The Ontological Status of the Driver – or why we need to change the law

[et_pb_section admin_label=”section”][et_pb_row admin_label=”row”][et_pb_column type=”4_4″][et_pb_text admin_label=”Text” background_layout=”light” text_orientation=”left” use_border_color=”off” border_color=”#ffffff” border_style=”solid”]

In the debate on autonomous cars, one document keeps popping up: The Vienna Convention for Road Traffic from 1968, which is ratified across Europe. Most debators point to article 8 in zeroing in on the legislative hinderance to deploying autonomous systems to our roads.

I am actually a big fan of article 8. In my opinion, the real problem lies in article 1, and here is why.

Article 8 basically states two things:

  1. The driver should have knowledge of the vehicle in question

  2. The driver should have control of the vehicle in question.

Both of these stipulations make a lot of sense. But when we take a closer look things get a little blurry.

What kind of knowledge do we actually have about our vehicles? We normally presuppose that taking a driver’s license involves some kind of theoretical knowledge about what an engine is, how how brakes and instruments work, etc. But in reality, our knowledge of the vehicle is often very limited. I know that if red lights appear on my dashboard I have to pull over and call a mechanic. The mechanic will run diagnostic program after which the computer tells him what the car perceives as a problem, and he can then relay the message to me. So in reality, knowledge of the vehicle is highly distributed, and thoroughly dependent on computer systems.

What kind of control do we have, then? In a technical sense, control is understood as mastery of energy. When running this means controlling energy from the engine to the wheels, and when stopping this means controlling energy from the wheels to the brakes. But in reality, again, those controls are highly distributed and computerized. In both going and stopping the flow of energy of is controlled by computers way beyond my detailed understanding and control.

Knowledge and control are shared by humans and computers. If we can accept this fact of reality, the “driver” is simply understood as a combination of human and systemic controls, article 8 stays intact, and we can all move on. But this is where article 1 enters.

Article 1 clearly states that a driver is a person and this is the real problem.

It means that a person – and ONLY a person – can perform in the role of a driver. This stipulation obviously has deep legal consequences, e.g., concerning liability. But from a practical point of view this position is untenable.

The reality is that even today the notion of a driver (of knowledge and in control) is a distributed entity where a person alone cannot be said to meet the requirements. Looking ahead, this also implies that even a system with far more knowledge than the average license seeker and with superior driving skills will not be able to obtain a license for the road. Why? Because only persons are eligible for such a license. To me that does not make sense.

Will someone PLEASE rewrite article 1 to reflect what we actually want on our roads?

We owe it to ourselves and to our children to build better and safer transportation systems, and right now that means: discontinue the obsolete legacy law in favor of real world solutions that ensure safety inside and outside of our cars.



Tesla charger robot

[et_pb_section][et_pb_row][et_pb_column type=”4_4″][et_pb_text admin_label=”Text” background_layout=”light” text_orientation=”left” use_border_color=”off” border_color=”#ffffff” border_style=”solid”]

Finally! A solution to this conundrum: We have advanced infrastructure, we have advanced cars, yet we still have to manually refuel our cars :/ One of the silly leftovers from an age where people were supposed to serve machines.

Tesla have revealed a prototype of a robotic charger. It may not be the stable or final solution we need, but its definitely a step in the right direction.


[/et_pb_text][et_pb_video admin_label=”Video” src=”” image_src=”//”] [/et_pb_video][/et_pb_column][/et_pb_row][/et_pb_section]

Google photo mistakes people for gorillas

[et_pb_section][et_pb_row][et_pb_column type=”4_4″][et_pb_text admin_label=”Text” background_layout=”light” text_orientation=”left” use_border_color=”off” border_color=”#ffffff” border_style=”solid”]

@jackyalcine caused quite a stir on the web – and rightly so – when Google Photos tagged pictures of her friends as gorillas.

Thinking about this as an A.I. problem, I do understand how something like this could happen. It is actually not very surprising for computer vision algorithms to make this mistake. This is because such algorithms looks for patterns of pixels WITHOUT ANY UNDERSTANDING OF WHAT THEY REPRESENT.

For a human being, however, the mistake is completely unacceptable. And this is because humans ALWAYS think in terms of what is represented. This is why it is impossible for a human being to make this mistake.

[/et_pb_text][et_pb_image admin_label=”Image” src=”” show_in_lightbox=”off” url_new_window=”off” animation=”bottom” sticky=”off” align=”left” force_fullwidth=”off” always_center_on_mobile=”on” use_border_color=”off” border_color=”#ffffff” border_style=”solid” /][/et_pb_column][/et_pb_row][/et_pb_section]

Mcity test facility for driverless cars

[et_pb_section][et_pb_row][et_pb_column type=”4_4″][et_pb_text admin_label=”Text” background_layout=”light” text_orientation=”left” use_border_color=”off” border_color=”#ffffff” border_style=”solid”]

U of Michigan has inaugurated a test facility for unmanned vehicles, comprised of a variety  of road types and “real world” obstacles that autonomous cars will encounter.

Good idea.

But I would like to point out that such projects MUST be accompanied by real world tests. There are a number of things that you will NEVER learn from simulation, be it soft or hard.

PS.: The idea of populating a test space with robots  that you can run over seems less than appealing to me. But perhaps I’m biased…

[/et_pb_text][et_pb_video admin_label=”Video” src=”″ image_src=”//” /][/et_pb_column][/et_pb_row][/et_pb_section]

Unmanned Factory in Dougguan City

According to People’s Daily Online, China is accelerating the use of robots in the manufactory industri. Fast.

Prognoses are, that 90% of the work force will be replaced by robots in the foreseeable future.


This will be an interesting project to follow, and a good indicator that we should be really careful in predicting the needs for work forces.