Nice to Meet You
How are you today? Because you’ve come to my site, I’m guessing that you want to know something about me. Let’s start with a cliche professional photo of me giving a talk in some event…
And here is a more recent, and a much more fun picture of me with my niece…
But I guess that you’ve come here to learn about my professional interests, like research stuffs, and not so much my niece, so ok.
After having built one of the first major eCommerce solutions in Europe in the mid-nineties, I spent over 20 years with internet technology startups as a founder, advisor and board member. In 2014 I co-founded and then chaired I-COM’s Data Science Board, and Botlab, the only non-profit foundation focused on advertising fraud research. I’ve probably wrote more articles, gave more talks, and spent more years than anyone else on uncovering the corruption in the online media supply-chain.
I have many global firsts in digital media, and my work has been featured several times on the cover of Financial Times. I’ve worked with artificial intelligence, mainly focused on Natural Language Processing, since 2004. I’m an advisor to Unicef and World Federation of Advertisers.
I loved numbers since my earliest memories, and started to dabble with computers when I was 6 years old. Around 1982. Thanks to my Mom, that’s when I got my hands on one of these guys:
Vic-20 from the legendary Commodore company. Soon the computer gave a plausible reason to start skipping school, and before time, I had dropped out entirely. I never made it to high-school, and when the time came to look for a job, I just kept going with what I had started as a kid. The first company I co-founded built the first major e-commerce solution in Finland around the mid-nineties. We had no idea what we were doing, but neither did any of our competitors. From there I sort of drifted along to data analytics and artificial intelligence.
I’m interested in how technology can help reduce suffering. At the same time, I’m painfully aware of how it’s causing it. Sometimes I think it would be better to sweep floors than write code, analyze data, or write articles.
My focus for the past few years have been on making available information and tools that help others do what I had myself learn to do after close to four decades of working with computers. A good example of this is a series of notebooks I’ve developed for the purpose of making it incredibly fast (and fun) to learn numerical computing (for all ages).
An inspiration for this project is my friend’s Mom, and the amazing math books by Yakov Perelman.
Another project, also related to children, has to do with my experience in being one of the most outspoken commentators regarding ad blocking and the threats related with online advertising. While helping Unicef to understand the role of online advertising in protecting children, I realized there was a major void in terms of understanding of the topic.
The upcoming guide is focused on helping children, parents, educators, and policy-makers to understand the threat that online advertising poses on our children. If you’re interested in contributing to it, don’t hesitate to get in touch.
My writings on ad blocking can be found in Thank You For Ad Blocking [website].
Because of the vast problems I came to witness as part of my research associated with the internet, I spent much time looking for technologies and ideas that provide reasonable alternatives to whatever it is that we have now. Perhaps the most interesting is related with mesh-networking, and the way it may facilitate for information equality and decentralization.
What we have in the above photo is a an actual example of the gear we used in one field deployment for a solar powered mesh-network high up in the Himalayas.
The field report Things I Learned Building the Skynet [article] and How to Eat Dust [article] explain the approach in great detail. Related, I’m a co-founder of Eka Foundation a think-thank focused on the human future of the internet.
But decentralization is about much more than network configurations. And no, Bitcoin has nothing to do with decentralization, actually in some ways it could be argued that it’s more centralized than the global banking system. True decentralization can act as a driver to more equality and transparency in the society, and it’s really important we get it right. Or at least less wrong than we’re getting it so far.
My manifesto for decentralization can be found in dcentralize.net [website] and in the article Thoughts on Pseudo-Decentralization [article].
As you probably already gathered, I love to tell stories. Being somewhat of a data nerd, I’ve focused extensively on the aspect of data visualization as a means to support storytelling. Because many of the greatest sufferings I’ve experienced with computers have to do with visualization libraries, I’ve created my own.
The goal I have with Astetik [code] is to make creating beautiful graphs to support storytelling without having to know how to code. The above example speaks for the simplicity, especially when you consider that this is an actual plot from a single command. Astetik is a representation of my view and experience, accumulated over more than 20 years as a researcher and a data-driven storyteller.
Some of the the ideas leading to Astetik can be found in Five Essential Points on Data Visualization [article].
The reason why I became originally interested in starting an internet company, had to do with intelligent agents. Having worked since the turn of the millennia with application of linguistic computation, and other artificial intelligence methods, in 2012 I started The 10x Computer Club.
The revised 10xcc manifesto Brave New Computing [article] and the 10xcc’s blog [website] with many writings on singularity, intelligence, and computation efficiency.
Three of the members of the 10xcc ended up starting Botlab, a research foundation that focused on developing of high-performance, open-source, machine intelligence solutions.
Botlab’s main contributions include Nameles [website] signals intelligence fraud detection and Autonomio [website] deep learning workbench.
You can learn more about my thoughts on Hyperparameter optimization in this article.
A somewhat comprehensive overview of my organizational background can be found on LinkedIn [profile].
Words of Advice
Be nice to people, and tell them that they’re doing a great job. Keep causality in mind; the kind of results that will arise depend on the causes that are put forth. Stop doing things that you love to do, and do things that bring great results. Great results are those that end up benefiting others. Why change the world, if you can save it. Never take anything about yourself or what you do too seriously, unless its about genuinely helping others.Be grateful for all kinds of things, regardless if you consider it good or bad. Try to meditate regularly, at least a few minutes each day.