I sent my first e-mail message in 1995, to a member of my development team. That was the only person I knew who had an e-mail address in those days. I also did my first web search around that time. I think I used Lycos for this. I entered some keywords into a text box, separated by Boolean operators, and received a list of web pages that I could click on that referenced these words.
Sixteen years has passed. I receive about 400 e-mails a day now from people all over the world. E-mail has become part of my life and has changed the way I communicate and the way I work. I don’t know anyone anywhere who doesn’t have an e-mail address. When I went to Sikkim, India, last year, a Buddhist monk in a remote Himalayan monastery even gave me his e-mail address. The web has also evolved in a similar fashion—it seems to be everywhere and connects everyone, for everything. Internet technologies are now toppling dictatorships in the Middle East.
But what has really changed in search? We still go to the same text boxes, enter expressions that we hope the computer will understand, get back lists of web pages that reference those words, and click on links to find the information we are looking for. The only real difference is that now the top links take you to spam sites—which want you to click on other links that make them money and that make Google money. Creating low-quality, low-cost information pages has become such big business that the leading content farm, Demand Media, just went public and is valued at $1.9 billion. According to Blekko’s spam clock, over 1 million spam pages are created every hour. So the web is becoming one giant heap of trash.
I had hoped to put Google and Bing on the spot at the recent BigThink conference by challenging them to fix the spam problem; perhaps to follow Blekko’s bold lead in blacklisting the leading polluters. But they instead got into a pissing match about who was copying whom. Google simply changed the subject. And when I asked the panelists about their long-term vision on search, I was really disappointed at the shallowness of the response. They weren’t talking about changing the world—just about fine-tuning what they\’ve been doing forever. You can watch the video below to see what I mean.
In the hope that I don’t have to wait another 16 years to see advancements in search, I’ll share my views on where it needs to go. Perhaps you can also share your views and we can inspire a new generation of startups to do to the current search leaders what Google did to Lycos and AltaVista: antiquate them.
In a nutshell, what I want is for my computer—or a new Google—to serve me. I don’t want to be serving it words that make sense to it, and then have to wade through pages of spam it delivers to me, to find the information I am seeking. I want it to learn what I like and what my friends like and tell me what I want to know or do what I need it to.
If I am visiting New York City tomorrow and want to eat dinner at a moderately priced North Indian restaurant near where I am staying, I want my computer to suggest the two or three places that I will like and that have space available. I book my flights and hotel reservations on line, have my calendar on Google, tweet my likes and dislikes, and talk to my friends on different social media sites, after all. So why can’t my new Google simply take my information and my friends\’ information and give me what I want?
Is this so hard? I don’t think so. This week, we witnessed a computer, Watson, beating the top Jeopardy contestants. It was able to parse human speech patterns, make sense of complex questions, do very sophisticated searches, and come back very quickly with the right answer. It didn’t respond with a series of links—it computed the probability that its answer was correct and responded accordingly.
Watson’s technology is a great start, but I want much more. I also want it to analyze my social graph and get recommendations from friends who matter. For example, when it comes to Indian food, I don’t care what my academic peers say or what my South Indian friends say; I want input from fellow Punjabis—they know their tikka masalas and saag paneers better than anyone else does. If I am looking for health-related advice, I want to know what doctors say. If I am shopping for a gift for my wife, I want input from women who share her tastes. This isn’t rocket science.
And then I want more. I want my new Google to automatically make a dinner reservation for me, buy me a ticket to a movie that I may want to watch, or place an order on the cheapest and most reliable shopping site. Yes, I know there are already many applications/sites that do this. Why aren’t my preferred sites integrated into the search function—so I never have to see the 90s-era text links?
What I really don’t ever want to see is the spam that Google delivers. The present page-ranking system is easy to bait—just add the right key words to some garbage content or pay Google for an ad, and your listing appears at the top of everyone’s search results.
Google took some good steps forward this week with its announcement that it will let users tag sites as spam via a Chrome extension and rank the websites of people they know higher than others. But this is still more of the same—the spammers still get top billing. Just do a search on a term like “digital camera under $200” to see what I mean. The results are practically useless. (Bing just added a nice option to let you search by price, but that is buried in its spammy results when you do a regular search).
We need some out-of-the-box thinking here. I doubt we will see this from Google, because it makes billions by serving up ads. So here is an excellent opportunity for entrepreneurs.