This is my CV.
More generally, this website is designed as a simple way to present complex information with arbitrary levels of detail.
It is often difficult to know how much information one should put in a presentation: Too much and the reader loses interest, too little and they don't learn what they want.
Additionally, different people focus on different aspects, exarcerbating the problem.
This style of presentation solves this problem: all information is here, but it is left to the user to explore it to any depth they like. Sections that are of no interest can be skipped, while anything that catches your interest can be perused in great detail.
Copyright: Florian Dietz.
Feel free to use this website as a blueprint for your own presentations so long as you mention my copyright in the source code.
It is easy to change this presentation's content using only HTML and (very little) CSS, without requiring any changes to the Javascript part.
I wrote this as a learning experience, so no guarantees of correctness are made.
Criticism and suggestions are always welcome at floriandietz44@gmail.com
Welcome to my interactive CV!
Navigate with the mouse wheel to zoom in on interesting slides.
Use the arrow keys to explore the slides sequentialy.
Click on the icon in the top-left corner for more information.
I founded and run the startup elody.com.
Due to the nature of startups, my responsibilities change from week to week.
I can't talk about what I did at Palantir due to the company's strict privacy and security regulations.
I worked on Big Data and Machine Learning problems for the VW Data Lab.
The broad variety of tasks encountered there, combined with the dual-nature of the company as both a start-up and a daughter of a large and old company, allowed me to gain experience in a large variety of areas very quickly.
After proving myself, was actually payed the salary of a PhD graduate despite only having a Master
I received a letter of reference when leaving Volkswagen Datalab.
Unfortunately, it is written in German:
Volkswagen hires slowly and uses subsidiaries to outsouce its workers.
I rose through these very quickly before joining VW itself.
I received a letter of reference after switching from Autovision Gmbh to Volkswagen proper.
Unfortunately, it is written in German:
This job was a paid research project that also resulted in my Master Thesis. It was funded by the DFKI.
The task was to create a semantic linking between documents in different languages.
This involved NLP and ML.
The goal was to make it possible for international students to read an academic text in English while being able to quickly look up difficult sections in their mother tongue.
This job was offered to me by Giesecke&DevrientGiesecke&Devrient is the world's second largest supplier of banknotes due to my performance in the national informatics competition.
It was a job over the semester break.
I tutored for several students from the area while working for "Studentenring", a company that brings tutors and students together.
The students' ages varied from elementary school to end of high school.
The subjects were English, Math and Physics.
I was a private tutor during high school, for around half a dozen people in total.
My focus was on Math, the natural sciences and English, but also occasionally other subjects.
I have been interested in AI since before going to university.
AI is my passion.
I have worked on my own theories on AI since my second to last year of high school, and soon afterwards began gathering ideas for developing an AI of my own:
I have been interested in the Cognitive Sciences and Philosophy of Mind since early adolescence (age 13), in tandem with philosophy.
My goal has not changed since then: figure out how thinking works. I want to understand the nature of thought and the mind.
There are two ways to do this: Psychology and AI, both of which I study.
My personal skills and abilities lend themselves more to AI than psychology, and AI has more applications than psychology, which is why I focus mostly on AI.
A particular boon I gained from my training in the cognitive sciences is my unusually strong control over my emotions.
I took all available courses related to AI and ML since the start of my Bachelor studies, as well as online courses.
In my Master studies, I added NLP as another, related focus.
Besides this, I read many books on AI and related topics in libraries.
Most importantly, I spent my free time conducting research on more unusual approaches to AI and how they fare compared to common ones. This gives me an improved understanding of the pros and cons of different algorithms.
Reached the final round in the 29. Bundeswettbewerb Informatik/BWInf (german national informatics competition), along with only 26 others nationwide.
Received the award for particularly creative solutions in the final round, which is given to only one contestant per year, in all of Germany.
This was my first and only attempt at the competition, and all my programming skills prior to this were completely self-taught, since I was still in high school at the time.
Membership in Junge Akademie, an honor society for extraordinarily talented and engaged students at TUM.
Within the Junge Akademie: Member of sub-group aiming to facilitate social behavior among students.
Received a term paper award in my last year of high school.
