On Talking about AVs (as Robots)……

Since I started working on policy for automated vehicles (AVs) in 2017, I have been struck by how difficult it is to talk about the role of robots in society. Not only does the lack of a lexicon inhibit the development of universal concepts, the lack of a conceptual framework prevents society from assessing how the development and use of robots in society should be guided.  

Today in tech policy conversations, it is easy for phrases like “artificial intelligence” and “machine learning” to be used at the drop of a hat, but the actual meaning usually has to be teased out and is context specific. This creates wide variability in how these phrases are used and defined, leaving most folks wondering how reliable the definitions are when describing technological advancement. 

To fill in the gap, I sometimes feel I’m stuck using imagery and language from sci-fi films to get complex ideas across on human-robot interaction. However, as I have realized, those descriptions are limited and do not adequately detail new interactions that require humans to think deeply about what is being asked and the implications of using more robots in society.  

For example, how might AV regulations need to account for an unaccompanied child riding alone? First, regulators have to look at the AV as a complex machine that if used incorrectly could pose great danger to both children occupants and other road users. Next, regulators have to work with designers to determine if a particular AV design requires additional equipment, such as a car seat, to ensure risks to children are minimized if a crash occurs. Then, the tricky task of how to structure child-robot communications will emerge: How should communication protocols be structured to allow children to roll down a window or operate an audio-visual entertainment system, but not give the software operating the vehicle instructions that could lead to a crash?  

Parents will likely ask these kinds of questions before they ever consider letting their child ride alone, but contemplating them may raise wider societal conversations that we should be prepared for in short order. It may be that society determines riding alone in an AV is too dangerous for children, or we create new social customs (and legal requirements) on human operation of an AV. Perhaps only licensed drivers should be allowed to ride alone in a vehicle, in the event that a human needs to intervene or assume control to avoid a dangerous situation.  

Only collective society will determine how we approach the issues that arise from greater adoption of robots. As AVs become a closer reality for the public, we will have to grapple with how to talk about robots becoming drivers that share the roads with humans. The sooner we can engage fully and authentically in that conversation, the more likely it is that we can develop and implement a governance structure for the use of robots in society.