While running i listend to a On The Media interview with the autor of the Wired piece on content moderation that was making the rounds this week. After a while the interview addresses the relatively obvious question of why content moderation is still a task that can only be carried out by humans:

Brooke Gladstone: I think that one of the things that struck me is that this work demands human beings clued in to American mores and laws. This has to be done by brute force of eyes and clicking fingers. Is there no alternative to human moderation?

Adam Chen: Well everyone i talked to said that there was no way a robot could do all of this. They can come up with programmes and algorithms that will make it more effective and more streamlined but there is always going to be somebody who has to looks at it. And also the kinds of moderation that si going on is becoming more nuanced and complicated. And so i think you always gonna need people and probably more and more people as time goes on.

What struck me when listening to this is exchange that this is just one instance of a much broader problem, namely the current inability to encode moral judgement in algorithms. Once we have ‘solved’ this issue those poor schmucks who have do do content moderation for the rest of us will be out of job (which sounds as a good thing) but that will also be the moment where have to start dealing with killer drones/robots that do not require human interventions before firing their weapons.

A while ago i used this space to express my skepticism with regards to delivery drones becoming a major thing in the developed world anytime soon (and also hedged that by pointing to the fact that they may be much more useful and economically viable in developing countries). A couple of days ago i came across an excellent essay (Build cargo drones, get rich) by J.M. Ledgard in which he makes the most convincing case for cargo drones i have come across yet.

While his scenario is entirely focussed on the use of drones (which he calles donkeys) in Africa it is interesting to note that the first step in the scenario that het is working on is exactly the same as an experiment just announced by German logistics company DHL. Here are the opening words from Legards essay:

My goal is to help set up the world’s first commercial cargo drone route in Africa by 2016. It will be about 80 kilometres long and will connect several towns and villages. The first cargo drones will carry small payloads of blood to keep alive children who would otherwise perish.

and here is the relevant passage from DHL’s press release announcing their experiment:

For the first time worldwide, medications and other urgently needed goods will be delivered to the island at certain times of the day by DHL parcelcopter. This research project represents the first and only time in Europe that a flight by an unmanned aircraft will be operated outside of the pilot’s field of vision in a real-life mission.

Aside from the fact that it seems that Ledgard has lost its bid for running the world’s first commercial cargo drone route it is interesting to note that the business case is the same here: Emergency deliveries of life-saving, small things by drone.

This is indeed where commercial cargo drones seem make some sense in their current state. What is happening here should be seen in the light of another recent innovation that started as an expensive solution for a first world problem but dramatically changed the lives of millions of Africans once it became technological mature enough to be commoditized: The mobile telephone. So while he may have lost his bet to be the first, DHL is most likely doing a bit of extra R&D work for Ledgard’s scheme to get rich with cargo cones.

p.s. also notice how both implementation are targeted at rural, non-urban environments. the point i made in my earlier post clearly still holds.

Friday morning when we left the cab that had brought us and our newborn from the hospital the mail carriers on their bicycles where swarming out around us (we live 3 houses down from a postnl distribution location where the mail for the neighborhood is transferred from delivery vans to bikes). Having spend some of the waiting time in hospital reading Alex Madrigal’s Atlantic feature on Google’s delivery drone programme this made me briefly contemplate if our daughter would grow up to see a world filled with mail carriers or one filled with flying delivery drones.

In spite of the hype caused by said article which seems to have made otherwise reasonable people abandon their analytic rigor, i am pretty certain that delivery drones will not play a significant role in my or my daughters life anytime soon¹.

Primarily because i can’t really see the economics making sense (certainly not for the silly examples of flying spare batteries or electric drills around) but more importantly because these things have clearly sprung from a suburban mindset that assumes that people have exclusive control over the vertical space belonging to a swath of land. For the majority of people in cities this is simply not the case and as a result they don’t have a place where the delivery drones could land (or lower their eggs).

The answer for the desire to have relatively quick access to things is living in a city. Cities provide quick access to shops and via mail carriers on bikes. If any part of the delivery process will be replaces by autonomous vehicles i would bet on the trucks serving the trunk routes and maybe the those serving neighborhood distribution centers. Autonomous flying delivery drones, may sound like the future, but they almost certainly aren’t, unless we envisage the future to be a version of suburbia that has been perfected so that there is no need to venture out into the world and interact with anyone anymore.

I would certainly hope that the future will look a whole lot more urban.

The Los Angeles times has a terrifyingfascinating reconstruction of a US airstrike that killed 23 (or 16 if you believe the US military) afghan civilians on 21 february 2010. The reconstruction focusses on the role played by the pilot and camera operator of a predator drone that observed three ‘suspicious’ vehicles whose passengers had the misfortune crossing paths with a team of US special operators early on February the 21st.

The article documents how the drone operators (in Nevada), the the commander of the Special operations team (on the ground in Afghanistan) and a team of video screeners (in Florida) interacted in falsely determining that the Afghanis were armed insurgents (in reality they were unarmed villagers). One of the most intresting aspects of the article is the insight it provides into the communication patterns between the different actors involved on the US side:

The Predator’s two-man team — a pilot and a camera operator — was one of the Air Force’s most-experienced. […] Also stationed at Creech were the Predator’s mission intelligence coordinator and a safety observer. In addition, a team of “screeners” — enlisted personnel trained in video analysis — was on duty at Air Force special operations headquarters in Okaloosa, Fla. They sat in a large room with high-definition televisions showing live feeds from drones flying over Afghanistan. The screeners were sending instant messages to the drone crew, observations that were then relayed by radio to the A-Team. On the ground, the A-Team was led by an Army captain, a veteran of multiple tours in Afghanistan. Under U.S. military rules, the captain, as the ground force commander, was responsible for deciding whether to order an airstrike.

So you have a ground commander who cant see the the target who is in radio contact with the drone operator who gets most of the image analysis provided by screeners in another state who provide their assessments via direct messages and emails containing image stills) but without having access to the radio traffic with the ground commander.

As it turns out this is about as much a recipe for disaster as it sounds… [read the full article here]