This is NOT Morgan Freeman! Scary!

The line between reality and deception is becoming blurred


Welcome to another issue of Practical AI, the newsletter that serves up news on artificial intelligence with a side of practicality.

This issue is about the darker sides of AI. Because let’s face it - who doesn’t love a little “help” from our robot overlords?

Cue Morgan Freeman…

This is not Morgan Freeman - AI & Deep Fake

The above video made its rounds in social media in 2021 and 2022 and was created by Dutch filmmaker Bob de Jong of

Despite being over a year old, I wanted to share this video and talk about the technology as it is very current right now (more on that below).

This video, more than anything, gives us a scary insight into what AI can do and what is to come.

This technology is blowing my mind!

But what happens when this falls into the wrong hands or is used for the wrong purposes?

I don’t want to talk politics in this newsletter, but in the context of today’s topic, I need to touch upon the Russian invasion of Ukraine.


Because Russia is already utilizing this AI technology in its propaganda.

Their output is not as great as Revel AI, for now... But I am sure they are scrambling to get their shit together, as they understand the power it holds.

This article/video on how Russia uses artificial intelligence and deep fakes in propaganda warfare is worth a watch. It is only 2 minutes (The Global News article)

In the Global News video, we are presented with another (scary) AI concept: Generative Adversarial Networks - GAN.

What a GAN is, is presented in this week's AI term of the week (further down).

The big question on my mind is: Are we prepared for what is coming down the pike?

AI firms warn that the world is not…

In the Morgan Freeman video, they used a voice impersonator rather than an AI.

However, there are AIs that focus on voice, and they are also scary in many ways. I think they could have added an AI voice to that video.

To avoid making these newsletters looooonger than they already are, I will cover voice AI next week.

Applications of AI video

I have no doubt you see the implications of AI used for the wrong reasons.

But there are use cases for work and business too.

And there is technology readily available that you can start using today.

It is not perfect, but I would say it is pretty darn good already.

And it is affordable too!

Two of the players in this space are Synthesia and Colossyan.

Both use real actors and people that has been digitized into AI actors and avatars.

All you have to do is to give it your script, choose your AI actor, add the voice of choice (multiple languages, accents, and dialects), edit the background to your liking (or use a template), and press the “Generate video” button.

5-30 min later, you have an amazing video created with a “real” human speaking the words you fed it.

It is so easy to do!

I made these two videos in a manner of minutes:

From the Synthesia AI

From the Colossyan AI

You can try making your own for free (and without a credit card) over at their respective websites:-> Synthesia -> Colossyan 

How can you take advantage of this technology in your business or work?

On top of my mind, I have these ideas:

  • Onboarding videos

  • Spice up boring power point presentations

  • Explainer videos

  • Welcome to our website videos

  • Novelty videos

  • Courses

What do you think of this? Hit reply and tell me!

I just launched my blog on Artificial Intelligence called The Future Handbook. It is a blog focusing on AI deep dives, case studies, AI industry interviews, AI tool reviews, and how-to guides on various AI technology and tools. Basically an extension of this newsletter. Check it out here.

AI experiment of the week

The ChatGPT from OpenAI that I covered last time, is continuing creating panic and uproard around the globe.

Whereas “the robots” took away a lot of manual jobs in factories years ago, it seems lawyers are starting to understand that their golden cash cow may be up for disruption next.

A friend of mine (Hey M) told me how she used ChatGPT to go over a contract for her business with a client and asked the AI if there were any sections in the contract that was in her disfavor.

She prompted the AI and copy and pasted the contract, and hit “Enter.”

She got excellent feedback from the AI and said she was alerted to several parts of the contract that was not in her favor.

Naturally, I had to find out for myself.

So I replicated the process on my own.

Here is what I did:

  • Googled for a “master service agreement example”.

  • Randomly chose one of the top results in Google.

  • Logged in to ChatGPT.

  • Asked the AI this question (prompt): Here is a contract from a client. Suggest issues to bring up to make it more fair for my company or any other issues to bring up for changes.

  • Copy/Paste the contract below the prompt.

  • Hit Enter

This is what the prompt looks like

And here is the reply (parts of it)

I wrote a full article about this experiment on my blog The Future Handbook.

The application of AI in the legal sphere is pretty clear. And I would guess there are plenty of other use cases in that industry as well.

But what about the implications?

It is easy to imagine the implications. Having customers with power and knowledge at their fingertips will be a huge disruptor in the field!

But I can also imagine law firms utilizing AI. It can either increase the productivity of their paralegals or save money by firing half of them…

NB! I have no affiliation with the company in this example "Mercy Corps" and the contract was found using a Google search. This example is not in any way created to hurt or damage Mercy Corps. It is just an example for entertainment purposes.

Practical use of AI in the healthcare industry

It has been said that ChatGPT does not have any application in the healthcare industry.

This is wrong…

I found a very cool use case on Twitter, where a doctor demonstrates how he utilizes ChatGPT to write an application to an insurance company for additional treatments/tests for one of his patients.

It is pretty amazing, and he states clearly how this benefits him and how other doctors can do the same.

(I think this is a very American problem/use case, as we over here in Europe do not need to worry about insurance companies paying our medical bills or not).

Do you know anyone in the medical industry?

Forward them this email and encourage them to sign up :)

PS! There are lots of applications of AI in the medical industry and I am sure I will cover many of them in this newsletter as time passes.

AI image of the week: Fantasy steampunk-style butterfly

I made this image using the AISEO Art AI.

AI definition of the week: Generative Adversarial Networks (GAN)

Warning: this week's definitions are complex and hard to simplify! I tried… Also, to understand why I decided on this definition, it is related to the Global News article mentioned earlier in the newsletter.

GANs are a type of machine learning model that is used to generate synthetic data. They consist of two neural networks, a generator, and a discriminator, that are trained together. The generator produces synthetic data, while the discriminator tries to classify it as either real or synthetic. The two networks are trained simultaneously, with the generator trying to improve its synthetic data to be more realistic and the discriminator trying to become more accurate at distinguishing real from synthetic data. The result is a generator that can produce synthetic data that is similar to the training data. GANs have been used to generate synthetic images, audio, and text.

Sorry, we need three (four) definitions this week: Synthetic data

Synthetic data is artificially generated data that is designed to mimic real-world data. It can be used for various purposes, including testing machine learning models, training machine learning models, and augmenting limited real-world data sets.

Generator & Discriminator

A generator is a neural network that is trained to generate synthetic (AI) data that is similar to some training data.

A discriminator is a neural network that is trained to classify whether a given input is real or synthetic (AI)

Sorry for the spoon-feeding.

And don’t worry; I had to read these definitions six times before I understood them… I so regret choosing this as the definition of the week!

ChatGPT created these definitions.

I asked for the simplified versions, but it is a very technical and difficult topic (sorry ).

Quick bites - interesting news articles on AI

Do you have any questions about AI? Please send them my way, and I can try to answer them in these newsletters. Just reply to this email with your questions!

That’s all for this week!

What did you think about this issue? Just hit the reply button and let me know. I am on the other side of this email!

And if you like it, please tell your friends and colleagues about it.

PS!Why did the AI get in trouble at school?




Because it kept trying to take over the class!

This is not the end, it is where the fun begins!


or to participate.