| Renato |
I’m Renato Beninatto. |
| Michael |
And I’m Michael Stevens. |
| Renato |
Today on Globally Speaking, we want to talk a little bit about artificial intelligence. I was inspired by a tweet that I saw from the GALA Conference in Munich that happened in March. The tweet quoted Mat Velloso who’s a technical advisor to the CEO of Microsoft. And he said: “If it is written in Python, it’s probably machine learning. If it is written in PowerPoint, it’s probably AI.” I love this quote. |
| Michael |
You love this quote because it’s identifying the hype curve. |
| Renato |
Exactly. We have already had this discussion here at Globally Speaking about how there is a lot of hype in our business, and artificial intelligence and neural machine translation are frequently associated one with the other and how it’s going to revolutionize the industry. And I tend to think that that’s a little bit too much. |
| Michael |
Yeah, and if we take a little step back first, artificial intelligence is a sub-section of the overall idea of machine learning. |
| Renato |
Absolutely. |
| Michael |
Machine learning is something that is proven; it’s in production. I’ve a quote here from Tom Mitchell. He says: “Machine learning is the study of computer algorithms so that they improve automatically through experience.” So, it’s a self-improving machine, and it’s done really, really well in image recognition. |
| Renato |
We’ve had episodes on the concept of neural computing and how this has had a significant impact in machine translation and the perception of quality of the machine translation, where the translation looks good, but it’s not accurate. It’s still in its infancy. It’s very early in the hype curve. It’s a classical situation that we see in business as a whole, but it affects the language services business, in our space. |
| Michael |
Absolutely. |
| Renato |
I once interviewed an executive, and we were talking about having somebody responsible for products. Who would be the person responsible for developing products inside a translation company? |
| Michael |
And, when is that product ready to go to market? |
| Renato |
And that was the question that I asked during the interview. And the answer was brilliant. He said, “Well, in Silicon Valley, the product is ready when the PowerPoint is ready.” So, PowerPoint is very, very powerful because it can take any story that you want to create. So, I have a question for you, Michael. |
| Michael |
Yes. |
| Renato |
My theory is that when you’re selling services, you’re selling a story. So, what is the role of PowerPoint? What is artificial intelligence? |
| Michael |
Right, my version of artificial intelligence or what I’m spinning as far as artificial intelligence? I did some time also working with offshore software development services. And what was really interesting about that is it truly was, when the PowerPoint was ready, you were going to clients with it. |
| Renato |
Yeah. |
| Michael |
So, we were pitching things that, there’s the term slide-ware… [Laughter] |
| Michael |
It did not exist in reality, but we had a concept, we had a hunch that we could support this concept, and it was out there, and we were selling it and we were taking money for it. And then it was a hustle game to find developers and offshore teams who could help us build such products. |
| Renato |
But, do you think that that’s wrong? |
| Michael |
I think it is ambitious. And there is a high failure rate with that sort of approach. |
| Renato |
Let me contrast slide-ware with shelf-ware. |
| Michael |
There is a difference there. Well, this is one of the reasons where I was more comfortable coming back to the language services industry. Because while there is some element of slide-ware in the aspirational work we do, ultimately you need to have French on Thursday. |
| Renato |
[Laughs] Yeah. |
| Michael |
Right? So, there’s accountability to that. |
| Renato |
Well, but the point is that French on Thursday is a deliverable. But how you deliver that is the slide-ware. Essentially, shelf-ware is that product that you develop, you invest a ton of money, you create a beautiful solution for a problem that nobody has. |
| Michael |
There’s absolutely no market. |
| Renato |
No market for it and it stays on the shelf. I love languages, you know that. I like to learn languages. And when CDs about languages came out, I bought a full Japanese course— |
| Michael |
On how to learn Japanese through a CD. |
| Renato |
—on how to learn Japanese. I did do the first two lessons, and I still can say *nihongo ga wakarimasen*—those who understand Japanese will get the joke. (Translation: I don’t understand Japanese.) That’s the typical shelf-ware: it’s the stuff that you buy and you think you’re going to use and you’re not going to use, or a product that you develop that you think there is going to be a market for it and there is no market for that. So, the point that I wanted to make is that this whole conceptual sale—developing an idea and selling the idea—makes sense when the sales cycle is long. |
