Larry Ellison's AI World Keynote - First Impressions from Industry Analysts and Thought Leaders

Last week, before Oracle's AI World event kicked off, I remember thinking that the company was on a serious roll coming into the conference.  They recently saw their stock price go up rougly 35% overnight due to a great quarterly earnings call - unheard of for a company that went public almost 40 years ago.  They're involved in a deal that would see TikTok sell its US assets to a group of investors, companies, and wealthy individuals.  And they are a significant player in the government's $500 billion Stargate project aimed at keeping the US in the catbird seat when it comes to AI innovation.

All this and more (way more, actually) has Oracle on a real heater.   It has them arguably being the most interesting tech company in the enterprise app space at the moment...whhich also is saying a lot.  And with all that going on leading into AI World, it felt like Larry Ellison's keynote had an even bigger spotlight shining on it. Making it, in my estimation, his most important one of these in years, given how every vendor is positioning themselves as THE agentic AI leader in the enterprise.

So with so much riding on this keynote, The Venerable Paul Greenberg and I held a special CRM Playaz watch party to check out what Larry had to say at this particular moment in time. We were also joined by two of our go-to industry analysts/thought leaders - diginomica co-founder Jon Reed, and industry veteran and leading expert Brian Sommer - to get their thoughts in real time on what they saw and heard.

Below is a clip of some of our initial thoughts directly following after the keynote ended, followed by summaries of each of our key takeaways as well as an edited transcript of the post keynote discussion.

Key Points from Paul Greenberg

• Stunning Ambition and Differentiation: He was initially concerned about Ellison's earliest AI conversations but now believes Oracle is "setting themselves apart from any company in the entire industry".
• Focus on Social Problems: Ellison is "literally working on solving with oracle technology and AI... some of the world's largest social problems" by building the necessary infrastructure and "partnering with the companies they need to partner with".
• Ecosystem of Ecosystems: Greenberg views Oracle's approach as an "ecosystem of ecosystems" and doesn't know a "company better situated" to execute it.
• Execution Over Talk: He is close to being "in awe" of "how much they've actually done as opposed to how much they're talking about" and was impressed by Ellison's "grasp and depth of that concern".
• The Cerner Acquisition and Legacy: Greenberg was "so impressed with the way he's been doing this," noting that it started for him when Ellison acquired Cerner. He realized that for Ellison, this is a deeply personal decision and "not a business fli- thing," tied to his personal sense of legacy.
• Non-Operational Role: He agrees with the idea that by stepping away from the CEO role and the operational "stuff," it allows "the big brains" like Ellison to focus on their capacity to do the big-picture context work and then "bring it back to the company".

Key Points from Jon Reed

• Keynote Quality and Disagreement: He considers it one of the most essential keynotes he's attended in a long time. However, he "profoundly disagreed with some things Larry said" and thought "some of the AI stuff was hyperbolic". Ultimately, he values that it "provoked the right conversation", which he considers rare.
• Highlights: The most significant parts of the keynote were the segments on "healthcare and the energy stuff" and "hammering home this issue of private data".
• Context Intelligence and Differentiation: Reed emphasizes that the issue Ellison is addressing is called "context intelligence", which is a "bigger deal than reasoning". This concept suggests that a large language model performs better when provided with the right context for a specific situation. Oracle is trying to provide this at a different level than other vendors because they control the entire infrastructure around the AI processes.
• The Full Stack Approach: He noted that instead of just building applications, Oracle is doing "the whole friggin' stack".
• Legacy and Profound Time: Reed concluded that Ellison's actions stem from a "combination of legacy" and "sensing a profound time in sort of human culture and in business both", where executives want to help their company and "make a difference". He compared Ellison's move to step back from an executive leadership role to similar shifts made by Aneel Bhusri at Workday and Sridhar Vembu at Zoho. He noted that a handful of significant enterprise software figures are embracing this period with a "legacy mindset".
• Customer Resonance: He stressed that while Ellison "set the tone for all of those audacious things," the next step is finding out what "resonates with customers".

Key Points from Brian Sommer

• Authenticity and Personal Narrative: Sommer was fascinated by the extent to which the talk was a "first-person narrative". Ellison did the "work", sharing his personal experiences with his health and what he saw, like the robots he questioned. This genuine effort is a major differentiator, as most CEOs are simply "fed them talking points"⁵. Whether one agreed with everything he said or not is "almost immaterial" because "he did the work".
• Vision and Big Problems: Ellison demonstrated the ability to "craft a compelling vision for the future", which Sommer considers the mark of the best top executives. While 99% of CEOs "can't see past the next quarter's AI... functional enhancements," Ellison was focused on "big problems to solve, not little tiny baby problems". This vision "does set oracle apart from competitors".
• The "Masterclass" and Inspiration: Sommer suggested the talk was a "masterclass" and that more software companies should watch a replay so they can "be prepared to talk about and inspire people in their audiences". He criticized companies that "resort to having showgirls and marching bands because you lack vision".
• The Best Executive Role: Sommer believes that the role of the best tech executives, like Larry, should be "permanently different in high tech". Ellison is at his best "playing in traffic, not necessarily with real customers", but by seeing the problems in hospitals firsthand, "get out there and feel it, touch it and embrace whatever the pain and the issues are". This allows him to internalize issues and then figure out where to take the technology. Sommer noted that this is not what CEOs generally do, as they spend time with the board and Wall Street. He suggested that every tech company needs someone like Larry, but "I’m not sure they have to be the CEO".

