Cloud Wars Live with Bob Evans

In this episode, I consider an argument that Palantir recently made: While traditional ERP has been valuable for many businesses over the last few decades, as we move into the AI Economy, is it possible that some of the rigor, discipline, and standards that traditional ERP has brought to organizations and their processes are beginning to cause a problem?

Highlights

00:58 — Palantir challenged traditional ERP in a recent blog post. Although there's a great value it can offer, the company prompted the idea that too much standardization — in the current economy, which is very fluid, and in the age of AI — can become a problem.

01:44 — Standardization leads to everything being uniform. So, how do companies stand out? Palantir noted that uniqueness is not a bug that software should erase but rather a feature that keeps you ahead. Another note it made was that it means sacrificing the organization's identity, its differentiated processes, and the ways it creates value for customers.

02:34 — This isn't a case of Palantir saying that ERP overall is bad; it's the mindset and how the business operates. Now, we have to pick and choose where the approach of standardization applies. This overly standardized traditional ERP approach is going to stifle the ability of companies to create competitive advantage by those unique capabilities, qualities, and personality that traditional ERP, in the view of Palantir, stifles.

03:42 — In its blog post, Palantir suggests a solution to this, that the company can help customers interconnect these new ways of doing things — a modern layer with traditional ERP. Palantir is deepening its partnership with SAP so they can deliver greater capabilities to customers.

Visit Cloud Wars for more.

In today's Cloud Wars Minute, I explore Microsoft's shift to usage-based Copilot Cowork pricing and what it reveals about the changing economics of enterprise AI.

Highlights

00:10 — Microsoft is moving Copilot Cowork from a fixed-price subscription model to usage-based pricing, and this is really reflecting the fact that heavy users are racking up massive compute costs compared to others.

00:55 — More and more, the focus is shifting to how organizations can scale those (AI) capabilities in a way that's financially stable, but beyond that, Microsoft has also said that it's considering a Microsoft-hosted version of DeepSeek as a lower-cost model alternative.

01:16 — Right now, at the moment, Copilot Cowork workloads are powered by models from OpenAI and Anthropic. We should expect to hear from Microsoft regarding DeepSeek, or another low-cost model choice, within the coming weeks.

01:32 — So, what are we really seeing here? Well, Microsoft's AI strategy is evolving beyond simply offering access to the most powerful models. Increasingly, it's about giving customers the right balance of performance, economics, and choice.

01:49 — This is also highlighting, for me, a big divide between how governments and businesses view the AI race. Governments often frame this AI race as a competition between nations, but enterprises are more likely to focus on which models deliver the best outcomes at the lowest cost for their customers.

Visit Cloud Wars for more.

In today's Cloud Wars Minute, I examine what OpenAI's latest move reveals about the maturation of the enterprise AI market.

Highlights

00:03 —My colleague Bob Evans has already covered the specifics of OpenAI's new partner network, so I don't want to spend too much time on the ins and outs of the program itself today. Instead, what I want to do is focus on what this announcement really tells us about the wider state, the broader state of the AI market.

00:25 — Now, for much of the past two years, the conversation around AI has focused on models, questions like: "Which model is best? Which company is ahead? How quickly are capabilities improving? Now, while those questions still matter, they're not the most important ones for many enterprises today. The challenge is more about scalable deployment.

00:44 — Most large organizations have already experimented at this point with AI in some way. They've run pilots, they've tested use cases, and identified areas where AI can really create value for them. The issue now is turning those successes into a scalable business strategy.

01:17 — And that's why OpenAI's partner network matters in this instance. For me, the announcement is less about OpenAI launching another program and more about the company realizing that, although it has the technology, that alone isn't enough for companies to scale in the AI era. They need an ecosystem that includes consultants, partners, and specialists as well.

01:49 — The winners in this next phase will not necessarily be the organizations with access to the most powerful models; they'll be the ones that can successfully embed AI into day-to-day operations and generate real business outcomes. When you look at it like this, OpenAI's partner network is not just a new customer program, it's a sign that the industry is entering a new chapter.

Visit Cloud Wars for more.

In today's Cloud Wars Minute, I discuss the growing challenge of balancing AI demand, budgets, and business priorities across the enterprise.

Highlights

00:02 — We’ve got an emerging dilemma here for CEOs in the early days of the AI Revolution, and a new solution from OpenAI has brought that to light. The dilemma is this: when everybody wants AI tokens, as many as they can get, as quickly as they can get them, but AI tokens are in limited supply, who gets them? Who decides that? How do companies measure that?

00:54 — So, OpenAI has come out with a new tool. This new solution from OpenAI is aimed at AI administrators, so they can sort of turn them on and off, see who’s using what, which ones align with business impact, more over here, less over here. As a tool, that’s all great. But who sets, on high, the policies that those AI admins then can use as their guide?

01:47 — The current state of reality in the marketplace is that the technology is outpacing a lot of corporate cultures. Perhaps there are some CEOs who have sat down with their executive teams and very rigorously hammered this out: a very clear, transparent policy. The current state of the market, though, is one where I don’t see that happening in too many places.

02:49 — The problem sits on top of that solution. Who sets the policies that the AI admins will then follow? You’ve got salespeople over here screaming, “I need more,” product development saying they need more, marketing saying they need more, and every part of the company saying, “I need more.” Who sets the guidelines for how that is determined?

03:30 — The enormous burden is resting on the shoulders of AI administrators who are just not equipped to see across the company and determine where these resources should go. For CEOs, the AI clock is ticking, not only to set a high-level strategy, but also to build the culture that allows companies to be successful in the AI Revolution and AI economy.

04:33 — Great tool from OpenAI, but there’s got to be a lot of education done at the top level for customers. As AI tokens begin to be recognized as incredibly valuable, companies may need entirely new approaches to allocation, governance, and business prioritization.

Visit Cloud Wars for more.

In today's Cloud Wars Minute, I examine three very different approaches to turning breakthrough AI technology into real-world business transformation.

Highlights

00:11 — The AI Deployment Wars involve Google Cloud, OpenAI, and Anthropic. Each has evolved from being almost lab-focused companies with tremendous AI models, and now the pace of innovation of these models is remarkable. The capabilities, power, is stunning, but what that leads to is a huge demand among customers now to not just slip in a piece of technology, but to change companies dramatically.

00:45 — So, that's why the focus here now is on these Deployment Wars. I want to take a look at the three different approaches that these companies are taking. How are they going to take this very cool technology and turn that into customer success at the point of deployment for these customers? There are different approaches here. Money's not the issue here.

01:45 — The challenge is, how do they set themselves up to be not just great creators of technology, but deployment enablers for some of the world's largest companies doing unbelievably complex projects and betting their futures on AI? The technology is enabling the real goals, which are transformation and the ability to move and grow quickly.

02:21 — How are they going to ensure customer success at every level? How are they going to change the processes companies have, the way they do business, the mindset, the technology, the opportunities they have, the strategy, and the cultures of these companies? How are they going to tackle technical challenges that nobody has really ever handled before? This is remarkably different.

03:56 — Google Cloud is going to go almost exclusively with partners. Both Anthropic and OpenAI are saying that they have both funded, along with a lot of partners, deployment companies that will use deployed engineers who are employees of either OpenAI or Anthropic out at the point of the customer to complement or supplement some of what partners are doing.

Visit Cloud Wars for more.