Blog

The Oldest AI Blog in the Czech Republic

We have been writing about artificial intelligence since 2017. 1000+ articles, thousands of pages of thoughts, experiments, and reflections. No sensationalism, no ads.

Tag Filter: Experiments × cancel

Browse by Topics

Showing 24 of 120 articles

DeepNude – an erotic application with neural networksArchive 2019
1 min

DeepNude – an erotic application with neural networks

A team of programmers from Estonia decided to fulfil their teenage dream and create an application that can calculate the areas covered by clothing. The result is a revealed...

Read
3 libraries that will help you interpret your modelArchive 2019
2 min

3 libraries that will help you interpret your model

A dry definition tells us that model interpretability is the ability to approve and interpret the decisions of a predictive model, allowing for transparency in the decision-making process. Simply put...

Read
God, they've turned against us! The robot uprising has begun.Archive 2019
1 min

God, they've turned against us! The robot uprising has begun.

Let’s start charging the global EMP – it’s the only way to stop them! https://www.reddit.com/…/comments/c0s0xl/maybe_maybe_maybe/…

Read
Neural Network Connected to Monkey Brain Generates Bizarre Images!Archive 2019
1 min

Neural Network Connected to Monkey Brain Generates Bizarre Images!

Harvard scientists attempted to connect a neural network to a monkey brain so cleverly that their XDREAM algorithm morphed remarkable images. If monkeys dream, these images are what the monkey cells dream about...

Read
The New Ferret Brain from Open AI Has Given Us Room for Further Experiments!Archive 2019
3 min

The New Ferret Brain from Open AI Has Given Us Room for Further Experiments!

And here it is! Open AI has released another slightly larger model, this time version 345M. So instead of the brain of a small bird with 117 million neurons…

Read
Open AI Comes with Another Interesting Project – MuseNet!Archive 2019
1 min

Open AI Comes with Another Interesting Project – MuseNet!

In this experiment with a deep neural network (using the same universal technology as GPT-2), instead of text, it attempts to predict musical notes. This is nothing new in AI,…

Read
Artificial Intelligence and The Hitchhiker's Guide to the Galaxy: 'What is the fundamental question of Life, the Universe, and everything?'Archive 2019
3 min

Artificial Intelligence and The Hitchhiker's Guide to the Galaxy: 'What is the fundamental question of Life, the Universe, and everything?'

What happens when we let artificial intelligence process the pentalogy of The Hitchhiker's Guide to the Galaxy by English writer Douglas Adams and ask it: 'What...

Read
Nietzsche Responded with AI to the Question: 'What is the Meaning of Human Life?'Archive 2019
1 min

Nietzsche Responded with AI to the Question: 'What is the Meaning of Human Life?'

Yesterday, I conducted a small experiment. I took my favourite model from OpenAI, called GPT-2. Imagine a vast artificial neural network the size of a bird's brain (with over 100 million parameters) trained on millions of interesting articles written by humans...

Read
Artificial Intelligence Transformed Star Trek from 1993 into Full HDArchive 2019
1 min

Artificial Intelligence Transformed Star Trek from 1993 into Full HD

Is this the new future of remasters? One Star Trek fan decided to realise their idea for an effective remastering of Deep Space Nine (DS9) from...

Read
Our Team Gave a Lecture at VŠE for the Student Society 4FIS!Archive 2019
1 min

Our Team Gave a Lecture at VŠE for the Student Society 4FIS!

Today we had a lecture at VŠE for the student society 4FIS. Some students had very nice questions and insights. Thank you for the pleasant atmosphere – you were...

Read
GPT-2 Interviews Alpha Industries Ltd.Archive 2019
5 min

GPT-2 Interviews Alpha Industries Ltd.

I am still enchanted by the possibilities of the GPT-2 model. Today, I fine-tuned the system for questions and answers. I provided it with information about our company, Alpha Industries, and devised four questions, giving it the correct answer to three. I was curious to see how it would respond to the fourth question, and I must say, the model surprised me once again!

Read
Artificial Intelligence Suffers from Human Prejudice!Archive 2019
1 min

Artificial Intelligence Suffers from Human Prejudice!

