AI to the test

Last week, I did a 5-day « write-athon” on AI.

Using, you can have “CONTENT CREATION MADE EFFICIENT”. is this real ? is it a dream ? well let’s try out and see where it goes with my next subject (spoilers !) : The technical aspects of Google Duplex.

Here is my question to the machine : “Google duplex creators and the technical aspect of google duplex”. I Could have Googled it, gathered some sources and written about. No, not in 2018 : I can use AI to do all of this.

After providing the email The page says : “Thank you! You will receive an e-mail as soon as your query has been processed. During alpha, this might take up to a few hours, but usually only takes a few minutes!”

Well after a few minutes (! Not the hours announced), the result is here and I got the email: “Your article ‘Google duplex creators and the technical aspect of google duplex’ has been written”

So here it is, my fake article from today’s POC:

First of all, it is a hastle copying the text ! hence the print screen.

Second, ai-writer is basically “googling for you”, if that makes any sense and picks up sentences for you. Not that smart he?

But let’s be positive, I have kind of summary plus the sources which really helps as a second source of information. And it kind of works as an article if re-ordered with my non-artificial assistant: me.

I let you be the judge if this looks smart!

“Google Duplex is impressive and surpasses Siri’s current functionality. Duplex is the culmination of Google’s work on machine learning and natural language technology – in other words, it sounds and comes across like a real person.

So far, Google Duplex is a tech demo not a product demo. But it will be tested with the assistant in summer 2018.

Its performance is very impressive on a technical level and it is easy to appreciate from a “good grief, that’s awesome” perspective, but Duplex also has some implications that you may not have considered:

It could be 911 responder and shave minutes of response time for emergency crews

Google might be about to put a lot of people out of work

It might replace Tier 1 call centre team

Business might be spammed by Google Duplex

Many people are already freaked out by the implication of AI – from its impact on employment to its lack of human discretion – and suspicious of Google because of the amount of data it holds.

Google CEO Sundar Pichai premepmtively addressed ethics concerns, saying : “it’s clear that technology can be a positive force and improve the quality of life for billions of people around the world.”.

Google Duplex is definitely something to keep an eye on in the future because the possibilities are endless.”

Time to write the article: 5 minutes. 220 words. FIVE MINUTES. TWO HUNDRED AND TWENTY WORDS! “And for five hundred words”, you might wonder, probably five minutes as well.

To respect the sources for this demonstration I will refer to them below. After all, this is cheating and ethic (only that) prevents me from taking any credit from it. In terms of pure creation, we are still far from having a proper original creation. After, all it did is copy sentences from other articles. I did have to re-order them to make them readable and put them in a literary form.

But I guess, if I did not mention you the trick, like for the Duplex demo, you would not have guess by reading it I spent just five minutes on it.

Missing also the whole “Google Duplex Creators” part, so I guess I will have to do it myself. Unless Google keeps it a secret!

Though there are some false friends in the results, I would say the assistance in great for providing sources to scout information from.

So soon, my friends, long will be gone the hours of reading books or article, same for writing. A simple “OK google, write me an article on a trending topic and publish it to my blog”, the machine will comply with a “it is done Yann, do you want we to read it for approval?”…

The whole creation aspect of Human mind may be at stake. First writing, music then what? Movies painting… Your next favorite Youtuber might be a machine… If we are not told, if we do not care, after all, there might not be much difference in the end.

Again, the machine cannot be hold responsible. If we want to sustain creativity (meaning human), we need ethics to respect and agree with decent controls over what we can or cannot do with this new technology.