5/5 (2)

In early 2021, I came across this advertisement for an application that uses AI in order to turn a picture into a small 10 second video. At the time, I scrolled past it with no interest until it kept being recommended on my social medias. I finally gave in a downloaded the application.

I had this picture of my grandfather as my background and thought it would be fun to try the application with it. I had just mentioned to my family how I thought we did not have enough videos of my grandfather, and his unique facial expressions. After watching a 30 second advertisement, the results were mind-blowing. It was like my grandfather was alive again through a picture that had been taken years ago.

The application, also available as a website, is called MyHeritage. With their service “Deep Nostalgia”, AI can convert photos of people either new or old, and animate them into a short 15 second video. Basically, “Deep Nostalgia works by using several pre-recorded videos of facial movements, consisting of a fixed sequence of movements and gestures, then using AI software, it will apply the one that works best for the photo you are trying to animate.” (MyHeritage, n.d.).

In my case, I found this service absolutely great. I still use it until this day to animate old pictures and play around with my camera roll. However, I can also see a potential issue that arrises from such a Generative AI concept. With any technology that has the potential to manipulate photos or videos, there is always a risk of using someone else’s picture with consent, and make it available to the public. For example, someone could use MyHeritage to create fake videos of people saying/doing things they never did, which could be harmful.

And again, the question of how freely should we let society use these Generative AI tools arises.

MyHeritage Deep NostalgiaTM, deep learning technology to animate the faces in still family photos – MyHeritage. (n.d.). MyHeritage. https://www.myheritage.com/deep-nostalgia

5/5 (1)

I have always had a passion for music. I started learning the trumpet and guitar at a young age, and still play until this day. Last year, I decided that I would try to get into music production, and give a go at creating my own song(s).

Therefore, I downloaded all required softwares, bought some gear, and ended up at my desk starring at an empty Ableton Live project, not knowing where or how to begin. It took me a very long time to get to know the software, come up with creative musical ideas and let alone assemble all of them into a backtrack that sounded good, and fit together. It got to the point where I pretty much ‘gave up’ for a week at some point, before coming back to it.

This year, with Generative AI taking over the way we do things, I heard about a new one by the name of SOUNDRAW. SOUNDRAW is an AI Music Generator which allows you to create your own music with the help of AI. You are presented multiple different “beats” or “soundtracks”, and can play around by increasing the tempo, and changing the “energy” of each segment. I am currently only able to access the free version, but the pro version also allows you to edit the tempo, length, structure, etc… You are also able to “create similar music” to a track you heard.

All in all, with SOUNDRAW, you are given a pre-made song soundtrack with which you can play around. This completely changes the game for music producers, who could now end up competing in the charts with artists who do not produce their own music, or use AI-generated soundtracks. This also rases a question in my head: Can someone who uses SOUNDRAW, or a similar AI Music Generator tool, get credit for producing music? What if an artist with a song in the top 10 global charts is found to have used AI Music Generator tools? Would it be a bad outlook on the music industry, or is it going to revolutionize the way people make music?

Personally, I have yet to make my own opinion as I believe that a successful music artist has worked hard to come up with creative ideas in a studio. However, it is also the case that some artists use “ghost writers”, or have producers produce beats/soundtracks for them. As a matter of fact, “Almost every major hit was written by two or more songwriters. In the 2020s, there has yet to be a single #1 hit written by a solo songwriter” (Pack, 2023). In the end, I think it would be the talent that artists have to write music that would be put at risk. It is just like asking ChatGPT to write a novel, and then publish it. It takes away the creative and hard work of other artists who do not use such AI tools.

If you are interested by SOUNDRAW, here is the link to their website: https://soundraw.io/. However, there are also plenty other AI Music Generator tools.

Pack, B. (2023, September 4). What Is A Ghost Producer? (And How to Become One). Mastering The Mix. https://www.masteringthemix.com/blogs/learn/what-is-a-ghost-producer-and-how-to-become-one