| Michael |
Yes, yes. |
| Renato |
And you understand why, because you sell a concept, you sell an idea, and it’s going to take months for that client to make a decision and to implement that stuff. So, very often, I hear people developing products that they want to bring to market, and they don’t know if the market is going to accept that they’re based on generic assumptions and not on real needs. And this is where the PowerPoint plays an important role because you can present a concept and then go and develop it. |
| Michael |
Yeah. And there are groups. There’s one here in Seattle called 6 Month Startup, and it’s an ideation group. And the whole idea is if you believe that you have a good idea for a company or a product, there are ways to prove that with very little investment. It’s called an MVP, this is the minimum viable product. |
| Renato |
Exactly. |
| Michael |
What do you need to do to see if somebody is willing to buy it? No more, no less. One of the common catch phrases is: “If you’re not ashamed of it, you’ve waited too long.” |
| Renato |
Exactly [laughs]. Yeah. It’s a whole concept about websites. If you’re not ashamed of your first website, you’re probably not doing a good job. You’re not advancing. [Laughs] |
| Michael |
It’s really helpful for people like you and I, Renato, who have a creativity and we hear the desires of our clients. We hear them talking about their aspirations and their goals and we go into this mode of: how do we deliver on that and how can we creatively find a way to do such things? Because the world should look the way our clients want it to look. It’s often good to have someone with a client delivery and production background to be with us to say, “Yes, but here is a limit.” |
| Renato |
Absolutely. |
| Michael |
And that tension is really helpful. I was just in a meeting a week ago where I had a solution architect and we were literally trying to solve the unsolvable. And we were in the client’s court making this case for how we could get to this place that hasn’t been done before. And, thankfully, we had one person with us just to go, “Uh, uh—just remember, there’s still project management.” [Laughter] |
| Michael |
“Just remember!” |
| Renato |
Well and again, this brings me back to the whole mantra that I’ve been using about the language services business: what translation companies sell is not translation; they sell project management. And how you use machine learning in artificial intelligence is how do you incorporate machine learning and algorithms to streamline the process. I was talking to a company that had developed this magic solution. I sat down with them and they’re describing me the whole process and how they are developing their product, and in five minutes I said, “This is a stupid product.” And I calculated something like 47 clicks between the moment that you receive the file and the moment that the file is going to the translator. And then it has to go to the translator; they have a ton of clicks, and then it comes back, and then there’s a ton of clicks, and so on. And this is the kind of area where you want automation to make a difference. You want automation to remove repetitive tasks and human hand-offs because every time you have a hand-off, you have an opportunity for a mistake to happen. Every time that you have a human intervention, you have an opportunity for a human error. There was this terrible accident with the Ethiopian Airlines airplane where it was a 737 Max that nose-dived after takeoff. |
| Michael |
Yes. And just recently, Boeing said that their software contributed to the crash and one other crash that was involved with the same plane. |
| Renato |
So, one of the interesting interviews that I heard was that one… an expert saying that when you try to remove human instinct from a process by using software, you deteriorate that process and you make it worse. So, you need to find a balance of how much human input you want and how much automation you want. And we’ve talked about this before here in our podcast: the power of management by exception, right? So, you want human input when things go out of the norm, and you want automation when things are repetitive, equal, always the same, always predictable. |
| Michael |
Because humans often are not very good at that kind of work because they lose focus. |
| Renato |
Exactly. Not all humans. I mean… |
| Michael |
I certainly do, though [laughs]. |
| Renato |
I was going to say that you’re an exception, but… [laughs]. I like your humility [laughs]. |
| Michael |
Yes. Yes, the long-term same task really tends to bore me, and I get distracted. There’s, like, squirrels all over the place in those moments. “Squirrel! Squirrel!” |
| Renato |
Absolutely! [Laughs] Squirrel! |
