Recently I saw another article in a newspaper where the author held some misconceptions about intelligence quotients. This article made the connection to artificial intelligence which sparked my motivation to explain some things about quantifying intelligence. Intelligence tests are often subject to discussions. Intelligence test measure primary the performance an individual on said test. The result is called intelligence quotient (IQ). This argument is often used to invalidate the relevance of the IQ score. However, studies showed that the result is a good predictor of success in other areas of life. This psychometric test is a useful tool in diagnostic psychology.

Although they are mainly a tool of diagnostic psychology sometimes the term IQ appears in discussions about artificial intelligence. When analyzing intelligence there is the problem that there is no known method to quantify behavior and possibilities of cognitive systems. Because IQ is a famous measure to quantify intelligence people often come back to referring to IQ. If we had a number for intelligence we could optimize algorithms and artificial intelligence would probably a solved problem. As an alternative scientist try to evaluate cognitive systems via the environments in which they act. OpenAI uses Atari Games to measure the performance of their algorithms.

An IQ test has to be carefully designed, which is an expensive process. When lots of people perform this test you get a result where the distribution of the results is gaussian-shaped. The scaling is then chosen so that the expected value is 100. Using this knowledge how do we interpret an IQ score? An IQ score can be compared to the shape to see in which percentile of the population a score is in. By definition, the expected value 100 but a Gaussian distribution has another value: the standard deviation. For many tests, this is 15 e.g. Stanford-Binet and Wechsler adult. Cattell Culture Fair III uses 24. What you need to remember is that the test results are not comparable as the percentile is different depending on the test.

There are two reoccurring misconceptions about IQ tests. First: it is not possible to give someone an IQ without the individual taking the test. The IQ is the result in a test. Although a person might appear very intelligent or less so the results on the IQ test cannot be predicted with high confidence.

Second: Very high scores are not possible. Most tests are norm-oriented, which means that they are designed for the typical human. If you score very high or low the result is not as accurate.

What is the IQ score if every question is answered correctly? It is not possible to get IQ results which are above 160 points on Stanford Binet.

When people talk about IQ numbers combined with artificial intelligence they want to talk about a qualitatively new type of intelligence. A new type, which is on a next level, called artificial general intelligence or superintelligence. Because it does not exist yet it is hard to imagine. We may imagine a smart person or the artificial intelligence we know from the movies, but a super intelligence is far more.

Related and Recommended Reading:

Robert J. Sternberg: International Handbook of Intelligence

Raven’s Progressive Matrices

If you do a web search for Deep Learning, Machine Learning and all the differences you will often find a graph which includes all those three terms. I think this can be done more detailed. I did not find any graphic showing more than three layers so I did my own. Open image in a new tab for full image display.

The structure is hierarchically so everything e.g. Convolutional Neural networks are used in Deep Learning, which is a part of the Neural Network Landscape etc. Circles which cut other circles mean that a part of this field or method is also part of another field or method or their classification is ambiguous.

Please write me if you have any suggestions or don't agree of my choice of circles.

Update: Thanks for Jonas K. for pointing out that Evolutionary & Genetic Algorithms are not a subset of Machine Learning but algorithms used in Machine Learning. I updated the diagram accordingly.

Diagram showing AI terms
Diagram showing AI terms


In this post, I want to clarify my motivation for setting up this blog and the topics I am interested in writing about.

Why is this blog written in English?

I was born and raised in Germany. German is my native language. I am pursuing a career in a time where it is needed to connect with people around the world. Therefore I am using the language most people can speak and to improve my English skills.

Why this blog?

Previously I have been using the blogging platform Medium. I chose Medium because it is easy to maintain but I disliked the recent developments of the site. As I wanted more control over the platform where I am hosting my content I am now using a custom solution. You know those good-old-days when the Internet was the „wild west“ where surfing on the internet meant exploring the web? I remember a time when the internet was not full of advertisement and websites by companies. People wrote their own websites and even own browser games. I still remember one specific site which is an archetype of this type. In school, I was tasked to write about China and do some internet research on that topic. I discovered a site of one guy who was an expert on China. He set up his own website and shared his expertise. It was fascinating. Nowadays you rarely find pages like that. Sometimes I see some decade old pares explaining some math. I want to bring this experience back to you. If you like my posts please link them on your blog or website. I am happy to answer feedback e-mails and answer your questions.

I still follow some blogs using RSS/Atom technology. RSS and Atom are protocols which can be used by some applications called feed reader, which check websites. You then get a feed like on facebook but you have the control of the sources and there is no company involved. Sounds great? You can try it out by yourself. Just use the URL of this site in any RSS reader. I am using NetNewsWire 4.

When you look at the state of society, you will see that social media has changed us a lot. I believe that lots of changes are not healthy changes. Although I grew up with technology and still believe that technology gives us useful improvements I stopped using Facebook, Instagram and avoid Whatsapp. People often cite that these are tools they need to stay connected with friends and family, but they miss two things. Every minute you spend on using these tools you could do something different. For example, instead of browsing Facebook for half an hour, you could call a friend - an experience which connects you deeper than Facebook could. This absence of some activities because you chose to pursue other activities is called opportunity costs. I want people to get updates on my life, and share my ideas. Maintaining a blog keeps me writing and finishing my projects and work on ideas. In order to further decrease my Facebook usage, I need people to get updates on my life and thoughts, which I would previously post on facebook, from a different source. Hence this blog.


I am currently pursuing a career in artificial intelligence research. Digital transformation, start-ups, and computer technology in general are also my interests.

Technical things

Now some technical things. You can skip this paragraph if you are not interested.
 This blog is hosted on uberspace. The software for the blog is grav, which is a flat-file content management system. I am using the hpstr theme. I customized it to make it GDPR (Germans know the GDPR as „DSGVO“) compliant by removing externally loaded resources.

Little people know that the audio quality of AAC > MP3 at the same bitrate. You can get AAC files from iTunes with the suffix m4a at the iTunes Store for example. Today I used the application “Spek” to visualize the spectrum of the same rock song with both encodings in comparison. You cannot really notice the difference by hearing (unless you are an audiophile with good equipment and very good ears) but if you want the superior audio quality with smaller file size just choose AAC files. For most people, the top audible frequency is, depending on age, at 18–20 kHz. In the diagrams, you can see that mp3 files cut frequencies at the top.

m4a Spektrogram mp3 Spektrogram
Left m4a, right mp3.

How to find the best place to live using the Google Maps API

The problem

Assuming you travel a lot e.g. for your jobs and you want to reduce the time you travel. I wondered in which city you should rent a flat to stay between your trips. In my specific case the only important factor was the travel time. Regarding this, what is the best city? Luckily we can solve this problem with a computer, some programming skills and an internet connection. Some small real-world problems like this are the best way to improve your skills like in a code kata. In Germany, almost every city is connected by rails and tickets are quite cheap. This reduces the cities we have to consider as small towns don’t have a train station. The cities to consider can be picked from a map.

The solution

I wrote the script in python, but a javascript version should work the same way. One input file is a simple list of cities with the concert locations. The other one is a list of cities which are a reasonable to consider. Then the mean travel time with public transport is calculated for each home city. The results are sorted, so you get a ranking. In her case the best city is Cologne. Cologne is quite well connected, so is not that much surprising. You need an API code for the directions API from the Google developer section in order to Google Maps API. A free API key allows up to 2500 request a day, which is more than enough for this task. Using the API is quite easy, but the documentation for the Python API links to the Javascript one. The code for my implementation can be found here.