"The perfect search engine would understand exactly what you mean and give back exactly what you want."
Larry Page, Google co-founder, 2006.
Google built the biggest advertising business in the world on the back of a near-perfect business model. It indexed a world of content it didn’t pay for. It bore no responsibility for that content. In using the index, users signalled their interests and intentions, in effect filling in their own advertising brief. Google set up a system where advertisers bid against each other for access to this audience it understood in detail. Everyone used Google, all the time. Even when not on Google, users were tracked through its widespread ad and analytics tech.
Instead of becoming a worse search engine as it grew, Google got better. Success fed success. Competitors couldn’t match its usefulness. People even liked the company.
There’s never been a better ad business, and maybe there never will be. But that era is over.
The flaw is there in Larry Page’s 2006 quote above. Page knew then that despite the beautiful business model, Google search was not perfect. It did not understand exactly what you meant, and it did not give you back exactly what you wanted.
Instead the audience had to be trained to work with the search engine, making multiple inquiries until the right results came up, assessing the credibility of web pages, and reading a lot of text.
That all seemed natural for many years. We, the users, adapted and got the answers we needed. Then along came ChatGPT and it turns out that actually, search is a pretty clumsy way of answering questions. It’s not how humans think or act. An AI which understands conversational questions and answers immediately is better.
In the past two months the number of Google searches I am doing has dramatically fallen because for anything that requires an answer - rather than a link or a map - ChatGPT is easier, faster and often better. My experience is backed up by experimental data. In research conducted last year, ChatGPT was 65% faster than Google search over a set of tasks that included retrieving facts, fact checking, and finding links.
Google is fighting back with the introduction of AI summaries at the top of search results. There are two problems with this. The first is that the AI summaries break the implicit deal Google had with the web: “let me index you, and I’ll give you traffic”. The second is that it’s still a results page. To the man with a hammer everything looks like a nail. To the company with a search engine, everything looks like a query. It will be hard for Google to move away from paid links on a page. That’s how it makes billions.
Thinking out loud
ChatGPT is still in its infancy, but already it is challenging the query-based view of the world. I don’t think OpenAI set out to do this when it made ChatGPT, but it is happening. The AI’s next-level voice-recognition and speech capabilities mean that users can not only answer questions, they can interact with the AI in a way that enhances their own thinking.
Thinking out loud to ChatGPT’s advanced voice mode is very useful. Speaking to a machine that can understand you is liberating.
Try it at the start of your working day (Advanced Voice Mode is now available free for a limited number of minutes per month). Tell it what’s on your mind, what’s coming up at work, and then ask it to summarise back to you. Maybe you’ll change priorities. Maybe you’ll remember something you forgot. Whatever happens, you’ll be thinking more clearly.
I think conversation-dominated interaction will lead, in time, to a different advertising business model. People will be speaking with AIs for an increasing number of minutes every day, and that two-way conversation presents a unique opportunity for monetisation. Media technologies that dominate attention (radio, TV, social media) have always been good for advertising, and AI conversation will dominate attention.
Search isn’t going away. Typing a query into a browser bar is still a default behaviour. But search’s use in finding answers will decline significantly, and that is going to mean massive change.
Good old Google
In tracking down the source of the above quote from Larry Page, I had to trawl through the Wayback Machine, the site that takes historical snapshots of the web and archives them. The quote is taken from Google’s corporate site on September 10, 2006, and appeared in the Our Philosophy section. The philosophy is a list of “Ten things Google has found to be true”. 18 years on, they haven’t aged well. Often Google has breached its own commandments, or they turned out to be wrong.
1. Focus on the user and all else will follow.
2. It’s best to do one thing really, really well.
3. Fast is better than slow.
4. Democracy on the web works.
5. You don’t need to be at your desk to need an answer.
6. You can make money without doing evil.
7. There’s always more information out there.
8. The need for information crosses all borders.
9. You can be serious without a suit.
10. Great just isn’t good enough.
I think Google’s biggest problem is the first commandment. With market dominance came an institutional frame of mind and a focus on advertisers above searchers. My own experience with Google Analytics and operating on YouTube indicates that even the advertiser focus has declined. The near-perfect business model bred complacency. That’s natural.
We’ll see how competitive Google can become, now it is facing a foundational challenge for the first time in more than two decades.
Siri crap for two more years
Last week Apple announced it would revamp its voice assistant Siri using the kind of AI technology that powers ChatGPT Advanced Voice Mode. The product will be released in 2026.
This is an astoundingly slow response.
Siri was released 13 years ago, along with the iPhone 4s, and the user experienced hasn’t changed. You can’t have a conversation. Siri knows almost nothing. It frequently mishears what you are saying, and can’t remember the simplest things about you.
It’s a mark of Apple’s dominance that it could continue to fumble the voice opportunity for a decade and still come away unscathed. Will its luck continue to hold until 2026?
Probably, because of Apple’s smart move to lock up OpenAI with a lucrative deal, and because it has such a tight grasp on the interface so many millions of people use everyday. To quote media commentator Brian Morrisey: “Controlling the interface is the closest you’ll get to commanding heights.”
Too bad consumers will have to put up with Siri for another 2 years.
Check out this 2011 ad for Siri if you think I’m exaggerating about how the product has gone nowhere.
Hope you are having a great week. It’s damn hot in Sydney.
Hal
Google Search has always been AI. It long ago got to the point where any phrase you throw at it will almost certainly return a page containing what you're looking for above the fold. It started surfacing answers in search results before ChatGPT was publicly available.
Kodak invented the first digital camera. Xerox invented the modern computer GUI. GM had an electric car on the market in the 1990s.
Google Search is now a legacy media company. It has a business model reliant on advertising. Unlike Fairfax's classifieds and Telstra's directories of 25 years ago, it doesn't need to move from paper to online. It has a smaller jump to make.
Most of it's advertisers are selling a product or service. It could fairly easily transition this advertising to paid results in a chatbot.