Categories
ai technology

The Humanity In Artificial Intelligence

Algorithms, artificial intelligence, and machine learning are not new concepts. But they are finding new applications. Wherever there is data, engineers are building systems to make sense our of that data. Wherever there is an opportunity for a machine to make a decision, engineers are building it. It could be for simple, low-risk decisions to free up a human to make a more complicated decision. Or it could be because there is too much data for a human to decide at all. Data-driven algorithms are making more decisions in many areas of our lives.

Algorithms already decide what search results we see. They determine our driving routes or assign us the closest Lyft and, soon, they will enable self-driving cars and other autonomous vehicles. They’re matching job candidates with applicants. They’re recommending the next movie you should watch or product you should buy. They’re figuring out which houses to show you and the whether you can make the mortgage payment.The more data we feed them, the more they learn about us, and they are getting better at judging our mood and intention to predict our behavior.

I’ve been thinking a lot about these systems lately. My son has epilepsy, and I’m working on a project to gauge the sentiment towards epilepsy on social media. I’m scraping epilepsy-related tweets from Twitter and feeding them to a sentiment analyzer. The system calculates a score that represents whether an opinion expressed is positive, negative, or neutral.

Companies already use sentiment analysis to understand the relationship with their customers. They analyze reviews and social media mentions to measure the effectiveness of an ad. They can inspect negative comments and find ways to make a product better. They can see when a public relations incident turns against them.

For the epilepsy project, my initial goal was to track sentiment over time. I wanted to see why people were using Twitter to talk about epilepsy. Were they sharing positive stories or were they sharing hardship and challenges? I also wanted to know whether people responded more to the positive or negative tweets.

While the potential is there, the technology may not be quite ready. These systems aren’t perfect. Context and the complexities of human expression confuse even humans. “I [expletive] love epilepsy” may seem to express a positive sentiment to an immature algorithm. The effectiveness of any system built on top of them is limited by these algorithms themselves.

I thought about this as I compared two different sentiment analyzers. They gave me different answers for tweets that expressed a negative sentiment. Of course, which was “right” could be subjective. But most reasonable people would have agreed that the tone of the text was negative.

Like a child, a system sometimes gets a wrong answer because it hasn’t learned enough to know the right one. This was likely the case in my example. The answer given was likely due to limitations in the algorithm. Still, imagine if I built my system to predict the mood of a patient using the immature algorithm. When the foundation is wrong, the house will crumble.

But, also like a child, sometimes they give an answer because a parent taught them that answer. Whether through explicit coding choices or biased data sets, systems can “learn wrong”. After all, people created these systems. People, with their logic and ingenuity, but also their biases and flaws. A human told it that an answer was right or wrong. A human with a viewpoint. Or a human with an agenda.

We create these systems with branches of code and then teach them which branch to follow. We let them learn and show enough proficiency and then we trust them to keep getting better. We create new systems and give them more responsibility. But somewhere, back in the beginning, a fallible human wrote that first line of code. It is impossible for those actions to not influence every outcome.

These systems will continue to be pervasive, reaching into new areas of our lives. We’ll continue to depend on them and even trust them because they make our lives easier. And because they get it right most of the time. The danger is assuming they always get it right and to not question an answer the feels wrong. “The machine gave me the answer so it must be true” is a dangerous statement, now more than ever.

We dehumanize these programs once they make contact with the cold metal box that they run in. But they are extensions our humanity and it’s important to remember their human origins.

Categories
future technology

Extending Trust In To The Home

Many years ago, I contracted with a popular pest control company to service my home. At first, I met technician at the house. I would leave work, wait for him to arrive, wait again for him to complete the job, and then I would head back to work. This routine went on for a few months before the inconvenience became too much. I decided to set up a temporary code on the garage for service days that he could use to let himself in.

At first, the arrangement worked well. But, one evening after he visited, I noticed my laptop that I had left on the side table had moved. I flipped it open to see a host of unsavory advertisements filling the desktop. There was no one else in the house that day, so I immediately deduced what had happened. After a round of disinfecting both the laptop’s hard drive, screen, and case, I devised a trap.

After the technician’s next visit, I returned home and again opened my laptop. While I didn’t see advertisements, a check of the Internet traffic revealed the truth. During the time that the technician said he was in my house, someone visited a plethora of adult websites. Needless to say, my next call was to the pest control company to report what happened and cancel my service. Since then, I have been reluctant to allow anyone into my home while I am not there.

Trust is a hard thing to gain but an easy thing to lose.

At least, it used to be.

Getting into the car of a stranger used to be terrifying. But now, people will jump into an Uber or Lyft because an app on their phone connected them. The sharing economy has us placing our trust in someone that we know nothing about. Why? Because other people did and provided a rating. Because a software algorithm turned those experiences into a score. Because that score has become a substitute for our own experiences and judgments. And because the score provides a mechanism for the transfer of trust.

We don’t need the system to provide real trust, though. We’re relying on it to provide ‘just enough” trust so that we can take advantage of the comfort and convenience at minimal perceived risk. We’re depending on the model to weed out the unsavory and to reward the top performers. And for the most part, it seems to be working. We’re so comfortable with the system now that people are getting into the wrong Uber or Lyft simply because a car is parked outside their house.

Other industries are using the same model. Airbnb hosts open their homes to strangers. Both sides use the score to decide if the benefits of the transaction are worth the risk. The rating in peer-to-peer lending provides an alternative to a credit score. Each of these industries creates their own scoring algorithm using relevant metrics. The success of each system depends on whether customers are willing to believe enough in the rating to complete an exchange.

Which brings me back to my story. Letting a stranger into my house unattended proved to be a terrible idea. Like most people, I don’t want to open my house all the time. I am a consumer of the convenience afforded by the sharing economy, not a provider. But I also don’t want to be home every time I need something done. Where is the intersection of inconvenience and my willingness to remotely unlock my door to a 5-star employee of a service provider?

If I can find a way to overcome the barrier, the potential benefit is huge. Devices are getting smarter and know when they need service or are about to fail. Instead of notifying me to call a technician, the device should make the call itself and cut out the middle-man. When my WiFi network can’t optimize itself to give me the best experience, it can bring in my service provider. Services that traditionally stop at the front door can enter the home. My grocery delivery should make it into my fridge and pantry. My dry cleaning should hang in the closet.

I want to find that place but, having been burned before, I don’t know if a community rating is enough to earn my trust. Technology (read: surveillance) might help me bridge the gap, though. The stream of data coming from inside my house includes cameras, door and motion sensors. I have the ability to peek in or, at least, the technician knows that I can. I know when a door opens and closes, so I know how long they are in the house. I can see what rooms someone was in and for how long. And I can always play the tape back if I suspect something is amiss.

Is that trust? Maybe not. But history has taught me to operate with a different philosophy: Trust, but verify.