Pareto says don't be evil

While working at a startup about content creation in media industry, I made a little innovation through Natural Language Generation. I created a code that generate new articles based on scraped articles. The code would just scrape the news from various sources and aggregate when such news were about the same issue. Then it creates a multi-document summarization that even changes the way how they are written in such a way they can’t be detected by the plagiarism detector. I thought about this in the context of scraping from sources that are press agencies and allowing for the reader to learn the source. But I didn’t think deeply about the consequences in case such code would get in the wrong hands and of course I cannot control what happens next once an innovation is leaked. I thought about it by wondering how most journalists work and how boring is often their work. Therefore, the concept is just the imitation of what many journalists do: they just search the information from different sources and rewrite it. One would wonder is it wrong if an AI does it instead of a human. I think many people would not appreciate it and maybe taking the jobs of people who just try to make a career in the media sector is not that ethical. Especially if such kind of job is done by interns and low-level journalists who are just trying to make a living and a career. The worst thing that could happen is for this code to be used to scrape and rewrite original content: not news but pieces of work like blog posts or opinions. In that case it is even more unethical. And it is unethical to my point of view especially because I think such content creators are underrated and they should be promoted and renumerated more than they actually are. I think they are the weak and underrated piece of the internet that must be enhanced and empowered. Therefore, if a tool like this goes in the hands of more powerful actors it might be used to decrease their value, and that’s wrong.

Therefore, I tried to think about an alternative that could be Pareto optimal. I had learned the concept of Pareto optimality during college, since I have a degree in Economics. A Pareto improvement choice is a choice that improves some actors by not causing any damage to any other. I believe in the case of innovation the concept of Pareto optimality is a very important concept that any innovator must understand. If an innovation you make is not Pareto optimal, that means that there are actors that would become your enemies and even endanger the stability of your innovation. Those actors are the group of people whose profit is damaged by the innovation.

The alternative of my innovative code is just about allowing an aggregating system but removing everything else. Just a news aggregator instead of a natural language generation system. I found out that it is possible to make a very cool news aggregator with the sources and tools that I found in the web. Therefore, I built an aggregator that allows to aggregate based on a lot of different information like: the source, the country, language, date of publication, words or sentences used in the article. That might be a very useful tool for the media sector. This choice is a Pareto optimal choice: nobody gets harmed by an innovation like this, and some gain through it. Because the aggregator allows for the people who use it to click on the article and open it on another page. The blogs and media outlets need clicks in order to make money, if instead of scraping and rewriting their articles one just allows for the users to click on it, both the user and the source is fine with it. The source gains the click, the users get the article.

I got the inspiration to change the tool from Google, particularly from Google News. Google News is basically an aggregator. Google News is Pareto optimal: everyone gains through it. The user gains, the news sources gain and of course, Google gains too. And in the long term, the content creators are more incentivized to create content: the tool serves as an incentive for more content creating. If too much people steal the content from the internet and use it on their own, it becomes a disincentive for content creation in the internet. And a disincentive for content creation in the internet is bad for Google too. Because it is true that content creators need Google to make money, and it is true that Google needs content creators too. Thinking more deeply about the ecosystem made me think about a Pareto optimal innovation.