E042 Ola Nilsson: CEO & Founder @ Lingmill

Ola Nilsson 0:00
then it's possible to ultimate it, you have to have some kind of process where you analyze data or numbers, and you can find what's most interesting about it. And then you through an algorithm write text about those things that occur in the data.

Abdulaziz M Alhamdan 0:24
Once upon a time, there were millions of businesses struggling every day, they wasted time, effort and money on repetitive tasks that added no value. One day, the better automation podcast by process CEO came to help them find a way. Because of this, these businesses save time, reduce costs, innovate, and make better decisions because of that. These businesses grow, scale, and use human creativity to change this world. Hello, my name is Aziz, and I'm your host that better automation podcast by process here where I interview the world's top experts and share their very best ideas on how to improve automation in your business, processes. And life. My guest today is Allah Nielson. Allah is a Content Automation expert and the CEO and founder at link mill, helping clients identify possible automations then automate content production through AI, Allah, how are you today?

Ola Nilsson 1:35
Oh, thanks, I'm fine. I've had a great weekend. So I feel energized and

Abdulaziz M Alhamdan 1:41
happy. I'm very happy for you, me as well, I feel positive and optimistic. And to begin this with a backstory about you and Ling mill and all that, how did you decide or come across the idea of becoming a Content Automation expert, and to create a company around it?

Ola Nilsson 2:03
Well, it started with, I was working as a reporter or an editor at a news agency in Sweden, and I was at the sports department. So my my goal for some nights was to write football, game reports, or match articles. And the instructions I got was right, as many as possible, because they're always very, very lots of games going on internationally. And I hadn't time to watch them, of course, because there were too many. So I just read the statistics about the games. And then you can you can create a kind of a good story, you know, when the goals were scored, if there were red cards shown, and so on. And I wrote as fast as possible that I maybe wrote seven or eight pieces each night. And I felt that I had more creativity within me than doing that. So I complained about this to my friend that he's an professor in cognitive science, I complained about this. And he immediately immediately said, that has to be automated. And I said, Okay, let's do it. This was like, maybe nine years ago, that we talked about, it's the first time and then we try doing it as more like a fun projects in the in the late nights. But with with my journalistic knowledge and with his AI knowledge, we could we realized that we could create something kind of fast that works we had so we in a few months, we were able to, to write 2000 articles every every day in in first Swedish and then Swedish and English.

Abdulaziz M Alhamdan 4:08
I have so many questions, but I'll begin with the basic one to have some context so that any listener will be up to speed, what is content automation? When is it useful? Can you explain it in a in very simple terms,

Ola Nilsson 4:27
the ways the different kinds of content is of course, gigantic. But if you if you go through what what's written in in, in news, news, magazines, and news sites, websites, you can find some of the content that is the same kind over and over again, even if this it's it's new information, but the type is the same. And when you can read recognize the type and you can identify the, the, the source of the news as, as numbers to be precise, then it's possible to automate it. And you have to have some kind of process where you analyze data or numbers. And you can find, find what's most interesting about it. And then you you write, through an algorithm write a text about those things that occur in the data. That's my first tried to explain it. Maybe I could do it better.

Abdulaziz M Alhamdan 5:46
No problem. Thank you. I would like to ask them the purpose, since you're speaking about news. Is that content automation, basically, to inform people but more in a story kind of way, so that it catches the attention and is entertaining? Or is it Content Automation, for SEO to have many articles on our website to rank on many keywords, is it to put on social media where you have some information and can be transformed into tweets into LinkedIn posts into something like that? Can you explain it more in terms of purpose, rather than process,

Ola Nilsson 6:27
our starting point was to automate automates the, the news as a way of storytelling as a way of telling it in, in in a good language and with a good headline, and with a compelling introduction, and so on. And also to, to go through lots of data, lots of material and find the news and then make headlines out of it. That was a purpose because you, as a human being you, you miss out on very much happenings, because you can't go through all data in the old world all the time. So that was the main purpose. But then we found that we could use our technique to to also create variety in telling almost some similar events for like companies. And then you can talk about the SEO, you have to, to create texts that are unique, and that are also nice to read, of course, for four people. But the combination combination of doing doing it in a way that Google likes, but I must say that Google is getting better for every day, as you all know. So the the aim of creating good stories in a good language, it. It serves both people and Google at the same time, because Google is getting more people like for every every day,

Abdulaziz M Alhamdan 8:16
thank you. And I'll play the devil's advocate a little bit. If it's going through algorithms, well, how much competitive advantage do you have? If someone else could in theory, copy it since it's automated? And therefore could someone just understand based on your analyzing your articles, or whatever it is, and just do what you do? And therefore, you're not unique anymore? What's your perspective on this?