This was for a computer simulation I created assist with solving simple physics problems.
Notably, I had no formal computer science education at this point and taught myself how to do this.
Received 3rd price in a regional competition of Jugend Forscht that I had entered on a lark in my last year of high school.
My presentation was the same program that I received a term paper award for.
Study at University of Saarland (Saarbrücken, Germany).
GPA: 3.6the grade according to the German rating system was 1.4
I chose to switch to this university from TUM because the
DFKIDeutsches Forschungszentrum für Künstliche Intelligenz (German Research Center for Artificial Intelligence)
This institute works closely together with universities and the private sector, acting as a bridge between both.
is located there and my focus was as always on AI.
I received the Graduate School scholarship, which covered all expenses of my studies.
As a member of the Graduate School, I received special courseworkWe had something called "Research Immersion Labs", which were designed to provide the environment and tasks of a PhD student to Master students, in order to prepare them for a PhD. and was on the fast-trackMy Master was supposed to take 3 instead of the usual 4 semesters and segue fluidly into a PhD position by skipping the Master thesis itself. I instead decided to take the 4th semester after all and finish my Master properly, since otherwise I would have had a risk of ending up without even a Master title if the PhD did not work out for any reason (which was always a possibility - I actually lost a chance at a PhD because the department that wanted to offer it ran out of funding). to getting a PhD.
My Master Thesis involved NLP and ML and was coupled to a job.
Unfortunately, I did not find an interesting-enough topic for my PhD despite the presence of the DFKI, so I decided to go into industry instead and gather some practical experience while staying on the lookout for interesting PhD topics.
Study at
TUMTechnische Universität München (Technical University Munich)
Internationally renowned
(Munich, Germany).
GPA: 3.4the grade according to the German rating system was 1.6
I Finished my studies in 3 semesters instead of the usual 6I was hurrying so that I could get to the more interesting parts of my studies sooner..
I focused on AI and ML, and took psychology as a minor to broaden my understanding of the cognitive sciences.
My Bachelor thesis was a theoretical work about the Technological SingularityThe technological singularity is the hypothesis that the invention of artificial superintelligence will abruptly trigger runaway technological growth, resulting in unfathomable changes to human civilization. According to this hypothesis, an upgradable intelligent agent (such as a computer running software-based artificial general intelligence) would enter a 'runaway reaction' of self-improvement cycles, with each new and more intelligent generation appearing more and more rapidly, causing an intelligence explosion and resulting in a powerful superintelligence that would, qualitatively, far surpass all human intelligence..
I went to high school at Apian Gymnasium, in Ingolstadt, Germany.
I had Maths and Physics as my majors and got a GPA of 3.7the grade according to the German rating system was 1.3.
I was a member of the student union.
I participated in two of the very first online courses ever offered, the success of which caused today's popularity of online courses.
These courses were a course on AI by Peter Norvig and Sebastian Thrun, in which I received a 98.8% score, and a course on ML by Andrew Ng, in which I received a 100% score.
These courses were some of the most interesting and useful courses I ever had and blow regular university courses out of the water.
In the course of theorizing about AI, I taught myself metaprogrammingWriting a program with the ability to write other programs.
Instead of writing a problem to solve a problem directly, one writes a program that either modifies itself or generates a number of new programs in order to solve an entire class of problems..
Surprisingly, this has proven to be a very useful skill even outside of AI, as the skills and mindeset I acquired while learning this allow me to automate many tedious programming tasks at work that my coworkers have to do manually.
I am interested in psychology for the same reason that I am interested in AI.
In addition to taking a Minor in Psychology during my Bachelor studies, I have read many books and articles on the subject in my spare time.
This is not a purely academic interest to me: I habitually analyze myself and others for signs of cognitive biases, in order to think more clearly and make fewer mistakes.
Game Theory is of particular interest to me because it is useful for AI.
However, I developed an interest in Game Theory beyond its usefulness for AI and came to appreciate it for its own value.
I like interpreting everyday situations in terms of Game Theory. Doing so always teaches me something new either about the way humans act irrationally, or the way traditional Game Theory fails because it does not take into account some more complex ideas that influence human decision making. In either case, I learn something useful (example)did you know that many apparently irrational actions of humans are actually results of group selection and are perfectly rational when viewed as Game Theoretical pre-commitment policies enforced at the genetic level? it's pretty neat..