| Michael |
There are tasks that machines do well, and they’re able to improve and move forward with them, and then there are tasks that people do well. And identifying those tasks and sending them to the right source to get them done is very important. Blockchain is another one of these ideas that we’ve spoken about before. And they’re so aspirational that it’s important to say, “What is it actually doing?” |
| Renato |
Yes, but let’s look at the other angle. Why is everybody talking about AI? Because the market wants that. The stock market wants that. When the iPhone came out, I had a conversation with a client and said, “Well, I don’t have any budget for localization, but if I tell any executive in the company that I’m going to develop an app for the iPad, money will flow from anywhere.” And we are in a similar situation with artificial intelligence, machine learning, blockchain: these are buzzwords that every business wants to be associated with. We recently had the story about Lionbridge acquiring Gengo in Japan because they had an AI component. The smart thing that Gengo did was associate itself with the concept of artificial intelligence. |
| Michael |
Absolutely. |
| Renato |
Keep in mind—you an entrepreneur, you translator, that are listening to us—if you associate whatever you do with an AI component, you buy a domain that ends with AI, you can [laughs]… |
| Michael |
…spark some interest. |
| Renato |
There was a company called Uniscape. We’re talking about early, early ages of the localization industry. And this guy was brilliant. He created an automated translation management system that he sold to everybody. Just recently I was in Silicon Valley and I met a person—she’s not in the industry—and I started talking with her and she said, “Oh, you’re in localization. I used to work for a company called Uniscape.” And I asked her, “Were you one of the monkeys behind the screen?” and she said “Exactly! Everybody thought that what we sold was that we were doing automated things, but in the background, it was a bunch of people running around and working very…” [Laughs] |
| Michael |
…people working very, very hard, very hard. I wonder if we will look back on this season and, similar to how every company had dot.com at the end, if we look back on this dot.ai and go, “Yeah, that was a fun time. There were some good things.” [Laughter] |
| Renato |
On the lighter side, you know how everybody is wearing beards and they’re trimming their beards these days? |
| Michael |
Oh, the manscaping? |
| Renato |
But, I saw a posting the other day on Instagram, I think, of a guy in the ‘70s with a mullet and a mustache looking totally ridiculous, and they said, “Twenty years from now, this is going to be you!” [Laughs] |
| Michael |
That’s right! Exactly. |
| Renato |
And that’s the same thing with technology—ideas age. |
| Michael |
Well, I think we are in an exciting time and the challenge is if we’re expecting artificial intelligence to respond and act like a human, we may be dumbing down the artificial intelligence. What do you think will be the actual true mark of artificial intelligence if it’s not PowerPoint? |
| Renato |
Well, artificial intelligence is a reality. It’s hard to stop the advance of it. In our lifetime, we are going to see amazing progress and changes due to this kind of automation enabled by artificial intelligence. But, I also think there is going to be an element, a renaissance, that is going to flourish in this environment. We tend to look at new technologies as something that is going to completely replace what we have in place today. My belief is that new technologies add on to what we have today. So, it’s not like what we do and the business that we’re in is going to disappear; what we do and the business that we’re going to be in is going to be transformed by these technologies. And the nature of the work that we do is going to be completely different and there is going to be new jobs still associated with the language industry. We didn’t have post-editors 20 years ago. We didn’t have software engineers, localization engineers, 30 years ago. We didn’t have computers 30 years ago. These are things that are going to co-exist as we move forward, and PowerPoint will be around. Maybe it’s going to be 3D, maybe there’s going to be PowerPoint VR, and we’re going to be like “Minority Report” where Tom Cruise is moving files around with his hands on a virtual screen. Maybe. It’s going to be fun and I can’t wait to be in that period. |
| Michael |
I agree with you on that. I have loved the question that’s been posed because we have machines now that are creating works of art. We have machines that are creating music. And as novel and innovative as those advances are, the question is: but do you want to pay for them? |
| Renato |
Yeah. Very good. On that note, I’ll see you next time, Michael. |
| Michael |
Sounds good. |