Ola Nilsson 8:47
In theory, it's possible that that we are dealing with natural languages, the variety is so, so huge, that it's hard to backtrack, almost, I'd say it's impossible to backtrack it all the way down to the precise algorithm, because there's so much variety, and there's so much so many synonyms that are used in so many ways of telling it, because we, we write it, not just we don't just put in words, we also created as a story that has to follow the drama curve, you know, so it can be told in so many ways, so it's hard to backtrack. In that sense, I'd say and, and of course, we can are out but we don't start with machine learning. We use that as a as a second phase when we have created the foundations of the texts. But when you add machine learning and, and train with the data from all over the world from all all times, so to speak, then then it also would be so much input ways and so much variety. It's that it's hard to to, to copy, and it will, it will stay unique in that sense, I think,

Abdulaziz M Alhamdan 10:24
thank you. And again, playing the devil's advocate, I will ask about competitors, there are so many AI writing tools or rewriting tools. And well, they don't write like a good journalist or like a good human being. Although they try some of them said that they read most of the internet that is available to the public, and therefore they write better, but often, they might put weird statistics or say sentences that don't match the context or whatever it is. So what allows you as two people or a small team or whatever, to beat all the other AI competitors? Who could be out there? Or do you see yourself as constantly improving, and therefore, what's your perspective, of course,

Ola Nilsson 11:14
we have to improve all the time, and we have to, to evolve. But our basic idea is a bit opposite of our competitors, with all due respect they have, we have many competitors, that does a great job in several areas, of course. But our main, main thing is that we start start from a setup with journalistic principles. And the data has to go through those principles to be presented. We don't just want you to say, we don't just map words to the statistics. Because to, to go through our, our system, it has to fit in a certain analysis structure, we have to in a way understand what we are writing about. And to be able to do that we only write in certain areas, if we write about, say the first example football articles, then we just write about football, football. And if we are supposed to write about product descriptions for for dishwashers, we just write about that. And that's another algorithm. So we, we, we we have very What if a strange thing happened in a football game, like aeroplane landing in the field, our algorithm can't write anything about that game. Because that's so that's so common that the system will just say, we don't understand this data, and they will send us an SMS, or an editor or an SMS. And the human being has to look into it. That's our way of not telling untrue stories. Because that's, that's a problem. If you start with all due respect, if you start using just machine learning, then you can tell strange things, sometimes. So we want to avoid that. Because that's, of course problematic in these times when, when we have fake news and all that. So we have to be extra careful to not telling untrue stories, even if people also tend to write untrue stories. So there will always be someone to stories, but we try to to narrow it down. So that doesn't happen at all.

Abdulaziz M Alhamdan 14:13
Thank you very much. And I'm really curious to understand the use cases of what you're doing. Is it really about transforming data into a news stories that will get an audience online and therefore your monetize through ads, or subscriptions or something like that? Or is it content about for blogs and all that so that companies will get more viewers and increase their brand and all that or what are you currently the strongest as when it comes to use cases for what you do so that I know specifically what to ask about

Ola Nilsson 14:54
the strongest cases are those clients who has their own They had to have their own data. Because we could, we could, of course, use some open API's of data or stuff like that. But we have, we have found out that those businesses have their own data in, in, in some domain, then we can use that to, to, to create content content that no one else can create, because no one else has the data. And they don't have to hire 200 people to write about that, we can write it, we can write several articles, automatically, and then we have are in control of the data. And we can also analyze it over time. So we we also have this way of feeling secure about but the inputs, so we don't have to worry about that. I can give you an example, without telling any names. But there are lots of companies collecting data for, for companies, all sorts of data. And they use usually send out press releases about this data of for, say, a specific region or a specific country. And we can take the same data, but write specific articles on each small city or even small, small town due to that data, and not just about the whole region, and then you get this hyperlocal understanding of what's going on in your, in your street or in your small, small municipality. municipality. And that's, I think that's a strong use case, because we both help the clients, but we also give some thing to the end user that they they need what they shouldn't have a gun, it's if it was supposed to be written manually, because it's too expensive.

