Tag Archives: Experiment

Reverse Robotics: A Brief Thought Experiment

Imagine a world in which manually controlled technologies simply do not exist. The very thought of manual technologies is, in actual fact, hardly conceivable let alone comprehensible. Instead, this seemingly alien world is seamlessly powered by intelligent and autonomous robotics systems. Lets call this world Planet AI.

PlanetAI

Planet AI’s version of airplanes, cars, trains and ships are completely unmanned. That is, they are fully autonomous—a silent symphony of large and small robots waltzing around with no conductor in sight. On one fateful night, a young PhD student awakens in a sweat unable to breathe, momentarily. The nightmare: all the swirling robots of Planet AI were no longer autonomous. Each of them had to be told exactly what to do by the Planet’s inhabitants. Madness.

She couldn’t go back to sleep. The thought of having to tell her robotics transport unit (RTU) in the morning how to get from her studio to the university gave her a panic attack. She would inevitably get lost or worse yet crash, maybe even hurt someone. She’d need weeks of practice to manually control her RTU. And even if she could somehow master manual steering, she wouldn’t be able to steer and work on her dissertation at the same time during the 36-minute drive. What’s more, that drive would easily become a 100-minute drive since there’s no way she would manually steer the RTU at 100 kilometers an hour—the standard autonomous speed of RTUs; more like 30km/h.

And what about the other eight billion inhabits of Planet AI? The thought of having billions of manually controlled RTUs flying, driving & swimming through the massive metropolis of New AI was surely the ultimate horror story. Indeed, civilization would inevitably come to an end. Millions would die in horrific RTU collisions. Transportation would slow to a crawl before collapsing. And the many billions of hours spent working, resting or playing in automated RTU’s every day would quickly evaporate into billions of hours of total stress and anxiety. The Planet’s Global GDP would free fall. RTU’s carrying essential cargo automatically from one side of the planet to the other would need to be steered manually. Where would those millions of jobs require such extensive manual labor come from? Who in their right mind would even want to take such a dangerous and dull assignment? Who would provide the training and certification? And who in the world would be able to pay for all the salaries anyway?

At this point, the PhD student was on her feet. “Call RTU,” she instructed her personal AI assistant. An RTU swung by while she as putting on her shoes on. Good, so far so good, she told herself. She got in slowly and carefully, studying the RTU’s behavior suspiciously. No, she thought to herself, nothing out of the ordinary here either. It was just a bad dream. The RTU’s soft purring power source put her at ease, she had always enjoyed the RTU’s calming sound. For the first time since she awoke from her horrible nightmare, she started to breathe more easily. She took an extra deep and long breath.

starfleet

The RTU was already waltzing with ease at 100km per hour through the metropolis, the speed barely noticeable from inside the cocoon. Forty-six, forty-seven and forty-eight; she was counting the number of other RTU’s that were speeding right alongside her’s, below and above as well. She arrived on campus in 35 minutes and 48 seconds—exactly the time it had taken the RTU during her 372 earlier rides. She breathed a deep sigh of relief and said “Home Please.” It was just past 3am and she definitely needed more sleep.

She thought of her fiancée on the way home. What would she think about her crazy nightmare given her work in the humanitarian space? Oh no. Her heart began to race again. Just imagine the impact that manually steered RTUs would have on humanitarian efforts. Talk about a total horror story. Life-saving aid, essential medicines, food, water, shelter; each of these would have to be trans-ported manually to disaster-affected communities. The logistics would be near impossible to manage manually. Everything would grind and collapse to a halt. Damage assessments would have to be carried manually as well, by somehow steering hundreds of robotics data units (RDU’s) to collect data on affected areas. Goodness, it would take days if not weeks to assess disaster damage. Those in need would be left stranded. “Call Fiancée,” she instructed, shivering at the thought of her fiancée having to carry out her important life-saving relief work entirely manually.


The point of this story and thought experiment? While some on Planet Earth may find the notion of autonomous robotics system insane and worry about accidents, it is worth noting that a future world like Planet AI would feel exactly the same way with respect to our manually controlled technologies. Over 80% of airplane accidents are due to human pilot error and 90% of car accidents are the result of human driver error. Our PhD student on Planet AI would describe our use of manually controlled technologies a suicidal, not to mention a massive waste of precious human time.