Amazon tried an AI-powered recruitment programme but found that the software favoured men. A Microsoft chatbot transformed into a sex-obsessed, neo-Nazi machine in just 24…

Read
The AI Model GPT-2 Generated an Article About Alpha Industries! What’s the Result?Archive 2019
2 min

The AI Model GPT-2 Generated an Article About Alpha Industries! What’s the Result?

I had the much-discussed AI model GPT-2 generate an article on the topic: “I love Alpha Industries – the best company in the world that...

Read
Scientists have created a language neural network that approaches human quality. It is said to be so good that its authors are afraid to publish it!Archive 2019
1 min

Scientists have created a language neural network that approaches human quality. It is said to be so good that its authors are afraid to publish it!

Experts from OpenAI have proudly announced another technological success – the language model GPT-2. In the article's headline, they claim that its authors are afraid of it…

Read
Hopfield's Neural Network is More Powerful When it Sleeps and Dreams!Archive 2019
1 min

Hopfield's Neural Network is More Powerful When it Sleeps and Dreams!

The article discusses how allowing Hopfield's neural network to have a proper sleep and dream enhances its performance. Scientists drew inspiration from the sleeping mammals during the development of this...

Read
Emotion Detector or How We Defeated IBMArchive 2019
1 min

Emotion Detector or How We Defeated IBM

Time and Space: 26.2.2019, 18:00 – VŠE Žižkov, SB 321 Is artificial intelligence and virtual assistants still a distant future? We invited Honza Tyl to prove otherwise...

Read
I bring bad news. We have improved the Czech and English emotion detector!Archive 2019
1 min

I bring bad news. We have improved the Czech and English emotion detector!

We have modernised the algorithms, so you can now see which words the neural network places greater weight on (the AI attention button in the bottom right). We also used much larger and more diverse datasets for training...

Read
Web Element – Lecture on AIArchive 2019
1 min

Web Element – Lecture on AI

Are you interested in teaching your computer to understand written text? How we wrote a programme with the capacity of two cockroach brains and defeated IBM with it? How our…

Read
Doctors Yearn for Surgical Robots Equipped with a Sense of TouchArchive 2019
1 min

Doctors Yearn for Surgical Robots Equipped with a Sense of Touch

Robot-assisted surgeries are the future of medicine. However, current machines still have a significant disadvantage. “Robots offer high-definition imaging and fine control, but I can’t feel if I’m damaging something by pulling on tissues…

Read
Unsupervised MT - Machine Learning Without a TeacherArchive 2019
1 min

Unsupervised MT - Machine Learning Without a Teacher

It is still more of a concept, with relatively weak results, but I am surprised that it works at all. Let’s take a look at machine translation. Traditionally…

Read
Publications on Machine Learning and NLP Written in 2018Archive 2019
1 min

Publications on Machine Learning and NLP Written in 2018

In this article, you will find a number of statistical graphs regarding what has been published in this field, in which journals, how much each author has published, and so on.

Read
AI Helps Researchers Predict Heart Attack MortalityArchive 2019
1 min

AI Helps Researchers Predict Heart Attack Mortality

Acute myocardial infarction (AMI) is becoming an increasingly common condition. It is expected that by 2035, roughly half of the adult population will suffer from some form of this disease….

Read
And here we are at the end of the year, along with the evaluation of our Toxic competition. Thank you for your significant involvement in AI development.Archive 2018
1 min

And here we are at the end of the year, along with the evaluation of our Toxic competition. Thank you for your significant involvement in AI development.

During the test, you tried nearly 1500 attempts. And roughly 100 times you disagreed with the statements of our detector. It must be said that sometimes you found it hard to believe,…

Read
By tomorrow (31.12.2018) you can compete to catch our detector of insults and toxic wordsArchive 2018
1 min

By tomorrow (31.12.2018) you can compete to catch our detector of insults and toxic words

Tomorrow we will ceremoniously draw one of you to receive a few symbolic prizes. 😊 PS. Don't forget to leave your contact details (email, or perhaps a screenshot...)

Read