Google IO / 2023

Laurent tagged me and asked me about my take of this years Google IO.
I do not have too much time right now and most probably do not manage to share my thoughts otherwise. So here are my very rough, unstructured, raw, probably partly wrong and unfinished thoughts. With lots of wild assumptions and hot takes that will probably be outdated right after my next lunch :).

Google lost their edge and open source is eating their lunch. Two things popped into my mind immediately:

They might be still leading in the niches and very specialized applications (AlphaFold) but that’s not where Google’s cash flow comes from. But this area may be the area where it could do the most good for everyone.

They are trying hard to figure out what to do and are very much in a “we don’t know yet phase, but this could go sideways VERY fast”:

  • Google labs: They try to tinker around (in the open) and find models / applications they can dominate and build on (and are hard to copy)

For me that’s also the reason why the keynote was all over the place. No one knows what comes next, it’s a tectonic shift and turns everything upside down at the moment, and their core business (that brings in the money) is being attacked on so many levels that they need to figure stuff out fast. In a way that is sustainable and not easy to copy.

  • Google is sueable and the perfect candidate for it.

  • Open ai is outsourcing many of the legal implications to the api users / product builders

  • If google adopts it in all of their products (esp. search) they might be also responsible for the content and therefore lawyers are looking closely.

  • Bing search has nothing to lose, google might lose everything. So bing can use open Ai in search without too many worries.

  • They use adobe for image gen for examples (ensured not trained on copyrighted material) because they might fear lawsuits when using their image search / user data.

  • I don’t think they want to be the forerunner when it comes to the discussions about rights and laws and copyright. They only can lose on this front, will be tied up in discussions while others will just pass by in the meantime not being examined and judged by the discussions.

  • Google search might be flooded with AI generated content which is currently based on hallucinating models but perfect for SEO.

  • They might face a massive quality problem / or everything looks and ready the same

  • If everything looks the same, I don’t need to search for alternatives

Maps might be currently their most valuable product being based in physical space. Search for proximity (Best doctor in town close to my place / best fit for me) on a human level that is not generative. Counter movement to a more and more confusing and diffuse and flooded online space is going back in the real world and meeting human beings. Places to hang out ( Even though the “immersive view” didn’t made me excited at all and was irrelevant for me For me this was a “uh that’s a neat integration and useful” moment
Doing stuff in second and third order with search is much more powerful than search itself. Maybe I would call it: Transformative interaction on and with (search) data. This might be the power google can provide better than others because of the amount of products and integrations they have, i don’t know.

Google photos and the editing is impressive on a technical scale but troublesome on a human level. It’s like tampering with your own memories. Like a tiktok beauty filter. Altering the reality to a non existing standard - everyday is a perfect day with sunshine? Not sure what that will do to the human psyche - for me it felt somewhat unsettling thinking about it. Looking back at old photos will be like showing a very distorted view that is somewhat conflicting with your own memories. Scary thought.

I’m not sure, maybe I’m a cynic, but I don’t believe the “build and deploy responsible” claim. As long as the business model is ad based, from my personal view, it conflicts with those claimed core values. Even though Google might not even be the worst actor in this area, ads in general are built to exploit your weaknesses and sell you stuff, mostly for profit. Not for best fit, but for the advertisers bid.

  • Watermarks and metadata is not the solution for the problem - and it’s weak fig leaf solution to show “hey look we do something and if people are not using it, it’s not our fault”
  • Showing context after a couple of clicks is neat but I feel it’s far away from the reality of how people use and consume that stuff.
  • We need stuff like this, yes. But it’s not the answer to the problem we face. And it’s more a deflection from a problem hard to solve - if solvable in a proper way at all.

I didn’t care about android at all. It was just being on par with iOS in many places (and it also goes the other way around). It’s a slow crawling battle to impress consumers with shiny things (wallpaper) but without real deep innovation and progress. Mobile phone innovation flattened massively over the last years and is just improving the status quo with better stats, battery life and cameras. We all have an iphone, a watch, a ring, a tablet. But it’s less and less important. There is no real relevant difference between an iphone 12 or 13 anymore. Or are we already at 14? I don’t know, and I don’t care anymore.

The tablet / hub mode is a small neat way to give the tablet a place to live in your home. But not a reason to switch ecosystems. I think we will see Apple copy this within a year.