Cloud Wars Live with Bob Evans

In today's Cloud Wars AI Minute, I explain how Microsoft's AI agent momentum could transform business applications at scale.

Highlights

00:00 — So we're now hearing, according to Microsoft, that it has 40 million agents deployed now in the Microsoft ecosystem. Now compare that to the 30 million seats that it actually has sold for the M365 Copilot, and that really starts to show us that AI agent creation and things of this nature are starting to take off when we're starting to see more than one-to-one ratios here.

00:33 — So this is a really big situation for Microsoft. It's actually a really good signal that custom agents and agents that are being built are starting to get more value off the backside. Now we can't necessarily completely get to this level because what you'll see is a lot of these agents are probably really simple right now.

00:54 — But with some of the new applications coming out and the focus that Microsoft also came out and announced — that they are looking at going through and producing a Super App Copilot — and this was something that Satya Nadella just talked about in the news just last week.

01:13 — So when we think about this and we start thinking about how is it that Microsoft's going to start combining all of its different Copilot assets or its AI agent assets, it really opens up a door where we might start seeing a much bigger multiplier effect happening on the number of agents created.

01:32 — Due to the fact that they're going to make it not so confusing to be able to determine where to get started and be able to help you get this all down. So this is really exciting news, and it's really a good thing for why you should start looking and understanding these agents because they're really starting to take off as business applications themselves.

Visit Cloud Wars for more.

In today’s Cloud Wars Minute, I break down Microsoft’s push to protect open-weight AI models and why the debate over openness could shape the future of the AI ecosystem.

Highlights

00:09 — Microsoft is among more than 20 tech companies, including NVIDIA, Meta, and Palantir, that have issued an open letter urging U.S. lawmakers to slow down in placing restrictions on open-weight AI models. These are models that users can modify and run on their own infrastructure.

00:38 — The backlash from the companies comes in light of discussions within the U.S. government about potentially restricting access to Chinese open-weight AI models in the U.S. The administration's floated the idea of potential restrictions or sanctions involving Chinese open-weight models around concerns about copying other AI models, intellectual property theft, and national security.

01:06 — The tech companies that signed the letter argue that open-weight models strengthen competition, and with technology benefits broadly shared instead of being concentrated into a few hands, and that, in fact, relying entirely on closed models is not inherently safe because they can be breached, misused, or fail in ways that outsiders can't detect.

01:30 — They say that keeping AI advancements in a small circle of companies just compounds that risk. Sharing the letter on his personal X account, Microsoft CEO Satya Nadella said, "Open-weight models are essential to a healthy AI ecosystem. Together with others across our industry, we are outlining a path for open-weight models to strengthen American competitiveness and expand economic opportunity while protecting national security."

02:14 — This is yet another example of how the democratization of AI is being promoted by tech companies. In my opinion, the idea of collaboration and openness has made it much easier for the public to trust AI. Not only does this strengthen the position of the companies themselves, but it also means that more people will benefit safely and securely from the potential of AI technologies.

Visit Cloud Wars for more.

In today's Cloud Wars Minute, I look at why Palantir's AI strategy is resonating with customers and disrupting the software industry's established order.

Highlights

00:03 — Well, we had a fascinating earnings report and earnings call with Palantir recently. A couple of things happened. Its growth rate soared. It almost doubled its growth in Q2, up 93% to over $1.8 billion.

00:19 — But intriguingly, the story behind this growth now is one that CEO Alex Karp is doubling down on, and that is to cut directly against the grain of how traditionally the software industry has worked. He feels that there's too much power being consolidated among the tech vendors, particularly now the AI labs and the AI model companies.

00:55 — Now, is there a direct correlation between the 93% growth rate and this refreshing, new, different customer-oriented approach from Palantir? I think there is. Whether you think Karp's antagonism toward the entrenched order is overblown or not, you've got to look at the financial results. Palantir is outperforming every major tech vendor on Earth.

02:36 — What Karp is promising, he said, what Palantir is offering, is a new, better way that allows customers to keep control of that data, keep the benefits of that data, keep what he calls the alpha, and let them control their own outcomes instead of turning over their data to model companies.

04:36 — I just find myself throughout the call listening and watching Karp, who always wears a white T-shirt. He is just completely unabashed in not holding back and stating his opinions. I think we need more of that in this business, especially at these very critical times when businesses are betting their futures on what's going to happen here.

Visit Cloud Wars for more.

In today's Cloud Wars Minute, I break down OpenAI's decision to make GPT-5.6 the preferred model for Microsoft 365 Copilot and why AI efficiency is becoming an increasingly important competitive advantage.

Highlights

00:09 — It's fair to say that OpenAI has had a changeable relationship with Microsoft over the past few years. However, ties between the two companies have always remained strong. To that end, OpenAI has announced that GPT-5.6 has become the preferred model in Microsoft 365 Copilot, which includes Word, Excel, PowerPoint, Copilot Chat, Teams.

00:33 — Now, for context, this is OpenAI's latest flagship model, which the company claims delivers higher quality outputs and stronger performance per dollar spent, and this efficiency is increasingly important as more work shifts to AI models.

00:58 — For Excel, there'll be deeper data analysis and faster insights in PowerPoint. Copilot will be able to create more polished presentations with less manual effort. Meanwhile, Copilot Chat benefits from GPT-5.6's improved reasoning and productivity capabilities for everyday tasks.

01:28 — For me, though, the biggest benefit is this idea of improved efficiency — the notion of getting the most out of every token and gaining more value from the tools you're using. I think this will become a new battleground moving forward, not just for model developers who balance power and budget friendliness, but also for providers of AI tools like Microsoft, who need to show their customers that they're making the best choices in how these products are powered.

Visit Cloud Wars for more.

In today's Cloud Wars Minute, I break down the reshuffled hyperscaler rankings and what AWS's comeback means for the AI economy.

Highlights

00:03 — The third hyperscaler reported its numbers late last week. That was Amazon's AWS unit. Fantastic quarter for AWS. Best in four or five years. It had what I called an awesome Q2, and in doing so, we've seen a complete reshuffling of the growth standings among the hyperscalers.

01:18 — Microsoft, for the same period, its fiscal Q4 grew 27% to $59.3 billion, and Google Cloud was up an incredible 82% to almost $25 billion. We see Google Cloud and now AWS on a very nice continuous upswing, while Microsoft has leveled off.

03:05— Inside AWS, the backlog soared, up 154% to almost $500 billion. Andy Jassy said Amazon will spend in 2026 $220 billion on CapEx, almost all of which will go toward building the data centers necessary to fund this expansion in cloud and AI.

04:19 — The smallest backlog number of the four hyperscalers is AWS at $496 billion. Microsoft is around $675 billion, Google Cloud is $514 billion, and Oracle is about $636 billion. Together they represent well over $2 trillion in backlog.

05:12 — I think it's great seeing AWS back in high-growth mode. It makes the competition better, and it certainly is a great thing for customers, who now have choices of fantastic suppliers across the four hyperscalers: AWS, Google Cloud, Microsoft, and Oracle.

Visit Cloud Wars for more.