challenge for all of us, but, you know, that’s what makes this the interesting space to be in right now because it’s a it’s a space with high innovation, as as everybody tries to navigate that inflection point, away from a dependency on purely human services.

Great answer. Absolutely.

So how can companies, you know, like MotionPoint, translation vendors, and LSPs out there, you know, better optimize and engineer content to create cost efficiencies, you know, maybe finding a balance between human and and machine translation there.

Yeah. I think I think the key the key thing is context. Mhmm. So as we think about, you know, any given piece of content, what quality do I need this to be at for the in terms of for the business?

If I'm a consumer of services from MotionPoint or anybody else, obviously, the content on my home on my homepage is a higher sensitivity. It has to be more right than something site map. Or it may be, you know, with certain types of content in certain industries in fintech, it may be highly highly quality sensitive in other in other industries less so. So I think that awareness of context and then being able to carry that awareness of context through the execution of the translation workflow.

So we say, okay.

Maybe I can use the machine to take me the first mile. And then based on context, I can kinda decide, like, where do I wanna go for it? Do I need to do a post edit? Do I not? Is it good enough already? And if we can get smart with how we do that, and I think there's other applications in that sort of decision making around Yeah. That are gonna be AI and ML driven, as well as the core NMT tech.

That's where we're gonna start to see this next step change of cost efficiency, being able to do, a given quality for a given language at a lower cost overall. Yeah. And therefore, being able to stretch the localization budget to enable me to go into more markets Yeah. At the same cost and have a bigger, larger impact for the localization budget.

So I think context awareness, I think, is the general answer I think that’s great.

It's almost a little bit of resource allocation where are you gonna put, you know, your best quality translation, where you're gonna put maybe lower quality, not good versus evil, but a little lower quality translation and other less trafficked parts of your website. Thank you very much, Steven. So that does it for three questions with MotionPoint. Thank you very much. Until next time.