I was planningwork on my AI research project is proceeding more slowly than planned and I have noticed that I still have a lot to learn in the area of business understanding before it is worth taking the risk of founding a company, so I have put this plan on hold. to found a company after university have therefore read up on business-related skills and taken a few courses on entrepreneurship.
My understanding of finance and business practices is limited to what I read on books and the internet, but it is still much better than that of the average computer scientist or data scientist.
Providing proof of soft-skills is really hard because it is so difficult to find something tangible and verifiable. How about you just ask me some questions in an interview?
Autodidactic learning about philosophy and ethics since age 13, writing of philosophy texts since age 14Most of what I wrote back then is of course outdated by now and I no longer believe in it, but I wouldn't have gotten to where I am now without these intermediary steps..
This was in parallel with my developing interest in the Cognitive Sciences
One particular result of my deliberations is that I evaluated what possible careers would allow me to have the greatest positive impact. The answer to this was to work in AI, as an advanced AI has the potential to solve a lot of problems, moreso even than curing cancer or eliminating world hunger.
The only reason I am listing this as a hobby instead of a skill is that it doesn't earn me any money. It's just useful for me personally.
I invented a fictional setting wherein the characters are aware of their own nature as fictional characters (metafiction).
I created several stories in this setting.
Features:
Unfortunately my ability to write is not on par with my ability to come up with ideas, so I haven't published anything, yet. It is an ongoing effort for me to write a story that is good enough to be worth publishing, while my collection of ideas keeps growing.
I read a lot. Academic or leisure, real or fictional, in a book or on the internet.
I also like to daydream a lot about the stories I read. I imagine what I would do if I was a character in the story.
By that I don't mean "if I was the hero I would shoot the bad guy so hard he would die to death!".
Instead, I mean "if I was that senator in the background (or some other authority figure with influence), I would try to tackle the problem of the story by introducing a law like X. Of course, that might anger interest group Y, so I would have to think of a way to deal with them. Maybe I could hire the hero to deal with them? He already has a grudge against them anyway. Also, funding research into Z could be useful in the long term.".
I do pushups as a strength exercise and go running for endurance.
I consider sports an important investment in the future, as it makes me less likely to randomly drop dead from a heart attack.
The increased charisma from looking fit also never hurts.
I have a tendency to think differently from other people. I am unsure whether my philosophizing in my youth was the cause of this, or merely a symptom, but it is definitely helped by my tendency to use game theory and psychology in everyday life.
One common way in which this shows is that I intuitively notice cognitive biases and statistical weirdnesses. This is especially useful at my job as a Data Scientist, where I am known for being able to quickly find the explanation for strangeness in data.
Another frequently useful aspect of this is that my thinking is always drawn towards asking questions like
"What is our actual goal here, exactly?",
"Is this really helpful, or are we just doing this because everyone else is doing it?".
I like to optimize and question the status quo, which often leads to useful improvements.
This is not to say that I somehow just think in a superior way. My way of thinking also has its drawbacks. However, if everyone thinks in the same way then the group is less effective than if everyone explores different options, so in the end it is mostly an advantage.
I have developed and mastered a technique for controlling my emotions.
Many people try to suppress their emotions to be more rational, but this often does more harm than good. Instead, I have learned how to control my emotional state by transforming a harmful emotion into another, more practical one.
As a result of this, I am very emotionally balanced and do not get distracted by stress, anger, or even personal tragedy.
In effect, I have a very stoic personality.
how does this work?The technique works by reflecting about the evolutionary purpose of the emotion, and putting it into a modern context.
I start by acknowledging that the emotion is meant to help me. I then talk to myself to explain to my subconscious why the emotion is actually unconstructive in our modern society. Finally, I propose an alternative emotion that deals with the situation more effectively, and will my mind to embrace this alternative.
With enough practice, the technique becomes automatic and practically instantaneous. It is now no longer necessary for me to meditate and talk to myself for any but the most unusual scenarios.
