What the heck is Electronic Mail?
What kind of man owns his own computer?
Why is old technology so funny?
The web is awash with old technology ads, whether they're for a cheaper hard disk, better color, or even an "Electronic Computer Brain" that's fun at parties, solves riddles and tells the future.
Technology advertising has existed for nearly a century, continuously promising a life made easier thanks to better, cheaper and faster gadgets. The only thing that changes is the art direction.
At the time these ads announced hard-won innovations in power and speed. Now, they invoke simple amusement at the way things used to be.
Why is this?
I posed the question to Don Campbell, a longtime tech watcher and IBM Distinguished Engineer. What ensued was a fascinating discussion about the pace of innovation, our relationship with the technology we build and what we'll find funny 20 years from now. Below are the highlights:
Why do we find old technology ads so amusing?
There was always a lot of romance in these ads about what a system could do for you - it would make you better, smarter, faster, richer, faster - all of those things. And yet they never drew the connecting point between the technology of the day and what you'd get as a result. The buyer needed to make the leap of faith that by the time you acquired the technology and figured out how it worked that the maker would have filled in the blanks between what it could actually do - which in most cases was add numbers together - and something that could make a meaningful difference in your life. It's the gap that we never got over.
What does this say about us? Why do we keep buying into the romance?
We want a partner in our lives. We want technology to be that partner - not just a tool that we use, but a partner that will take care of the boring stuff that we don't want to do. And we humanize it into being a partner.
At every leading edge of technology, there is a little more of that human element that we want to give it. I think we have some of those feelings for our iPads or our favorite mobile devices. But when we realize that a particular device didn't become our partner, then it looks like one of those old ads.
Watson wasn't human - but it used human language and was built to compete against humans using human rules. Could Watson be that partner?
What interested me about Watson wasn't the system called "Watson." I don't have a lot of applications in my business for winning on Jeopardy! What did interest me were the analytics possibilities that it opened up if you were to combine its components with other analytics capabilities. For example: it breaks down a problem and attacks it in multiple domains; it understands its confidence in each of those domains; it combines those confidence levels into an overall confidence level and a threshold that determines how to act. Sometimes it had the right answer but wasn't confident enough to buzz in.
These are attributes that you can apply to your business - we're getting all kinds of unstructured content from blogs, tweets and message boards, but we're not sure how much we should trust them. Now, we can apply Watson algorithms and capabilities to this data to help us understand whether or not we should trust what's being said, whether we should respond and how.
There's also the gaming aspect, which is quite appropriate for our businesses. In some cases, I might want to make as much money as I can as quickly as possible; in others, all I might need to do is beat a specific competitor in one area at a particular time. In that case I'll want to reduce my risk and choose different tactics to drive a better outcome. So when we think about the gaming part of our business, I think Watson has a lot to teach us.
Ken Jennings said that on most nights, the top Jeopardy! players know most the answers; winning is more a matter of who buzzes in first. Our businesses are less about the specifics of Jeopardy! and more about trying to maximize how often you're right given the information you have and how well you can pull it all together. The algorithms in Watson help it accomplish this and that's extremely valuable in a business context. I was very impressed with how well it dealt with the nuances of the language and the complexity that was behind those questions to get most of its answers right.
Does Watson eliminate the disconnect between the promise and performance? Does it follow through on the romance?
Watson's success is a tremendous accomplishment. But no matter what we say today, you know that 20 years from now we'll be having this same conversation. No matter now good our technology is now, we'll all be saying, "Can you believe it took a room full of rack-mounted servers to do that? Let me just ask my watch." But the fact that Watson had more human characteristics - it spoke, it played a game that humans play in a way that felt more human, it took natural language as an input and produced natural language as output, and even conversed as it chose each question - that will drive it more into our comfort zone.
Technology will always change. But will what we want from technology change as well?
I think our expectations will continue to grow in the same space. We've made tremendous progress in many domains. But in some respects we've not yet lived up to the ads of 25 years ago. Until we do, there's a lofty goal out there for a system to support us in the way we want to be supported. We're getting better at allowing more people to participate in technology and benefit from what it can do for us - for example we don't have to type in lines of code from the back pages of Byte Magazine to make our system run - but I think in general we still want that partner in our lives that will make our world easier to manage and that will free us up to do more things. We've made progress, but too much technology still makes us perform tasks on its own terms.