The Panini Formulas

The Panini sticker book for the upcoming UEFA Euro 2016 football tournament features 680 stickers this time. We can calculate how many sticker we need to buy to fill the album by using the amazing (due to its relationship to music and architecture) harmonic series formula:

Latex formula

Buying 4829 stickers, it will give me 4149 stickers to swap, assuming that stickers are evenly distributed. If we take into account that stickers are not sold separately but in packs of 5 stickers, the result is slightly different.

Having bought my first 160 stickers today, I have 3.3% of these already, caught in the coupon collector’s problem. We can calculate the probable duplicates I will have from these 160 stickers by:

Latex formula

and I ended up with 17 duplicates, which is nearby.

Today’s Neural Networks Are The Machine Code

In a great new piece in Wired, Jason Tanz announces the end of coding. That might be a bit tendentious but it shows that machine learning that comes with neural networks has the effect of computers programming themselves, with us humans reduced to providing some training data sets:

If in the old view programmers were like gods, authoring the laws that govern computer systems, now they’re like parents or dog trainers.

A neural network takes lots of input data and create its own internal wirings, matching input data to the desired output. It is self-learning, so to say. The neurons of a neural network effectively learn by adjusting their own relative weights (expressed by a number value) as well as by adjusting the weight of the relationships between different neurons (also expressed by number values). As all of the generated stuff is just lots and lots of numbers, we might not be able to say for sure how neural networks work and how they will react on specific input data:

The code that runs the universe may defy human analysis. Right now Google, for example, is facing an antitrust investigation in Europe that accuses the company of exerting undue influence over its search results. Such a charge will be difficult to prove when even the company’s own engineers can’t say exactly how its search algorithms work in the first place.

While this is surely true, it is perhaps too early to be pessimistic.

Continue reading “Today’s Neural Networks Are The Machine Code”

Automation and Artificial Intelligence

Steven Cherry has written an essay about automation and artificial intelligence for the New Scientist. I took two insights from it:

Malone’s third law: “Every technology breakthrough takes twice as long as we expected and half as long as we are prepared for.

So, let’s give real automation with artificial intelligence a few decades, from today.

In fact, all jobs are at risk, even the software programmer’s.

I agree that automation will target all employment sectors, but I would argue that the degree of risk will be divided very unevenly between different professions.

WhatsApp Must Speak German Because

Legal journalism is no easy thing, especially if it is about foreign law or court judgments. All too easy, all those details making the difference, or just the legal context, don’t make it to a published article.

David Meyer’s coverage for Fortune of a recent German regional court of appeal decision in a legal case that WhatsApp is fighting against a local German consumer protection association:

WhatsApp Must Speak German to Germans

Yes, but. This is by no way a suprising decision. It is long established in German legal practice that foreign companies that actively target German consumers are subject to German law, including legal provisions regulating Internet commerce. It is easy to actively target German consumers, with e.g. a German language website (like being sufficient. If you want to avoid German regulation, don’t target German customers specifically.

The appellate court said that while many Germans can understand everyday English, few speak the English legalese in which terms of service are written. Therefore, it said, it was unreasonable to expect them to do so, and WhatsApp’s terms of service and privacy policy are essentially ineffective.

Yes, but. In its decision, the court stressed that it probably wouldn’t agree to any terms of service not being in German language: “Alltagsenglisch mag verbreitet sein, für [..] überhaupt kommerzielles Englisch [..] gilt das aber nicht.”, meaning that while everyday English might be found commonly in Germany, this isn’t true for any commercially relevant situation. This is also a common line of German courts. So, if you want to do business in Germany, you need to follow German law.

If the ruling is finalized

WhatsApp has a chance to lodge a kind of final appeal to the German Federal Court of Justice (Bundesgerichtshof). However, as the recent appeal decision in this case is in line with the common sense of courts in Germany on the applicability of German law, I doubt very much that such a final appeal would have any chance to go through.