I am very creative and good at thinking outside the box. Of course, everyone claims this about themselves, but I can prove it:
I read about "The Hardest Logic Puzzle Ever" and found a way to solve the puzzle more effectively than anyone before me, including the people who had written actual academic papers about it. I wrote the section on beating the puzzle in a single question instead of the three questions required by the puzzle. (Check the article's edit history to see my name listed as the author.)
Having a different way of thinking from other people is a double-edged sword.
I frequently find myself dismissing topics on the news as unimportant and not worth the time, only to find out later that many others do not share that view, so that I am the only one present who doesn't know what everyone else is talking about. Meanwhile, topics that seem important to me are unpopular with others and do not allow me to strike up a conversation, or worse, alienate others.
I also sometimes react in very logical and unemotional ways when everyone else is reacting emotionally. While acting logically is itself good and useful, it can also occasionally be harmful for interpersonal communication.
exampleMy reaction to a recent terrorist attack was to make fun of the terrorists for killing so few people despite having had so much time and resources at their disposal.
The reason: the entire purpose of terrorism is to spread terror, so by making fun of terrorists I am denying them their victory. If everyone had this reaction, terrorism would no longer work and it would eventually just cease.
This logic seems straightforward to me, but lead to awkward glances from people who don't subscribe to this reasoning and just think I am being insensitive.
Most of my work so far focused on relatively small projects, with around half a dozen members and only a few months of duration.
While I have read about the theoretical differences between large and small projects (there are some interesting books about this in psychology and economics), I have not worked on a very large project before in practice.
Notably, my own AI project is rather large by now, but since I work on it alone I have yet to experience the difficulties in communication and organisation that arise when working in teams of several dozen people or more.
I'm really arrogant.
Fortunately, I am awesome at practically everything else, so this is not a big deal :-)
(note: this is a joke. Don't take it seriously.)
I love research, but I hate writing papers and all the bureaucratic and unnecessary formal requirements that weigh down the publishing process. Moreover, it feels like academia nowadaysActually, it's probably always been like that, but I wasn't around a hundred years ago, so who am I to judge? is not about the pursuit of knowledge but the accumulation of prestige.
For these reasons, I decided to go into industry directly instead of pursuing a PhD. While industry is of course even less about the pursuit of knowledge, it is at least honest about its goals, which is to earn money, and I can work with that.
I am still conducting research on my own time, I just don't like having to spend countless hours on overhead work just so that I earn the right to put three letters next to my name.
I was originally planning to do a PhD.
However, I was contacted by someone from the Datalab who had noticed me on LinkedIn.
The position sounded a lot more attractive than writing papers:
Customers often inhabit a very different world from programmers.
Learning how to most effectively translate difficult concepts into a language understandable by someone with no background in the field is practically a science of it own.
Conversely, it is often difficult, but always extremely important, to understand the customer's demands correctly and read between the lines where buzzwords blur the actual task.
While working on my own AI project in lisp, I stupidly decided to implement a multithreading task in a very non-standard way because I deluded myself into thinking that what I was working on was an exception to the rules.
I tried to implement a non-cooperative form of multithreading, because the code I wanted to parallelize contained automatically generated turing-complete code, and it was entirely possible, and acceptable, that this code would occasionally go into an infinite loop, so that I would have to break it up from the outside.
When I started getting random crashes, I spent two months on and off taking apart all of my code to see where the crashes were coming from. In the end, I finally found out that the problem was caused by a component so small and primitive that I never bothered to test it thoroughly. It always worked fine in other contexts, but the multithreading sometimes gave it unusual input that caused it to crash.
Ever since then, I care a lot more about exhausting every possibility when setting up component tests.
I have fortunately never had any major conflicts with teammates.
The closest I have come to that was a deep philosophical difference with a coworker about how a problem should be tackled most effectively. Realizing that none of us could convince the other of his way of thinking, we simply split the task in such a way that there would be no friction between the two approaches and then each of us worked on his half of the problem in his own way.
Besides such professional disagreements, I have never encountered a coworker I could not get a long with.
I am fine with relocating. I would prefer to live in Silicon Valley, or failing that, another place with English as the dominant language and a similar culture.