Screen Shot 2016-07-07 at 11.11.08 AM

An average person in the US spends 101 minutes per day driving (which totals to more than 4 years in their life time). There are 214 million licensed car drivers in the US. This means that over 360 million hours of human time in the US alone is spent manually steering a car from point A to point B every day. This results in more than 30,000 people killed per year. And again, that’s just for the US. There are over 1 billion manually controlled motor vehicles on Earth. Imagine what we could achieve with an additional billion hours every day if we had Planet AI’s autonomous systems to free up this massive cognitive surplus. And lets not forget the devastating environmental impact of individually-owned, manually controlled vehicles.

If you had the choice, would you prefer to live on Earth or on Planet AI if everything else were held equal?

How Can Digital Humanitarians Best Organize for Disaster Response?

I published a blog post with the same question in 2012. The question stemmed from earlier conversations I had at 10 Downing Street with colleague Duncan Watts from Microsoft Research. We subsequently embarked on a collaboration with the Standby Task Force (SBTF), a group I co-founded back in 2010. The SBTF was one of the early pioneers of digital humanitarian action. The purpose of this collaboration was to empirically explore the relationship between team size and productivity during crisis mapping efforts.

Pablo_UN_Map

Duncan and Team from Microsoft simulated the SBTF’s crisis mapping efforts in response to Typhoon Pablo in 2012. At the time, the United Nations Office for the Coordination of Humanitarian Affairs (UN/OCHA) had activated the Digital Humanitarian Network (DHN) to create a crisis map of disaster impact (final version pictured above). OCHA requested the map within 24 hours. While we could have deployed the SBTF using the traditional crowdsourcing approach as before, we decided to try something different: microtasking. This was admittedly a gamble on our part.

We reached out to the team at PyBossa to ask them to customize their micro-tasking platform so that we could rapidly filter through both images and videos of disaster damage posted on Twitter. Note that we had never been in touch with the PyBossa team before this (hence the gamble) nor had we ever used their CrowdCrafting platform (which was still very new at the time). But thanks to PyBossa’s quick and positive response to our call for help, we were able to launch this microtasking app several hours after OCHA’s request.

Fast forward to the present research study. We gave Duncan and colleagues at Microsoft the same database of tweets for their simulation experiment. To conduct this experiment and replicate the critical features of crisis mapping, they created their own “CrowdMapper” platform pictured below.

Screen Shot 2016-04-20 at 11.12.36 AM Screen Shot 2016-04-20 at 11.12.53 AM

The CrowdMapper experiments suggest that the positive effects of coordination between digital humanitarian volunteers, i.e., teams, dominate the negative effects of social loafing, i.e., volunteers working independently from others. In social psychology, “social loafing is the phenomenon of people exerting less effort to achieve a goal when they work in a group than when they work alone” (1). In the CrowdMapper exercise, the teams performed comparably to the SBTF deployment following Typhoon Pablo. This suggests that such experiments can “help solve practical problems as well as advancing the science of collective intelligence.”

Our MicroMappers deployments have always included a live chat (IM) feature in the user interface precisely to support collaboration. Skype has also been used extensively during digital humanitarian efforts and Slack is now becoming more common as well. So while we’ve actively promoted community building and facilitated active collaboration over the past 6+ years of crisis mapping efforts, we now have empirical evidence that confirms we’re on the right track.

The full study by Duncan et al. is available here. As they note vis-a-vis areas for future research, we definitely need more studies on the division of labor in crisis mapping efforts. So I hope they or other colleagues will pursue this further.

Many thanks to the Microsoft Team and to SBTF for collaborating on this applied research, one of the few that exist in the field of crisis mapping and digital humanitarian action.


The main point I would push back on vis-a-vis Duncan et al’s study is comparing their simulated deployment with the SBTF’s real-world deployment. The reason it took the SBTF 12 hours to create the map was precisely because we didn’t take the usual crowdsourcing approach. As such, most of the 12 hours was spent on reaching out to PyBossa, customizing their microtasking app, testing said app and then finally deploying the platform. The Microsoft Team also had the dataset handed over to them while we had to use a very early, untested version of the AIDR platform to collect and filter the tweets, which created a number of hiccups. So this too took time. Finally, it should be noted that OCHA’s activation came during early evening (local time) and I for one pulled an all-nighter that night to ensure we had a map by sunrise.