AI learns to create valuable content - a threat for Steem?

Good morning Steemians,

@neavvy wrote an article which describes superb the evolution of AI, the possibilities, his own project and the consequences it could have for the Steem blockchain. In his blog he mentions the idea, what if AI could scan trending Steem articles, analyzes successful patterns and create a brand-new post with no copy right errors. @crypto.piotr draw my attention to his post and I honestly recommend reading it.

Source

Concerns

Steem would be a mess or no one would take some time to prepare their post themselves. The gratitude of upvotes for your small intellectual masterpieces would mean nothing. Everyone would suspect that people would use the bot, so even true “artists” & good writers would lose motivation since trust on this platform would decrease.

In my opinion this scenario would be far away from happening and this is why I left this message:

In my opinion you don't have to be scared AI dominating social media. In my opinion the concept of social media is that people can connect with other human beings, sharing their interests, ideas and having great conversations. AI might join social media, but people might leave and find something else. Because the true benefit is that you can exchange personal relationships.
It may happen that people will have more face to face contact again.

Doing some research

I thought I was completely right – but then I did some research and found an interesting article about AI and the creation of texts.

It describes a program from a Californian university program called GPT2, which started with word prediction. The program was trained to predict the next word in 40 GB of the internet text. The developers made progress and learned the machine to create logical sentences on it’s own. In the beginning the program forgot the start of the sentences, due the lack of enough memory, and the paragraphs made absolutely no sense.

But then

In an experiment, the software was given about two sentences saying that a group of English-speaking unicorns had been discovered in the Andes. The program continued the crazy story, invented experts who were quoted, reported about details of the expedition, about DNA samples or possible references to aliens. The AI thus wrote one of those countless nonsense stories that can be found today in the lowlands of the social network and which are also clicked on abundantly.

Researcher mentioned that the program is already good enough to score a moderate result in higher education (e.g. language test). They also added that the software will be kept safe, since other people could use it for “bad” purposes.

Source

Personally I couldn’t believe that the development of KI is already that far. I don’t want to scare people around here but I thought it is an add-on what @neavvy already pointed out in his post.
In case you are interested in more information you can find the website project here
Also visit @crypto.piotr and @neavvy that help Steem to be an unique experience.
I would like to know what you guys think, so leave a comment below :)

---

I hope that you guys had a good start in the week. In case I add some valuable content I would appreciate an upvote and make sure you hit that follow button for more informative content.

Cheers,
M