AI: Why Google’s ‘AI powered Search’

...will win the AI ‘ChatGPT’ round

I’m calling it now in 2023. Google is going to do to OpenAI's ‘ChatGPT’ over the next 12–18 months, what Microsoft did to the Netscape browser between 1996 to 1998. And this time, Microsoft, even though it’s on the opposing team, will not be able to change that outcome. Yes, the ‘Browser Wars’ then, to the ‘AI Wars’ today’. 

Several reasons for calling this now in the doldrums of summer, and seemingly early. 

First, I was there at the front seats the last time, seeing and studying it up close and personal. And the pattern recognition is strong and vivid across time. 

Second, the Force of incumbency is strong with this one (Google), as it was for Microsoft in the 1990s, powered especially by Windows 95. 

Third, now as back then, it matters not what the early adopters love and want (the Netscape browser then and ChatGPT now), it matters what regular users are used to right in front of them. Habits are hard to change, especially at Scale. 

Windows 95 with the Internet Explorer browser built right in for free then, and AI powered Search built into Google Search now with SGE (Search Generative Experience) and Bard now, and Gemini soon. It’s all going to be more than ‘Good Enough’. Even if ChatGPT remains the technically better mouse trap. 

Habits matter. Especially habits ingrained for hundreds of millions and billions. They’re hard to change for the mainstream. Not to mention all the partners with an economic stake in the current ecosystem. And almost impossibly expensive. 

As Google is fond of underlining every opportunity, and as CEO Sundar Pichai did in the  just released quarterly results call:

“With 15 products that each serve half a billion people, and six that serve over 2 billion each, we have so many opportunities to deliver on our mission.”

He didn’t have to add that the world’s 3-4 billion of the total 8 billion, search on Google over 9 billion times a day, before they search for anything video related in Google’s YouTube. 

That’s 90% plus of the global Search market OpenAI ChatGPT and Microsoft Bing notwithstanding. Eight months after OpenAI’s ChatGPT moment. And of course more OpenAI LLM AIs to come beyond GPT4. 

As The Economist pointedly asks in a detailed piece this week with its very title,

“Is there more to Alphabet than Google Search?”

My answer is, that’s enough. Even after 25 years of trying so many ‘Other Bets’ & things beyond Search. The core thing they did was to do and fund the Basic Science behind this generation of AI, not the least of being the seminal Google ‘Transformer’ “Attention is all you need” AI paper in 1917. And Deepmind with their AI powered robots for years, and RT-2 now. Regulatory headwinds notwithstanding.

And all the post ChatGPT ‘Moment’ ‘Code Red’ actions for their Empire to rise to the LLM AI rebels. Even if they seem at times to nudge, shove, talk, and/or plod Back, they’re in the game. 

It’s worth repeating that Google isn’t coasting on the flywheel of their incumbency. They’re hard at work on new stuff, like infusing Generative AI into robots, as I described yesterday. Of overhauling their ubiquitous voice product, Google Assistant, with AI technologies like Bard. The giant who has long considered themselves an “AI First’ company, is indeed thinking of AI first in most things they do, from CEO Sundar Pichai on down.

So, for all those reasons and more, in my view, they’re going to win this AI ‘Chat whatever’ round. Just like Microsoft did against Netscape almost a quarter of a century ago. Stay tuned.