Key Points from Brent Leary

• Comparison to Jensen Huang: The only other keynote Leary could equate to Ellison's was the one given by Jensen Huang earlier this year. He sees Ellison as doing "the enterprise version of the big picture", similar to what Huang presented.
• Focus on Physical and Real-World AI: Ellison's talk focused on "bigger things" like physical AI and real-world, not digital-world, challenges, rather than spending a lot of time on "agentic" AI, which many others have been talking about all year. He finds that the repetitive talk about agentic AI "feels smaller and and much more insignificant" when compared to how Oracle is tackling these bigger challenges.
• Ellison's Dedication and Grasp: Leary was deeply impressed that Ellison sat in a chair for an hour and a half and went through "some serious stuff at length and at detail and had the full grasp of it" without a hiccup. He noted that as the second richest man in the world with his own island, he wouldn't even be thinking about this stuff, but Ellison is "not only thinking about it" but clearly spends a lot of "deep focus" on connecting the dots between big challenges and the potential impact of AI.
• Setting Oracle Apart: This dedication and focus on big challenges over repetitive agentic AI talk helps Ellison "set himself and oracle apart from the others".
• Masterclass and Follow-up: Leary ultimately called the talk a "masterclass" and stated that he hasn't seen anything like it in the enterprise world in a long time, though he acknowledged that there are things that will need to be followed up on.
• Shift in Role (CTO/Chairman): Leary speculated that Ellison's shift from CEO to CTO and Chairman, handing off operational stuff to a great executive team, is what allowed him to focus on these big-picture items. He compared this situation to what Sridhar Vembu was doing with Zoho.

Edited Transcript

Brent Leary: First impressions?

Paul Greenberg: It was stunning, really. I mean, we all, I think, probably have some, let's say, concerns and disagreements with his earliest conversation around AI. I actually think that oracle is setting themselves apart from any company in the entire industry. He's literally working on solving with oracle technology and AI and so on, some of the world's largest social problems and they're building infrastructure that will do them, and they're partnering with the companies they need to partner with to do them.

Again, I'll go back to my continuous refrain of ecosystem of ecosystems because that is pretty much what I'm seeing here. And I don't know a company better situated. I don't wanna say I'm in awe, I'm not quite there, but I'm close to being in awe of how much they've actually done as opposed to how much they're talking about and of course, his grasp and depth of that concern.

Jon Reed: I think I profoundly disagreed with some things Larry said. Some of that I articulated, some I didn't. But look, I mean, this is one of the must-hear keynotes that I’ve attended in a long time.

I think Larry delivered today something where you didn't know what was coming next, uh trying to grapple with real world problems. I do think I will judge Oracle also on how they progress on those things. But like I said, a very meaty keynote. I thought some of the AI stuff was hyperbolic. But the highlights were the healthcare and the energy stuff and also, really hammering home this issue of private data and getting into some of how that works.

I know Paul had some lingering questions about all of that, but I think Larry took a good crack at the differentiation around that. 'Cause that's actually a bigger deal than reasoning. Reasoning is just a technique that sometimes helps and sometimes hurts and can be expensive and blah, blah, blah. But the issue that Larry's getting at is one called context intelligence.

The idea is that a large language model can deliver much better to your needs if you provide it with the right context for your situation, and that's what Oracle is trying to provide at a different level than other vendors because of all the infrastructure that they control around the ai processes themselves. Instead of just building applications, They're doing the whole friggin' stack.

And so Larry kinda set the tone for all of those audacious things and now we have to find out what resonates with customers. But for me, a keynote weary guy like myself, I'll sit around for something like this and I'll listen, and who cares if i didn't agree with it all. At least he provoked the right conversation and that's so rare.

Brian Sommer: I did think it was probably one of the meatiest and toothiest talks I’ve heard all year. What's really fascinating to me is how much of this is a first-person narrative that he's talking about, his experience with his health stuff and what he personally saw and, the robots that he saw and asked questions about. It's not that he's taking you on this journey with him as much as he did the work, which is more than I can say for most CEOs who somebody fed them talking points and they come out on stage and pound them to death in front of us.

