Interacting with Machines

I have a complicated relationship with my GPS unit. On the one hand, it rarely works. Here's what happened the last time I turned it on. First, there was a five minute delay while it searched for the satellite signal. Then, it couldn't find the street I was searching for. Then, it found the street but lost the satellite signal. Then, it regained the signal but sent me in the wrong direction. And then, after I'd already gotten accurate directions off my phone, the GPS unit finally decided that it knew where I was going. In other words, the device sucks.

But here's the funny part: I still use the device every time I'm even a little lost or unsure of where I'm going. In fact, I sometimes turn the machine on even when I know exactly where I'm headed. Why? I'm not quite sure. Although the device drives me crazy, and I'm constantly complaining about it (see above), I also enjoy interacting with that posh British voice emanating from the gadget, as it mispronounces every street name and tells me to take the wrong turn. When I'm alone in the car, the stupid piece of plastic feels like a companion.

We spend our days surrounded by machines and microchips. With few exceptions, these machines feel like machines. We treat them like mindless tools because they are mindless tools. My laptop, in this sense, is just a fancy version of a can opener - it performs a nifty function, but I don't worry about its interior life. As a result, if the tool malfunctions - if the can opener gets rusty, or if my computer crashes when I'm working on a file - I get furious. I curse the damn device and threaten to throw it in the trash.

Why, then, am I so indulgent of my GPS unit? The answer, I think, has to do with the facade of agency. This machine speaks to me, calmly telling me where to go and why it's failing to telling me where to go. Sometimes, when the gadget is really struggling, I get the sense that it wants to apologize, that it feels bad it's so utterly ineffective.

Of course, I know my intuitions about my GPS unit are ridiculous. I know that the genteel voice isn't real, and that the device is just another piece of plastic crafted in China. But I think my response to the device captures something interesting about the human mind, which is that loves looking for other minds. Consider this elegant study, led by Scott Huettel at Duke. The scientists demonstrated that merely watching a computer "play" a video game - this is roughly equivalent to looking at a device on demo mode - led to increased activation in the posterior superior temporal cortex (pSTC), a brain area associated with the perception of agency. (Interestingly, subjects who showed more activation in the pSTC while watching the computer were also more altruistic in real life.)

The moral is that it doesn't take much before we start attributing feelings and intentions to a machine. (Sometimes, all it takes is a voice giving us instructions in English.) We are consummate agency detectors, which is why little kids talk to stuffed animals and why I haven't thrown my GPS unit away. Furthermore, these mistaken perceptions of agency can dramatically change our response to the machine. When we see the device as having a few human attributes, we start treating it like a human, and not like a tool. So here's my advice for designers of mediocre gadgets: Give them voices. Give us an excuse to endow them with agency. Because once we see them as humanesque, and not just as another thing, we're more likely to develop a fondness for their failings.

And this doesn't just apply to gadgets. Every company is now on Twitter and Facebook. For the most part, companies use these platforms as just another way to pump out banal promotional messages. However, I think these platforms actually have potential for corporate communication. The reason returns us to perceptions of agency. Consider this study led by the experimental philosophers Joshua Knobe and Jesse Prinz. They wanted to understand which psychological concepts people were willing to apply to corporations. They quickly discovered that people were perfectly happy to give companies intentionality, so that it would be acceptable to use sentences such as:

"Acme Corporation believes that its profit margin will soon increase," or "Acme Corporation intends to release a new product this January."

However, the vast majority of people refused to allow companies to have feelings or subjective experiences, so that sentences like this were deemed ridiculous:

"Acme Corporation is now experiencing great joy," or "Acme Corporation is getting depressed."

I think this asymmetry comes with important implications. When we fail to endow an organization with emotions, we're much lower tolerance for mistakes. Instead of reacting with empathy when our plane is late, or when the product is imperfect, or when the item goes on sale the week after we bought it at full price, we react with unmitigated anger. This is why I think companies are missing the real potential of social networking platforms. Twitter and Facebook and whatever comes next aren't just echo chambers for corporate messaging. (BP, as usual, is a case study in how not to do it.) Instead, these interactive platforms are a rare opportunity to endow a brand with agency, to inject some emotion into an emotionless corporation. By interacting directly with customers, by showing us a sense of humor, or making grammatical mistakes in tweets, or even apologizing for a mishap, we suddenly see the corporation as more than just a monolithic organization. We treat the company more like a GPS unit and less like a can opener.

