Google’s earnings call identified some major changes and unexpected outcomes, including the performance of AI Ads, the importance of hardware compared to algorithmic efficiency gains, and the groundswell of developer adoption for Gemini models.
AI algorithms have had a thousand times reduction in inference cost over the last three years. There is another cost reduction curve, which is the compute per watt of electricity. Google reveals that this is decreasing at half of the rate of algorithmic efficiency. But it can be just as meaningful.

“Google data centers deliver nearly 4x more computing power per unit of electricity compared to just 5 years ago. These efficiencies, coupled with the scalability, cost and performance we offer are why organizations increasingly choose Google Cloud’s platform. In fact, today, Cloud customers consume more than 8x the compute capacity for training and inferencing compared to 18 months ago”

Google’s interest among developers is growing quickly now, representing 2 of the top six models according to OpenRouter. It’s about a third the size of Anthropic but more than 5x the size of OpenAI.

That’s why more than 4.4 million developers are using our Gemini models today, double the number from just 6 months ago…Vertex usage is up 20x last year.

The new modalities for engaging with computers will be the main theme of AI in 2025. Whether it’s deep research or local process automation, these agentic systems will transform the way that we all work with our computers.

And we have opened up Trusted Tester access to a handful of research prototypes, including Project Mariner, which can understand and reason across information on a browser screen to complete tasks. And Project Astra. We expect to bring features from both to the Gemini app later this year.

Multimodal will be critical. Photographing an object and circling it is the state of the art today, but imagine recording a video and your own voice at the same time as a way of catalyzing a workflow or task.

Those who have tried Circle to Search before now use it to start more than 10% of their searches.

This was the most surprising part for me, that ad experiments within AI perform at similar levels of monetization to current search. This means that the calls for a less profitable Google as Gen AI Search becomes a dominant methodology are premature.

“We recently launched the ads within AI Overviews on mobile in the U.S., which builds on our previous rollout of ads above and below. And as I talked about before, for the AI Overviews, overall, we actually see monetization at approximately the same rate, which I think really gives us a strong base on which we can innovate even more.”

With $75B of CapEx in cloud, the growth of the cloud business will be stunted this year until those systems come online. You can imagine that all the general contractors building those data centers feel an enormous amount of pressure to deliver. This is similar to other public clouds.
“We exited the year with more demand than we had available capacity.”
Amazon said something similar :

“And though we expect growth will be lumpy over the next few years as enterprise adoption cycles, capacity considerations and technology advancements impact timing, it’s hard to overstate how optimistic we are about what lies ahead for AWS’ customers and business.”

The implications are clear: 2025 will be a pivotal year for AI, marked by a renewed focus on hardware, the rise of powerful AI agents, and the surprising monetization potential of AI-driven experiences. These shifts will change how we use technology.
—————
Boost Internet Speed–
Free Business Hosting–
Free Email Account–
Dropcatch–
Free Secure Email–
Secure Email–
Cheap VOIP Calls–
Free Hosting–
Boost Inflight Wifi–
Premium Domains–
Free Domains