5 likes
AI painting
How does the exhibit work?
Move in front of the screen. Do you see yourself on the screen? It seems like you are part of the painting.
What is the science behind it?
How would you look in a painting by a great artist? You will discover that here through artificial intelligence (AI).
AI allows machines to perform tasks independently and learn through lots of practice. In this setup, the powerful supercomputer Genius learned the style of some famous paintings. He did so using images of them. This knowledge was put into an AI model. And that model is applied to the images filmed by the camera.
When you move, you also move on the screen. So the conversion from you to painting-you happens in real time. In the blink of an eye, the very best Picasso, Munch or Hokusai appears.The most beautiful? Sure, because you're on it!
When you look up more about AI, you may come across terms that are not so easy. That is why we have created a glossary for you.
AI: Machines that can learn and act autonomously.
AI model: A programme that can recognise patterns and make decisions with the help of algorithms (see below).
Algorithm: A mathematical formula. In programming language, it is an instruction, a piece of code, that helps analyse data.
Bias: Bias in AI occurs when the data (the examples) an AI system learns from do not accurately reflect reality, causing poor (sometimes even racist or discriminatory) results.
Deep learning: Machines that use artificial 'neural networks' to perform complex tasks. Such a neural network is inspired by how the human brain works. Deep learning is a form of AI and more specifically of machine learning (see below).
Machine learning: All techniques by which machines improve themselves in performing tasks through practice.Machine learning is a form of AI.
Training: Teaching an AI system things by giving it data (examples).
AI pictures
Having words converted into an image is getting easier and easier. Imagine, I want an image of 'Tyrannosaurus rex eating a biscuit'. Then I give that text to an AI system like DALL-E or Midjourney and it will turn it into an image.
Nice! But you also have to be careful. There are a lot of these pictures going around on social media that look just like the real thing, but are actually fake. Sometimes people use them to make you believe things that didn't really happen at all. Tip: by paying attention to small details, you can often tell that the images are fake. For example, you can count the fingers or look at the ears or teeth. There are often flaws in those.
AI at ChatGPT
You may have already heard of ChatGPT. That's a chatbot that uses AI to have a 'human' conversation. It learned that by being given lots of examples. In this case: things written by humans on the internet. We also call this 'training' AI.
Not everything written on the internet is true. And ChatGPT was only given examples until January 2022. So occasionally ChatGPT may make mistakes. We wouldn't let ChatGPT do your homework. But looking for a fun topic for a talk? ChatGPT can feel free to brainstorm about that!
How does the exhibit work?
Move in front of the screen. Do you see yourself on the screen? It seems like you are part of the painting.
What is the science behind it?
How would you look in a painting by a great artist? You will discover that here through artificial intelligence (AI).
AI allows machines to perform tasks independently and learn through lots of practice. In this setup, the powerful supercomputer Genius learned the style of some famous paintings. He did so using images of them. This knowledge was put into an AI model. And that model is applied to the images filmed by the camera.
When you move, you also move on the screen. So the conversion from you to painting-you happens in real time. In the blink of an eye, the very best Picasso, Munch or Hokusai appears.The most beautiful? Sure, because you're on it!
When you look up more about AI, you may come across terms that are not so easy. That is why we have created a glossary for you.
AI: Machines that can learn and act autonomously.
AI model: A programme that can recognise patterns and make decisions with the help of algorithms (see below).
Algorithm: A mathematical formula. In programming language, it is an instruction, a piece of code, that helps analyse data.
Bias: Bias in AI occurs when the data (the examples) an AI system learns from do not accurately reflect reality, causing poor (sometimes even racist or discriminatory) results.
Deep learning: Machines that use artificial 'neural networks' to perform complex tasks. Such a neural network is inspired by how the human brain works. Deep learning is a form of AI and more specifically of machine learning (see below).
Machine learning: All techniques by which machines improve themselves in performing tasks through practice.Machine learning is a form of AI.
Training: Teaching an AI system things by giving it data (examples).
AI pictures
Having words converted into an image is getting easier and easier. Imagine, I want an image of 'Tyrannosaurus rex eating a biscuit'. Then I give that text to an AI system like DALL-E or Midjourney and it will turn it into an image.
Nice! But you also have to be careful. There are a lot of these pictures going around on social media that look just like the real thing, but are actually fake. Sometimes people use them to make you believe things that didn't really happen at all. Tip: by paying attention to small details, you can often tell that the images are fake. For example, you can count the fingers or look at the ears or teeth. There are often flaws in those.
AI at ChatGPT
You may have already heard of ChatGPT. That's a chatbot that uses AI to have a 'human' conversation. It learned that by being given lots of examples. In this case: things written by humans on the internet. We also call this 'training' AI.
Not everything written on the internet is true. And ChatGPT was only given examples until January 2022. So occasionally ChatGPT may make mistakes. We wouldn't let ChatGPT do your homework. But looking for a fun topic for a talk? ChatGPT can feel free to brainstorm about that!