Whether you agree with everything he said or not is almost immaterial because he did the work. And I think that's something I wish more of the top executives of software companies would do.

I would say this does set oracle apart from competitors. I know I've been harping on this point for better part a year, two years, is the best CEOs—and I know he's not that anymore, he's the CTO and chairman of the board—but the best ones are the ones who can craft a compelling vision for the future. And I'm sorry, 99% of the CEOs out there can't see past the next quarter's AI, you know, functional enhancements they're sprinkling a little bit around the product line. He was thinking big problems to solve, not little tiny baby problems.

I don't wanna use the word masterclass, but damn it, I’d say we ought to make more software companies watch a replay of this so they can be prepared to talk about and inspire people in their audiences. And if you have to resort to having showgirls and marching bands because you lack vision, we see that as analysts and we will call you on it.

Jon Reed:  Go get 'em, cowboy! You got 'em this time.

Brent Leary: The only keynote this year that i can kind of equate to what this was is Jensen's keynote earlier this year. where everybody has been talking about agentic and he didn't really spend a whole lot of time on that stuff. He was he was looking at these bigger things as well with physical AI, as opposed to the agentic stuff and thinking about those in kind of the real world, not the digital world realm.

Jensen has his own style and flair and Larry definitely had this more relaxed approach. I couldn't help but think about Larry's kinda doing the enterprise version of the big picture like Jensen did in his conversation.

And I agree with you guys, I'm sure there's certain things you could nitpick if you really wanted to. But the big picture to me is this guy just sat there in that chair for an hour and a half and without a hiccup, went through some serious stuff at length and at detail and had the full grasp of it. And I'm thinking to myself, "If I was the second richest man in the world and i had my own island, not even thinking about this stuff." And he is not only thinking about it, but you could tell he spends a lot of deep focus on these things and connecting the dots, to  the big challenges and how the ai can and potentially will impact it. And just by doing that kinda thing, I think he sets himself and oracle apart from the others that repetitiously all year long talk about agentic AI, and what it will do.

That stuff really feels smaller and and much more insignificant when you hear how oracle is kinda tackling it. Yeah, it was a masterclass. I would say that there are things there that'll need to be followed up on, but I haven't seen anything like this in the enterprise world in a long time.

Paul Greenberg: I have been so impressed with the way he's been doing this and again, for me, it started when he acquired cerner. even then, I wondered why he did that. And then I started realizing he was probably what, 78 then?

Brent Leary: I think it might have when he decided not to be the ceo and was like, "I've got great people that I can hand over all that operational stuff to." And I think when he went back to being the chief technology officer in addition to the chair, I think it was it almost sounds like what Sridhar Vembu was doing with Zoho. You got such a great executive team that you can feel comfortable about turning over all that operational stuff and that allows these the big brains, the folks that really have this capacity to do all that in context and then bring it back to the company.

Paul Greenberg: But I think the Cerner thing was a huge part of it, probably started a little bit before that with that co-ceo legacy starts, you know, It kind of starts when it starts with you, whenever it is. Um you tend to be older, but it starts and with him, I think the uh one thing about legacy, they're personal decisions. And th- this is not a business fli- thing for Larry. This is he's deeply personal for him.

Jon Reed: I think it's the combination of legacy but also sensing a profound time in sort of human culture and in business both and wanting to help your company see that through but also make a difference in that. To your point around Sridhar and Larry, same thing happened with Workday with Aneel Bhusri. Another significant enterprise software figure all realizing like, "to take this on in the way that I need to do it, I can't be the the lea- the executive leading the charge." Like and and I think it just shows you and I'm not as optimistic about AI in a global sense as larry perhaps, but I would still agree that it's profound that these executives a handful of them are really embracing that in a legacy mindset like you said, Paul.

Brian Sommer: Listening to you guys describe the ceo role is making me think maybe it needs to be permanently different in high tech.

Larry’s at his best playing in traffic, not necessarily with real customers, but seeing the problems in the hospitals and, you know, seeing this stuff firsthand. he's a get out there and feel it, touch it and embrace whatever the pain and the issues are and internalize that and then he can figure out where to take the technology beyond that.

That's not what you generally have a CEO do. And CEOs are spending all kinds of time with the board, with wall street and a whole bunch of other things that I'm sure Larry would be happy if he never had to deal with that kind of stuff ever again. It's why he had killer people like Safra and Chuck Phillips. They dealt with a lot of those kind of issues. Then you had Mark Hurd who dealt with a lot of the sales and revenue kind of stuff and we could go down the list. They've had other executives handle some of these things that might be ceo responsibilities.

But the best thing you could do is every one of these tech companies need somebody like Larry, but I’m not sure they have to be the CEO. But you gotta give them enough, you know, turn them loose in the front 40 and let 'em run.