Online bootcamp Art & AI

Program:
—————–

Since the legendary Dartmouth Summer Research Project, which marked the official birth of the field of artificial intelligence in 1957, the development of AI has been a journey full of unexpected twists and turns. It is a story with as many AI summers as there have been AI winters. Recent breakthroughs due to highly improved computational power and optimalisation of algorithms have seen numerous unprecedented and groundbreaking applications.

New AI systems and software are introduced on a daily basis and used at an exponential rate. From ChatGPT4, DALL-E, Stable Diffusion, MidJourney, Sora, MusicLM, MusicNet to open-source libraries and systems such as TensorFlow, Keras or PyTorch; they all show impressive results and an even more impressive future ahead. It is an evolution which is poised to have an everlasting impact on practically all artistic disciplines.

But as fast as these new tools are being used, their development is not completely unexpected. Most AI techniques are based on principles and techniques with roots in information theory, mathematics and data science. One of the goals in this course is to give an insight into these fundamental techniques and methods from a layman’s perspective. In this way the summer course will provide you with a bottom up and deeper conceptual understanding of the various AI models and systems. As such it will also enable you to incorporate AI in a meaningful and effective manner in your own artistic practice.

This bootcamp is intended for artists and creative minds alike. No previous knowledge is required. You also do not need to have coding experience to follow this bootcamp.

This bootcamp is hosted and organized by the Center-MCS (Center for Music, Coding and Sound Design).

On a practical side the bootcamp is organized into three sessions:

Sesssion 1: (17/01 : 19u00 – 21u30 CET)

In this first session you learn to get acquainted with the basic principles and key concepts of AI and it’s usage in the artistic domain.

Concrete program:

-Foundations of AI

  • Origins of AI and Art and AI
  • Subfields in AI
  • Forms of (A)I
  • Limitations of AI
  • Learning strategies
  • AI and data
  • Fundamental principles of (artificial) neural networks (ANN) and Deep Learning

Session 2: (18/01: 13u00 – 18u00 CET)

The second day of the bootcamp we’ll dive further into the key concepts used in AI and how to design and setup your own AI systems. This will be followed by applying this knowledge to the domain of music and sound art using MaxMSP.

-Key concepts in AI & Art

  • ANN’s and Deep Learning in artistic context
  • Understanding the mechanisms behind generative AI

-How to design and use custom AI systems

  • Design of use problem
  • Data management
  • Choice/design of AI algorithms/model

-Practice Session 1: AI applications in music and sound (art)

  • Introduction into MaxMSP
  • Introduction into AI libraries in MaxMSP
  • Elaboration of practical examples of using AI in MaxMSP for electronic music
  • General strategies how to use (open source) AI models and techniques for electronic music composition and sound installations

Sesssion 3: (19/01: 13u00 – 17u00 CET)

For the third day of the bootcamp we’ll be covering topics related to ethics, copyright, authorship and responsible use of AI in artistic context. This will be followed by a practical session focusing on the domains of visuals and multimedia art.

-Implications of AI for the arts

  • Ethics of AI
  • Authorship, copyright and AI
  • Future of (human) creativity
  • Building responsible AI systems for artistic practices

-Practice Session 2: AI applications in visual and multimedia art

  • Introduction into Python and google Collab
  • Practical example using Keras Python library
  • Principles of latent diffusion image and video generating AI models
  • Basic principles of prompt engineering
  • Practical examples of working with Stable Diffusion
  • Installing and using open source image generation models e.g. Stable Diffusion

Requirements:
————————-

Laptop and an internet connection. This bootcamp does not require any previous coding or sceintific knowledge. All techniques and principles will be explained from the start. You’ll need a laptop to follow this course. As to software we’ll be using open-source software Python and free libraries, Google Colab and MaxMSP. As to MaxMSP you can follow this course using a demo version of this software.

Planning and duration of this bootcamp:

——————————————————————-

12,5 hours in total. This 3 day bootcamp runs from 17/01 till 19/01. On 17/01 from 19u00 till 21u30 CET, and on 18/01 and 19/01 from 13u00 till 18u00 CET.

Cost of this bootcamp:

—————————————

Early bird ticket: 195 € (runs till 20/12/2024)
Regular ticket: 250 € (after 20/12/2024)
Student ticket (*): 175 €
A maximum of 10 participants can join this bootcamp
///
(*) For a student ticket please contact us via info@valeryvermeulen.net and include a scan of a valid proof of student enrollment. Upon validation of the proof of enrollment you’ll receive a discount code that you can use when purchasing your ticket. Students from a University or a University College are eligible for student price.

Enrollment / additional info:

————————————————

You can enroll via the following eventbrite link:

Once enrolled you’ll receive a Zoom link and reminder of this link in the week before the start of the bootcamp. On the day of the start of the bootcamp you’ll get to the Zoom meetup as of 18u45.

-If you would like to have more info on the bootcamp please send an email to info@valeryvermeulen.net.

About the teacher of this bootcamp
————————————————————–
Teacher of this bootcamp is Valery Vermeulen. He’s a Belgian electronic musician, mathematician, visiting professor at the Music Production department of the KASK-Conservatorium School of Art in Gent, guest lecturer the Master Artscience KABK Den Hague and researcher at the Department of Mathematics and Computer Science at Antwerp University (BE). He holds a Phd in pure mathematics from Ghent University (BE) as well as a masters degree in Music Composition at the Royal Conservatory of Ghent (BE). Apart from his artistic and teaching activities Vermeulen has been working professionally as a data scientist/statistician and AI/ML engineer since 2005.

In his work Valery focuses on he intersection between electronic music, data sonification, AI, data visulaisation and theoretical physics. His work has been widely shown in Belgium as well as abroad and was featured in numerous magazines and news outlets such as BB6 (UK), Music Radar (US), Tech Radar (US), The Wire (UK), Electronic Sound Magazine (UK), New Scientist (NL), Neural Magazine (IT), DJ Mag (US), Mixmag (UK, Asia and Korea), Trax Magazine (FR), Deutschland Funk (DE), Radio Eins (DE), Radio Canada (CA), Yle Radio1 (FI), The Wire UK on Air (UK), VRT Radio1 (BE), VRT Radio2 (BE) and VRT Max (BE) to name but a few.

With his project Mikromedas AdS/CFT 001, merging electronic music and black hole astrophysics, Vermeulen received an Honory Mention from the prestigious S+T+Arts Prize Ars Electronica, a yearly prize for artscience projects issued on behalf of the European Commission.

More info on his background can be found via www.valeryvermeulen.net.