Platforms like Google and Instagram are our habits. Content are the commodity peanuts on the bar. Getting to be a habit is really hard and wildly under appreciated. AI-infused search has to be better than results. They have to give us that warm-fuzzy feeling that makes them habitual.
— — —
Ewe sort of know his name when you hear it. He has an advanced computer science degree from one of those fancy west coast schools. Stanford or Berkeley or something like that. He’s a wiz when it comes to search engines. In fact, his company created the world’s most advanced way for you to find things online. Sorry, *You.
The year was 1998. Yahoo was our entry point to the world wide web. It organized links into categories and put the major ones on its home page. It looked like this…
You would narrow your search down from a topic like News & Media to category like magazines and to a publisher like People to a story. And, if you knew what you wanted you could also type it into the big empty search box atop the page. You almost missed it there, didn’t ya?
Fun fact, my college roommate coded the tech that became Yahoo and Lycos and Excite and the search engines for today. His other 235 patents are the reasons our phones access the Internet and how take most of mobile computing for granted. If I ever write a book, it will be called, Asymptotic to Life. My seat really close to people who do really cool stuff.
Any-who back to our story. Along came two smart guys that changed the core of the search engine. I’m not nearly smart enough to do it justice. So, here – again – I’ll ask my uncharles AI bot, açaí to do the heavy lifting.
Açaí: How did early Google differ from Yahoo?
Well, let me tell you. Yahoo’s index was hierarchical. You know like layers or levels. In fact, Yahoo means Yet Another Hierarchically Organized Oracle. I think Mike told you this. Remember? Google was more organic. Instead of just looking at pages title and how many times it was clicked, early Google looked at how our pages linked to each other to get a sense of value.
Trainer note: I need to find a way to tone done its snark and condescension. And, trust me, that, “Let me tell you,” thing is getting really tiresome. I don’t know why it looks for catchphrases.
In short, early Google’s search results were 10% better than Yahoo’s. There is no objective measure. That’s how I remember it. So, that’s my number.
Google didn’t crush all the search that came before it because it was better. Or because it had some quirky inside baseball name derived from geekdom. Or because it’s a simpler page. No, they won because it captured user search data and linked it to ads.
Yahoo may have looked modern… trust me, when it launched in 1994, it was the world’s most futuristic card catalog machine – ever. But it was playing last generation’s game. Ads were endemic. Buds ads were pre-coded to Sports Illustrated. Max Factor ads to People. IBM PC ads to Fortune. BTW, Time, Inc. owned SI, People, Fortune, and more when each was ad powerhouse.
Google got to know you. Charles searches for cars. It doesn’t matter where he goes, show him car ads. Trust me. Their search engine could sense where ads could go. It’s like it had a sixth sense. An ad sense. Which they did. When they launched Ad Sense in 2003. If you’re looking to know when modern digital advertising started it was June 18, 2003. Ad tech folks should throw a party for it and buy it beer. It turns 21 this year.
Google went public in 2003. In 2004, it doubled its revenue and quadrupled its profit. It didn’t take the smart folks at Google long to see that they had a perpetual money motion machine. Get asks, collect data, link it to ads, show ads, make a ton of money, create new products to get more asks. Repeat.
People never see the, “Create new products to get more asks.” That’s what this entire post has been about. Google didn’t win the search was because their PAGE is empty or their search is marginally better. They won because they became habitual. Ewe see, that’s the hard part. And, that’s why their coffers are filled to the BRIN.
There’s a new search dog in town. It’s called Perplexity.ai. Its name derives from the geekdom of information theory and probability distribution. Aravind Srinivas leads the team who built it. He has a PhD from Cal Tech. He was a research intern at Google for a year. For a year after that he was a research scientist at OpenAI. He left OpenAI to build a search engine based on AI.
He is damn sure his search is better. So, I tried it.
Google: What should I have for dinner?
It showed me eight choices from fajita parchment-baked chicken to oven-baked French bread pizza.
Perplexity: What should I have for dinner?
It showed me 18 choices from fajita parchment-baked chicken to hoisin glazed pork chops.
I have no objective way to measure, “Better.” I’d say, “Marginal.” You can decide for yourself.
It’s doesn’t matter if it’s baked feta pasta or cedar planked salmon. This game is not going to be won by a slick or slicker interface. It’s not going to be won on the guts of the tech that collects data or whether it’s a hierarchical index, and organic one, or an artificial one. It’s going to be won by where ads go and how the group that gets ad dollars uses those ad dollars.
You see, there was an AI search engine. It came out two years ago. It’s called You.com. It’s founder and CEO is Richard Socher. He has a PhD in CS from Stanford. He worked in AI, and for Microsoft, and for Salesforce. Marc Benioff the CEO of Salesforce gave You $20M in seed money. It’s based on OpenAI. And, despite Dr. Socher’s claim that You delivers better results, it’s flopped.
Because it never became a habit. You didn’t say, “I need to You this thing.”
You didn’t even care enough to say, “Eww.” Not nearly enough people replaced the Google app on their Android phone to make it their default. They couldn’t pay Apple about $30 billion to get Siri to search with it. Which means it never collected enough data to serve smart ads. Which means it didn’t have the oodles of money to build Gmail, Chrome, Docs, Android phones, Chromebooks; or buy Waze. It didn’t have all the entry points Google started to figure out we wanted 25 years ago.
When you look at these screens, you realize the data is the data. For — admittedly — this one search, all the results come from the same few top source: Tasty.com. There’s only so many ways to present these results. So, it all comes down to one thing: Habit.
Unless Perplexity figures out how to become habitual for you, it doesn’t matter what you call it, it’s going to fail. Sorry, *ewe.