Abdulaziz M Alhamdan 17:12
Thank you. That's absolutely fantastic. And therefore, I will ask you, you said instead of hiring 200, people who understand maybe the journalistic principles, they can use your tool, they can use the what you have created, and therefore on more philosophical level is journalism is skill that can be automated. And therefore, there are you know, most journalists can be replaced, or what can human made journalists do to compete in a world where AI could create content fast or better or as good as them, It's always

Ola Nilsson 17:53
risky to look into the future. But I can I can try to do it in is some years ahead. And I'm, I'm almost sure that we aren't replacing people, because the automation, the kind of automation that we do is, it's more of the kind that we would do stuff that hadn't been done. If it was supposed to be done manually. But we reach more people instead. And instead of people doing a kind of boring writing jobs, so to speak, we can do the basic the groundwork, and then the journalists or the content creators can add more more quality or more some extra add ons in the text, you can use our texts as as a ground as a as a bottom line for an article and then you as a person manually add some extra some some quotes may be or whatever it could be, and then you in just a few minutes, get a much better content piece than if you had had to do all the work from the beginning. And I guess that you don't have the time or you won't prioritize to do that. If you haven't been given it as a as a ground or a basic article, automated automatically done. So I think that we more we more than replacing people we are helping people to to have more meaningful, meaningful jobs. That's my that's that's how it started. For me. I didn't want to write those EC articles. I want to zoom out analyze and do more creative stuff. And I think I am not the only one that has been thinking like that. Of

Abdulaziz M Alhamdan 20:07
course I do as well believe that automation and AI is not replacing people, it's enhancing and leveraging their time so that they will do really the things that put them in a flow state because they're unique. They're creative, they're not the boring, repetitive tasks. And I'm curious about something. Your tool is creating a story. And you said that follows the story curve. So about storytelling, some story writers say, No, you cannot create a story based on a formula. It's something that emerges from your subconscious. It's something that you feel you cannot really just put it into a structure organized from the beginning. While you are saying basically that you can follow a formula that even an AI or machine learning cookbook to have a nice and good story. What's your perspective on this? Can stories built on formulas be as fantastic as stories that people dream about? Or create? Without thinking? Or what do you see when it comes to storytelling?

Ola Nilsson 21:21
Interesting question. Because there are many parts in storytelling that are possible to put in numbers. I go back to the first example again, if you write about football games, we have found analyzing many, many games that there are a percentage in how many goals you should score as a as a player to become a game hero. If if the if the result is within, if the two teams are close, in the in the end, like say they want with three one or so it's, we can put a percentage at 40% where the player becomes a game hero. And there are other other stuff that occurs and when you look at it, and compare it over time, you see those those structures that yeah, if you, if you do five out of seven goals, you You are the game hero, and so on. And, and that's that you can find in other areas, but football is, it's a good way of showing it. So there are many things that could be put into numbers. In storytelling, especially when you're in the normal spectra, very extraordinary happenings are harder than you need to be creative. So if it's never happened before, it's it's hard to, to, to, to tell it in a in a good way as an algorithm. And I think that the the creativity that is it's possible for a human being is it's very hard to to reach that with with AI or how you will want to describe it. But on the other hand, you can also you can also find new ideas using AI that you you as a person never have been able to think about. But so as you said, I think the combination is the best if you can, if you can add the human creativity to to an AI system or what you prefer to call it that's then you can make the real magic happens so to speak.

Abdulaziz M Alhamdan 24:10
Thank you so much. This is so interesting. And to finish, can you explain in any way you can? What is the service you offer where people can go to learn more if they want to be customers? What can they do or should do so that they will know exactly how to benefit from a base that your offer and I'll make sure to write how to go to your website in the episode description.

Ola Nilsson 24:42
The bottom line is, is about needs if you you have a need of of content in perhaps maybe gaining customers or gaining traffic, web traffic then it's Almost always possible to find some data that could be used for for creating automated content. And in those cases, you know that you have your own data within your company, then that's the first first place to look for us. But otherwise, we can always find something that relates to your business. And then we can make news and content for you to gain web traffic or, or in in the next step again, customers.

Abdulaziz M Alhamdan 25:38
Thank you so much. And I do believe that content is very valuable. And we live in a world where content is king. And of course, before we finish, I first does your service have an API? Yes, yes, perfect. So it would integrate perfectly with processor to processor is the modern low code, no code platform for advanced automation and creating an enterprise grade back end for your software. Any viewer or listener can request access to a totally free account at processor dot app, and for those with higher business needs, there is a very generous 50% off discount it's better 50 of one word in capital letters, more information in the description. Thank you. Oh, Allah. This was my privilege, my honor, wonderful conversation. And I wish you to keep going because you're creating the future.

Ola Nilsson 26:38
Thank you. Nice to talk to you.