Warnings of possible consequences of the increasing and rapid development of artificial intelligence (AI) are published on a regular basis. In particular, in conjunction with the equally rapid development of automation, millions of jobs would disappear, according to the general narrative. The media industry of course will be hit by this trend as well…
Many experts agree that in the foreseeable future most of the articles will be written by “robot reporters”. Literature itself will be dominated by the AI — at least according to their forecasts. And indeed, a computer-generated novel has already made it into the second round of a literary competition in Japan. In fact, software programs have long been used in the cration of content. Bloomberg, just to name one example, uses AI to create almost one third of its news. Weather and geology services as well are already using for a long time now algorithms to generate their messages. And just last week, the Australian edition of The Guardian published its first article written by a software program. The topic was donations to political parties.
However, AI has also become a buzzword — everyone uses it, but hardly anyone understands what’s really behind. The huge influence it already has on our daily life is mainly overseen and the same goes for the real risks of its use. Instead an unspecific fear and distrust is growing towards this praticular kind of technological progress.
Nothing new here
Actually the concern that machines literally steal people’s jobs is not a new one. Back in the 18th and early 19th century machine breaking was a well-known phenomenon. The swing riots are a classic example. At the beginning of the 20th century people were afraid of cars as job killers: In the UK a posterwarned that more than 100.000 jobs might be lost due to “reckless motorists” who “kill your children,” and “poison the air we breathe”… Also in Germany concerns about automation are widespread. Since the 1960iest the most important news magazine of the Federal Republic of Germany, DER SPIEGEL, regularly raises the topic to its title page.
However, we will not run out of work — on the contrary. Despite all warnings from magazines, experts and such like there are today more jobs than ever before in the history of the UK.
Can AI create beauty?
Back to journalism and AI. The concern that in the nearer future mainly software programs will write our articles, texts and books is as well largely unfounded. It is not by chance that almost all the texts written by “robots” are strongly related to numbers. Weather and geologic reports or the latest quarterly figures. Here the algorithm can “work” with just a few text modules. It’s mainly about the numbers. The quality, beauty or depth of the article is less important.
The fact that programs today can beat not only the people in chess, but also in Go, is impressive, especially given the far higher complexity of the Asian board game. But Go also works on the basis of given, strict rules. “Surprising”, “new”, “irrational” and “completely different” are not to be find here. AlphaGo Zero, an algorithm which taught itself to play Go (only the rules were given to it), is undoubtedly an outstanding example of “reinforcement learning” and of the ingenuity of human(!) code writing. But the programm wouldn’t stand a chance in Ludo against a five year old, simply because it doesn’t know the rules of this game and can’t learn them without the help of its coders.
And above that writing is much more then calculating and following rules, although of course there are certain basic guidelines to be followed. Most of all, when writing texts, it’s all too often about things that software can not and does not have access to. It was William Faulkner who once said that “the human heart in conflict with itself is the only thing worth writing about”. So how could an AI write about it? After all, that’s exactly what it lacks: a human heart.
Littler helpers
The AI will not replace authors in the foreseeable future. Instead it will provide further help for content producers to create good, interesting and well-researched content. Spelling and dictation programs are getting better and better. They have already become indispensable in daily work. The same goes for search engines and online databases. These are all forms of AI. Especially large data sets can be searched much faster by AI for certain patterns or conspicuous accumulations of certain events. JP Morgan, for example, uses contract intelligence analyzing records in seconds, while legal personnel would need hundreds of thousands of hours to do so. The same applies, for example, to the analysis of financial data. However, the interpretation of the analysis results will still be done by humans.
Also, the media could — at least partially — reinstall their credibility through AI. The “old school” of journalism assumed that journalists should never attach themselves to any cause, even and especially not to a good one. The younger generation sees it differently. They want to promote the good and right causes and perspectives. But this is a quite dangerous path. Well-intentioned motives quickly become an occasion not to be as accurate as possible with the truth. In late 2018 Germany was shaken by a hugh press scandal: A well known and highly decorated reporter of Der Spiegel turned out to be a fake news producer — almost all of his articles were more or less completely made up. If the control and review department of Der Spiegel were software-based, Claas Relotius would probably have been exposed much earlier.
Don’t loose knowledge
Moreover, the process of automation is not a one-way street. Toyota, for example, one of the first automobile companies to focus fully on automation, is now bringing its workers back to the factories. Mainly for two reasons: First human beings are still working more precisely in decisive details than robots. Even more important are the concerns of Toyota in regards of a potential knowledge loss. If there are no humans any longer involved in the production processes, the knowledge about these processes will be lost. No AI in the world could bring it back then. And if this knowledge is lost, it will be impossible not only to improve but in generally to install further automated production.
crumbl supports authors
The same is true with the algorithms for creating content. To write their code well, not only experienced human programmers are needed, but also experienced authors, journalists and content creators. Without their input in regards of their needs, demands and work flows — in other words, their knowledge — it will be impossible to write use- and helpful content producing algorithems. And of course it would no longer be possible to improve the existing content AI.
That’s why we cooperate with experienced and renowned authors and content creators working for the closed beta version of crumbl. We want to adapt our “marketplace” as best as possible to the needs of authors and publishers. We are committed to provide the best possible results for all those involved. With crumbl authors have a tool at their fingertips, which not only helps them to create content, but gives especially support when it comes to publishing. Thanks to distrubted ledger technology (DLT), authors can track exactly where and how their content is published. A fair participation in the generated advertising revenue is guaranteed. Also crumbl facilitates the tracking of the contents after publication as well as the spreading in the social media.
Learn more about crumbl here or test us out in the closed beta phase and convince yourself.
All you have to do is send Felix a one-liner at felix@crumbl.org expressing your interest.
