Microsoft, OpenAPI Tighten and Loosen Their Relationship

Microsoft's relationship with OpenAI agreement removes some restrictive elements and arguably strengthens the alliance. 

The agreement locks in Microsoft's priority access to OpenAI's technology and massive cloud spending, while granting OpenAI the flexibility and corporate structure it needs to accelerate its growth than it perhaps used to be.  

Microsoft's intellectual property rights to OpenAI's models and products, including those developed post-Artificial General Intelligence, are extended through 2032. 

OpenAI will purchase an incremental $250 billion in Microsoft Azure cloud computing services, confirming Azure as the primary platform for OpenAI's massive compute needs.

But OpenAI gains the ability to work with other partners to jointly develop some products (though API products remain exclusive to Azure), as well as use other cloud providers for non-API products. 

According to the latest agreement between OpenAI and Microsoft, Microsoft holds an investment in OpenAI Group PBC valued at approximately $135 billion, representing roughly 27 percent of the total. 

But the new agreement, while preserving some early features, also allows each firm more flexibility to pursue other relationships. 

OpenAI remains Microsoft’s frontier model partner and Microsoft continues to have exclusive IP rights and Azure API exclusivity until Artificial General Intelligence (AGI).

OpenAI can now jointly develop some products with third parties. 

API products developed with third parties will be exclusive to Azure. Non-API products may be served on any cloud provider.

Microsoft also can now independently pursue AGI alone or in partnership with third parties.

Search and AI Chabot Lines Blurring?

Without question, AI for e-commerce is displacing some amount of shopper activity. For example, AI chatbots seem to be displacing some other sources, most notably perhaps, search. But there are some nuances. 

Anecdotally, I find Google’s AI Overviews to be useful much of the time, so for casual online shopping research, Google search continues to function suitably, without the need to go to an AI engine. 

And that might not be unusual, though many younger users probably prefer using chatbots rather than search for e-commerce-related queries. 

Some 58 percent of consumers prefer to use AI tools instead of traditional search engines in 2025, up from 25 percent in 2023, according to research by Capital One Shopping.

Still, when AI overviews are available, they seem to get used, suggesting that the difference between search and AI chatbots is narrowing and perhaps even merging. 

Perhaps 60 percent of searches now terminate without users clicking through to another website, according to Bain & Company, when AI summaries are present. Shopping queries on ChatGPT doubled in six months from January to June 2025, growing from 7.8 percent  to 9.8 percent of all searches.

The distinction between "AI chatbots" and "search with AI Overviews" is important. Most consumers still use search engines, but those search engines now incorporate AI features, blurring the line between “search” and “AI chatbot” use. 

The data shows that while standalone AI chatbots remain a small fraction of total search activity, AI-generated content within traditional search engines has become dominant in how consumers receive information.

source: McKinsey 

How Small Routines Lead to Big Outcomes

And it all starts with small things like making your bed. 

Anthropic Expects 1 GW of Compute Power Online in 2026

Anthropic has announced a multibillion-dollar partnership with Google giving Anthropic access to one million Google Tensor Processing Units and more than one gigawatt of compute power by 2026.

It might not be unrealistic to expect the other leading model suppliers to keep pace, as the evolution of model generations has seemed to require an order of magnitude increase in power capability with each succeeding generation of models.

Facility	Organization	Power Capacity	GPU Count
xAI Colossus	xAI	~100-150 MW	200,000 H100s
Meta AI Research Cluster	Meta	~50-100 MW	100,000+ GPUs
Microsoft Azure AI	Microsoft	~200-300 MW	Distributed
Google TPU Clusters	Google	~150-250 MW	Millions of TPUs

AWS CEO Projections for Future Training
Generation	Power Requirement Per Model	Timeline
Current (Gen2-3)	50-200 MW	2023-2025
Next Gen (Gen3-4)	1-5 GW	2026-2028
Gen5+	5-10+ GW	2029+

Planned/Under Construction Mega Data Centers
Project	Organization	Total Power Capacity	Investment	Timeline
Stargate (Total)	OpenAI/Oracle/SoftBank	10 GW	$500B	2025-2029
Stargate Abilene	OpenAI/Oracle	1-2 GW	$10B+	2025-2026
Stargate Phase 2	OpenAI/Oracle	4.5 GW	$300B+	2026-2028
Stargate Phase 3	OpenAI/SoftBank	1.5 GW	$50B+	2026-2027
UAE Stargate	G42/OpenAI	5 GW	TBD	2026+
Meta Louisiana	Meta	2-4 GW	$10B+	2026-2028
Microsoft Wisconsin	Microsoft	3-5 GW	$15B+	2026-2029
CoreWeave Pennsylvania	CoreWeave	0.5-1 GW	$6B	2025-2027

Compute Capability of Current LLM Models
Model	Organization	Release Date	Training Compute (FLOPs)	Estimated Cost
GPT-4	OpenAI	March 2023	~2×10²⁵	~$100M
Gemini Ultra	Google	Dec 2023	~10²⁵	~$100M
Claude 4 (Opus/Sonnet)	Anthropic	May 2025	~10²⁵	~$100M
GPT-4o	OpenAI	2024	~10²⁵	~$100M+
GPT-4.5	OpenAI	2025	~10²⁵ - 10²⁶	~$100M+
Gemini 2.5 Pro	Google	March 2025	~10²⁵ - 10²⁶	~$100M+
Grok 3	xAI	Feb 2025	10²⁶	~$500M-$1B
Grok 4	xAI	July 2025	~10²⁶ - 10²⁷	~$1B
Llama 3	Meta	2024	~10²⁵	~$100M

At least up to this point, each generation requires approximately 10 times more compute than the previous generation to achieve significant capability improvements.

Expected Compute for Coming Models
Model	Organization	Expected Release	Training Compute (FLOPs)	Estimated Cost
GPT-5	OpenAI	Late 2025-2026	10²⁶ - 10²⁷	$1B+
Claude Next	Anthropic	2026	~10²⁵ - 10²⁶	$1B+
Gemini 3.0	Google	2026	10²⁶ - 10²⁷	$1B+
Llama 4	Meta	2025-2026	10²⁶	$500M-$1B