More like this

Over at the Experimental Philosophy blog, Joshua Knobe has a post about a series of experiments that he has run with Jesse Prinz on people's intuitions about consciousness, and he includes a link to a draft of the paper they're writing on the experiments. The experiments were motivated by two…
If William James were alive today, I'm pretty sure that he'd be an experimental philosopher. (He'd also be a cognitive psychologist, a public intellectual in the mold of Richard Rorty and a damn fine essayist, filling the back pages of the New Yorker and New York Review of Books with incisive…
There's a new and very timely paper out this week that looks at the cortical mechanics of charitable giving. While it's been known for a few years that giving away money activates the dopamine reward pathway - that's why doing good feels good - this latest paper attempted to investigate the…
One of the interesting subplots of this new research on the intellectual benefits of sadness - it seems to bolster our attention and make us more analytical - is that it helps illuminate the intertwined relationship of mood and cognition. For decades, we saw the mind as an information processing…

I think it's funny that you list your laptop in the collection of mindless tools. Perhaps I simply play too many computer games, but, like the subjects of that Duke study, I psychologically assign all sorts of agency to computers (although I do also get furious with them when they don't work properly).

It does seem to be a good strategy for companies at this point. But since we're almost never completely satisfied with what we have, by the time we're adjusted to the idea of machines having genteel human voices, we'll probably long for more human-like interactions. Isn't this already true with telephones? Well, at least in my personal experience, I find it much more pleasant to leave a voicemail message when I hear a personalized recording, as opposed to merely a phone number or name.

Right, we hear commercials where companies try to convey to the viewers that they'd be talking to a "real person" not a "machine" when they dial this 800 number :)

This tendency to attribute agency to nonhuman entities may also explain such things as Pantheism and Animism. Though, as an environmental philosopher, I applaud anyone who sees a tree as alive (and as 'conscious') as a moose or a human being. In contrast, the personification of truly mindless devices makes me sad. As Nature cedes to the machine, will this sort of agency keep us company in the lonely corridors of the megalopolis?

One good example of what you are talking about is the Transportation Safety Administration blog. They use the blog to answer questions and provide information about the TSA rules, but do it through a personal (rather than bureaucratic) voice, so it at least seems as if a real human being is responding. Even a little humor is employed (although I emphasize "little"), as in this entry about shoe scanners:

"I can assure you that our officers want this technology to work just as much or more than the passengers do. This would allow them to focus their attention on other things and there would be much less clutter on the X-ray belts. So the passengers and officers would be gaining all of this added convenience and security to boot. (Shoe pun intended)"

Trying to add the illusion of agency to faceless corporations or bland products is what the process of branding has always been about. Way back in the 1970's Stephen King wrote a seminal paper called 'What is a Brand?'in it he quoted research where housewives described the 'personalities' of different products.

Social media offers a brand new approach to achieve the same end, brand managers and agencies are still figuring out how best to use them and some will, of course, always be better at it than others.

Cliff Nass at Stanford did a lot of work in this vein. Be wary of execution though--his work led directly to the development of Microsoft Office's much-derided Clippy.

Haven't we known this since at least Eliza?

By Peter Lund (not verified) on 30 Jun 2010 #permalink

The facade and perceptions of agency. Very interesting article. Thanks.

Admin Note: When I try to share this on my favorite social networking site I get HTML tags and not the article itself. I don't expect anyone to get to this today (Happy 4th of July).

By Robin Redden (not verified) on 04 Jul 2010 #permalink

I agree that branding is the personality of the corporation, and social media is an excellent tool for conveying that. This means, though, changing our idea of branding and making it more human and less rigid. Just like humans who display a variety of emotions, so you will find the same via social networking. The idea of a one-size-fits-all elevator pitch seems small in comparison. My question is whether all companies and products need these personification? B2C, in general, seems better suited than B2B, or will personification (i.e., "meet the iPad) always win out?

With the introduction of Kinect, your wish for computers to understand our feelings is becoming reality. Recognizing posture plus facial expression plus voice can help determine mood, sentiments.