Jan Böhmermann in ZDF Magazin Royale: Artificial intelligence kills

Jan Böhmermann no longer cooks and shows in ZDF magazine Royale what he can do best. He doesn’t like artificial intelligence. The TV review.

Jan Böhmermann is back. Well, already for a week and only from an excruciating summer break in two senses, but at least. First things first: he does what he can, should and must do again. In other words: he doesn’t cook. In the end we, blood-sucking critics, had to make do with extremely bland food.

Because “Böhmi sizzles” was, let’s be honest, scary. In order to tackle the current episode of his ZDF Magazin Royale from behind, I would like to quote a (not the!) Wendler who “says” what was missing: depth of content. Further: “It was uninspired, poorly researched and not at all funny. I am honestly disappointed. ”A small, subtle self-irony on a meta level, which could lead to one or the other critic feeling a bit caught at this point.

Jan Böhmermann serves a three-course menu in ZDF Magazin Royale

So it might be. If one thing has not changed despite the summer break, it is that a ZDF Magazin Royale is usually served as a three-course menu. As an appetizer, there are jokes from the retort. Fast cuts, Harald Schmidt memories. Late night. Must be. Likewise, Annalena Baerbock, Armin Laschet and Olaf Scholz should be mentioned briefly. After all, it belongs in every good online article. Keywords. Good for the algorithm. That an AI reads. But more on that later.

The second course is home-style cooking, everyone knows it. May also be out of the microphone. Like the difficult to digest (and please soon removed from the menu) format “Thintank21”. Who comes up with that? Are these the same people who do two shows for RTL like “the 10 best sauce binders”? With a funny voice from the off and so? And did this computer-scripted slapstick “KI Magazin Royale” have to be real? For free, it goes by, it’s more teeth cleaning than root canal treatment. Picks, annoys, but is quickly forgotten.

Jan Böhmermann in ZDF Magazin Royale after the summer break

In the third part, Jan Böhmermann doesn’t spoil it, it’s mostly delicious. Quasi the (also temporally) extended dessert that everyone shovels full of because it didn’t taste that good before. And it will be inspired, well researched and even funny at one point or another. Today’s topic: Artificial Intelligence. The ones from the computer. Even if it is ultimately not entirely artificial, as “Böhmi” opens almost exclusively to an astonished world public.

Because this artificial intelligence is actually fed by humans. So of real intelligence. Or somehow not, because (unfortunately) there is no intelligence check before someone is allowed to mess in their mustard (at this point we should remind you of the first episode of ZDF Magazin Royale after the summer break, the one with Wikipedia). And because this intelligence is not that intelligent as long as it is mostly fed with white bread. From white bread bakers who convince the machine brain that it would be wise to only listen to the whites.

Jan Böhmermann shows the limits of artificial intelligence in ZDF Magazin Royale

Jan Böhmermann quotes several media outlets and MIT computer scientist Joy Buolamwini, which leads to the fact that the facial features of black people are simply recognized rather suboptimally. Buolamwini reports that her facial features were only properly read out when she literally put on a white mask. It can be found grotesque. Or scary. Or just wrong. Just as wrong as the fact that this leads to facial recognition software incorrectly assigning black people and thus initiating a process that can result in innocent people coming into conflict with the law.

On the other hand, that’s also consistent somewhere. It’s not the case that the racial profiling that real white police officers use with their very conventional intelligence doesn’t often produce a similar result. If you fancy a fact check, we recommend that you take a seat at a central station armed with a pad and pen and make a tally sheet while friends and helpers do their job.

Jan Böhmermann then continues in his ZDF Magazin Royale on his little round trip through the abstruse effects of a world in which we click on all traffic lights and zebra crossings. To be able to look at pages. And, as Böhmermann explains, Google’s self-driving cars will soon help to recognize things as well as we do.

Jan Böhmermann fillets the EU’s AI rules in ZDF Magazin Royale

Incidentally, a dystopian idea for anyone who has overlooked a hydrant on the fourth try. Because as funny as it may be sometimes: AI is not harmless. Böhmermann shows here, suitable for the masses, that it is not funny when Israel develops kamikaze drones that detect self-controlled targets which they can then switch off automatically. Foreign Minister Heiko Maas can look at the camera so faithfully that it runs counter to good, German “ethical standards” to let an AI shoot people down on its own – it’s not that reassuring. If only because “Böhmi” unpacks a few documents.

This begins with the “Future Combat Air System”, which, according to Spiegel, will replace the Eurofighter and the French Rafale from 2040. Also on board: “unmanned missiles”, which combine the “capabilities of guided missiles and drones” and “have at least some autonomous functions”. At least the EU wants to ban biometric mass surveillance. This is what it says in the “Ordinance on Artificial Intelligence”, which Der Spiegel dug out in April 2021. The EU wants that too. With a few tiny exceptions, of course.

Jan Böhmermann reads from the European Commission’s announcement 2021/0106 / COD that the police in the EU may use AI for automatic surveillance as long as there are very, very, very rare occasions. That would be: search for potential victims, danger to life or physical integrity, danger of a terrorist attack, identification or prosecution of perpetrators or suspects. And of course, as long as it is not a question of remote identification in real time, when an authority other than the police carries out the surveillance, when a private company is hired by public authorities, or, as Jan Böhmermann adds, when someone is “following a husky sledge Drives home pulled by six robotic vacuum cleaners ”.

It’s good that he doesn’t cook: Jan Böhmermann is in top form in ZDF Magazin Royale

Uff. Böhmermann thinks that this anti-AI surveillance law has “more loopholes than Didi Hallervorden’s underpants”. You can see it that way. Sure, you could also find that these are all highly justified exceptions. After all, nobody likes to play with terrorists and all that. They’re the bad guys. Or? Just like all those who sit in Turkish prisons as “terrorists”. Because they are Kurds or critical journalists. So terrorists. Highly official.

The fact that the Chinese company Huawei trains its facial recognition software to recognize Uyghurs and, according to the Washington Post, reports them to the government immediately and automatically, is also mentioned by Jan Böhmermann. Not even the quote from the Guardian, which refers to a report by Amnesty International, when he speaks of these Uyghurs being interned “in a dystopian hell landscape”. And on which the Chinese police then have AI emotion recognition software tested, according to the BBC. The customary treatment of a Muslim minority in China in times of self-flying, self-aiming and self-killing kamikaze, so to speak.

So it is once again one of those episodes of ZDF Magazin Royale that are no laughing matter. So one of the best. Because Jan Böhmermann is in top form when he is not playing the thigh thumper or small talker. When he cultivates his greatest strength. When he takes off his apron and mixes snappy satire with relevant topics and strong research. Before he sits down on a husky sled and lets six robotic vacuum cleaners drive him out of the studio. (Mirko Schmid)