Google is probably going to be fine. The web is going to get a bit weird in a few years and they're going to be right in there.

The web is likely going to layer up soon. As more and more people utilize AI it changes traffic. It's impossible (and undesirable) to keep training with current affairs, LLMS need to go out to the web to find reality and they need to do search in order to get there.

More and more they're going to get blocked from doing so via "human" channels. I predict we're going to get a sub-web. basically MCP but for websites, AI interacts with a sites content through that door, humans use http.

When there's a gold rush sell pickaxes. Forget Gemini, that's a money pit. Google is in a great position to shape that sub-web through chrome and it's current incumbent position as gatekeeper to the web. They'll be selling search deals to all the Big AI agents, indexing the new MCP-like layer so AI can search and interact with site information and blurring the line between your harness and chrome such that browsing the web is more like asking AI for crap and getting pages there in the harness rather than traditional browsing.

before you know it nobody will browse the web the way we do now and no agent harness will function correctly without Google.

I’ve never encountered in my life a new technology that was adopted this quickly. In just a few years AI is being used in some way by almost everyone, no matter if they’re tech-savvy or not. It’s also interesting that everyone uses it but it’s also pretty widely hated at the same time.

The demand is there - I just don’t think the revenue streams have been fully figured out yet. Ultimately I think the AI providers will converge on allowing their AI to respond to queries with sponsored suggestions, quite simply because there’s too much money being left on the table to not do that.

Google is making these massive investments because they’re terrified that they’re going to be eclipsed or turn into IBM. But Google has way more of a buffer to play the long game than OpenAI or Anthropic.

It means the spending spree is on hold. The credit window for the neoclouds remains closed. Now the hyperscalers are feeling the pain. The narrative was that the hyperscalers could lean on their prodigious cash flows if the data center rush didn't pan out. This round of earnings reports adds some doubt to that narrative which is spooking everyone. Hyperscalers are still projecting a further increase in capex in 2027, but one has to wonder how long that will last if conditions stay the same.

The next step is whether data center capex discipline begins filtering to Nvidia and the DRAM semis revenues. At that point pure play companies must drive revenue growth or else.

There’s too much actual value for a flaming crash.

i'm disappointed that Mythos is an Anthropic model because it would be the perfect naming for Google's models since they only actually exist in benchmarks or press releases