AI & Culture Year in Review and Outlook for 2017
Last year was tough, not just because of all the celebrities who died. 2016 proved tough for artificial intelligence (AI) because every news story that trumpeted stunning advances in AI capability was balanced by other reports that highlighted AI’s limitations, its foibles or massive social or economic implications of an AI-powered society.
For once, what a new technology meant became as important as what a new technology could do. Effectively speaking, we discovered that AI can be just as biased as us because AI learns what data sets can teach it. And if those data sets contain our best and worst biases, so will our AI system
Along with being biased, we also learned in 2016 that AI can kill you. In May, a Florida man died when his Tesla Model S slammed into a large truck while the car was in self-driving mode. Of course, compared to the annual toll of 40,000 deaths on American roads with human drivers, the Tesla crashes in Florida and China haven't stopped development of autonomous vehicles.
Moreover, voice activated virtual assistants were no more immune to being connected human deaths than self-driving automobiles. A legal tug-of-war erupted between Amazon and the state of Arkansas over access to potential evidence in a murder case that might have been captured an Amazon Echo device.
If 2016 brought AI into the mainstream warts and all, what might we expect in 2017? What should we hope for? What should we expect? And what should we fear?
The most significant and doubtless difficult improvement we hope to see is a new cultural platform for talking about privacy in an augmented world. To be blunt, there is no real privacy in an augmented environment. We need to talk about what are we giving up and what are we getting in return.
What do we expect in 2017? That one is pretty easy. It's a full blown rush to grab augmented territory. This is a DOT.COM stage for AI. Any and everybody is going to slap on the augmented label to their marketing and products whether or not it makes sense.
Another thing we expect is that as more intelligent assistants like Alexa penetrate homes and automobiles, there will be efforts to develop distinct "child modes" for how these devices interact with kids.
We expect to see more legal cases involving intelligent products in 2017. While the Arkansas murder case is the most extreme example of unexpected outcomes, far more likely will be divorce proceedings where recorded activities by intelligent products and services become material to the final outcome.
The simple truth is that the more enmeshed these products & services become in our lives, the more enmeshed they will become in ALL aspects of our lives, including the messy ones.
At the Augmented City, we aim to chronicle this evolution for the point-of-view that AI is neither a pure blessing or a pure curse on humanity. AI is a new, extremely powerful medium for extending our senses and augmenting our minds.
When you change how people experience reality, you simultaneously change what people think is normal. So we better ensure AI bends toward humanistic outcomes or machines will definitely kick our ass this time around.
There is no time better than now --- 2017 --- to think critically about how culture and code will evolve in tandem.