AI

What Google I/O tells us about BigTech's strategy with AI

Nikhil Kapur

• 3 min read
What Google I/O tells us about BigTech's strategy with AI
Google I/O or AI

In continuation with my previous post about AI in BigTech vs Startups...

By now you're probably been inundated with several memes of Sundar Pichai, stuck in a time-loop droning “AI”.

Sundar Pichai - AI

It reminded me of a talk I attended as a MSFT dev with a sweaty Ballmer jumping up and down the stage shouting “Developers”. Well that didn't end well for him :)

Steve Ballmer - Developers

Google I/O Announcements

Based on the new announcements from Google I/O, it's quite clear that as of now, in the Application layer, big tech is busy injecting AI left, right and center into its existing applications - Consumer and Enterprise. We are likely to see bots, copilots, recommendation systems emerge in pretty much all the apps we use within a short span of months, if not weeks. Here is just a sample list of announcements made during Google IO.

What's also becoming clear is that BigTech is not currently thinking of how to build AI-native or AI-first apps. These AI-native apps need to work at a higher level of abstraction than the current input control parameters that applications rely on, but the incumbents seem to be bogged down by the same thing that enables them - incumbency.

  • Think Spreadsheets. Instead of a giant table of text and numbers entered by data, and features such as Autocomplete, Suggestions, etc, what the user really needs is insights and analysis on top of this data presented to them in a much more intuitive manner? In near future, there might not be even the need to “pull numbers into a spreadsheet” and “crunch them”.
  • Think Word Processing. Instead of a creators-block-removal paragraph, or a blog post written and edited by a bot - as already implemented by the likes of Notion, Word, and Docs, what the user really needs is a way to transform ideas into text in their own original voice? Or a way to restructure the flow of their existing text and make it short and to-the-point or long and detailed as per their requirement? Does the interface even need to look like a blank canvas any more?

Business Model Disruption

In VC there is a popular theory, and I somewhat agree with it, that for startups to win during a technological revolution, the fundamental question comes down to whether “Incumbents will adopt Innovation faster than Startups find Distribution”? While I do lie in the camp that simply adopting AI as an API will generate much more value in the BigTech companies, I also feel that this statement disregards true business model innovation.

If the blank canvas or an empty spreadsheet is not as valuable anymore, then why do I as a user pay $8 per month for that piece of canvas. If the power is shifting from platforms to intelligence, then I might prefer to pay $10 in the months when I create more content with the help of my intelligent co-worker, but $0 in the months I take a vacation - basically a Credits-based model. Do Microsoft and Google have the willingness to reorient their business models towards pay-per-use-of-intelligence instead of the current platform subscription model? As of now, I doubt that they have any incentive to do so, until the time their core business starts getting threatened, by which time competition would have sprung up already. Does this mean that BigTech gets disrupted? I doubt it - they will be an eventual follower of the AI-native apps and remain relevant because of distribution.

But a handful of startups will emerge true winners of this AI applications race as they will be the first ones to innovate on the delivery, native use, and business model in the AI age.