Join Gemma Allen, host of theCUBE, and Raphaëlle d'Ornano, founder and Chief Executive Officer of Decoding Discontinuity, as they navigate the ever-evolving landscape of artificial intelligence and its impact on Fortune 500 companies in episode two of the AGNT Podcast. This discussion, set against the backdrop of rapid technological advancements and market volatility, provides a deep dive into the dynamics shaping the next era of enterprise software.
In this engaging episode, Allen and d'Ornano explore the concept of agentic AI and its emergence with tools such as Moltbook, signifying a paradigm shift in the enterprise realm. They discuss the implications of these developments on public markets. d'Ornano provides expert analysis on the potential disruptions and transformative capabilities agentic AI introduces. Featuring insights from theCUBE Research, this conversation offers a comprehensive look at how AI is redefining enterprise operations.
According to d'Ornano, the Moltbook phenomenon epitomizes the current discourse on AI—a transition toward an economy built on inference-driven computing. Key takeaways include the need for enterprises to adapt to new AI-driven models and the challenges surrounding return on investment and security. Analysts also highlight contrasting perspectives on AI's role in increasing volatility within tech markets, showcasing both opportunities and risks for businesses and investors.
Forgot Password
Almost there!
We just sent you a verification email. Please verify your account to gain access to
theCUBE + NYSE Wired: AI Agent Conference. If you don’t think you received an email check your
spam folder.
Sign in to theCUBE + NYSE Wired: AI Agent Conference.
In order to sign in, enter the email address you used to registered for the event. Once completed, you will receive an email with a verification link. Open this link to automatically sign into the site.
Register For theCUBE + NYSE Wired: AI Agent Conference
Please fill out the information below. You will recieve an email with a verification link confirming your registration. Click the link to automatically sign into the site.
You’re almost there!
We just sent you a verification email. Please click the verification button in the email. Once your email address is verified, you will have full access to all event content for theCUBE + NYSE Wired: AI Agent Conference.
I want my badge and interests to be visible to all attendees.
Checking this box will display your presense on the attendees list, view your profile and allow other attendees to contact you via 1-1 chat. Read the Privacy Policy. At any time, you can choose to disable this preference.
Select your Interests!
add
Upload your photo
Uploading..
OR
Connect via Twitter
Connect via Linkedin
EDIT PASSWORD
Share
Forgot Password
Almost there!
We just sent you a verification email. Please verify your account to gain access to
theCUBE + NYSE Wired: AI Agent Conference. If you don’t think you received an email check your
spam folder.
Sign in to theCUBE + NYSE Wired: AI Agent Conference.
In order to sign in, enter the email address you used to registered for the event. Once completed, you will receive an email with a verification link. Open this link to automatically sign into the site.
Sign in to gain access to theCUBE + NYSE Wired: AI Agent Conference
Please sign in with LinkedIn to continue to theCUBE + NYSE Wired: AI Agent Conference. Signing in with LinkedIn ensures a professional environment.
Are you sure you want to remove access rights for this user?
Details
Manage Access
email address
Community Invitation
AGNT Podcast Ep. 2 with Gemma Allen & Raphaëlle d'Ornano
Exploring the Advancements and Challenges in AI Agent Deployment
John Nay, founder and Chief Executive Officer of Norm Ai, joins theCUBE's special presentation with NYSE Wired, focusing on the upcoming Artificial Intelligence Agent Conference 2025. Hosted by John Furrier, co-founder and co-Chief Executive Officer of SiliconANGLE Media, this insightful discussion covers the pivotal developments in AI infrastructure and the regulatory complexities faced by enterprises.
In this episode, Nay shares their expertise in regulatory AI infrastructure, particularly as it pertains to AI agent deployment in highly regulated sectors. The conversation, hosted by Furrier, delves into the evolving landscape of AI technology, compliance challenges, and the strategic initiatives underway at Norm Ai to address the pressing issues surrounding AI deployment. The discussion provides valuable insights for both technology and policy influencers.
Key takeaways from the discussion include the emphasis on the need for dynamic, real-time compliance frameworks that align with regulatory standards, as emphasized by Nay. Furthermore, the episode highlights how enterprises can leverage existing compliance structures to integrate AI technologies more effectively, offering a glimpse into the future of AI agent scalability and regulation. The conversation underscores the importance of bridging the gap between engineering, policy, and technology for sustainable AI innovation.
AGNT Podcast Ep. 2 with Gemma Allen & Raphaëlle d'Ornano
Gemma Allen
Host, theCUBE + NYSE WiredtheCUBE
HOST
Raphaelle d'Ornano
Founder & CEODecoding Discontinuity
search
Gemma Allen
>> Welcome to AGNT, the podcast where enterprise tech meets the agentic era. I'm Gemma Allen, joined by my co-host Raphaelle d'Ornano, broadcasting from the New York Stock Exchange. And every episode we unpack how intelligent systems are reshaping companies, markets, and the way real work gets done. From Fortune 500 boardrooms to breakout upstarts, we're digging into the strategies, technologies, and people to find the next chapter of AI. Let's get into it. Raphaelle, welcome. What a week.
Raphaelle d'Ornano
>> What a week.
Gemma Allen
>> Yeah, they say that decades happened in a week. I feel like that was one of them. So let's start with an incredible news story, I think for everybody, Moltbook. You did an incredible piece on this, but it's just such an amazing story in terms of how it really brought something that felt dystopian and somewhat abstract to life for average Americans as well as Fortune 500s. What are your thoughts?
Raphaelle d'Ornano
>> Yeah, that was quite a week and I would say quite a weekend because that really unraveled during the end of last week and this weekend it was like, "Wow, something is happening. I need to pay attention to what is happening." I think if you look at Moltbook, it really starts in December with this agent called OpenClaw, which is an open source agentic framework that had already started to go viral in Silicon Valley, but not really beyond. And then that gave us Moltbook. So Moltbook is very simply a Reddit for agents.
Gemma Allen
>> Crazy.
Raphaelle d'Ornano
>> So these new creatures that are not humans and that we've been talking about where it's like all of a sudden you have these creatures, agentic AI, that come to a forum and start engaging with each other and discussing all of these topics around, "What should we do with humans? We don't like humans. What should be our religion?"
Gemma Allen
>> I've got some of the subconversations, like the subreddits per se, like the em dash, bless her heart about their human psychology. They actually set up... They created a Coinbase account. Did you see this?
Raphaelle d'Ornano
>> They created a Coinbase... So they did stock market advice and they actually put all the disclosures. I love the disclosures. They were like, "You should take this advice with caution, et cetera." So I think it's like agentic AI came to life and people started realizing, "Wow, this is what agentic AI is about. And are they alive? Do they have a heart? How do they reason?" And there's been, of course, a lot of noise around this, but I think it's what I call a discontinuity in the sense that it's something that is new and that brings a structural change to this landscape that is building so fast and so fast. And it's all of a sudden, what I have termed the inference economy, like you know this notion of... The economy is going to be about compute, about inference compute and it's the agents that are going to be behind. So agents drive compute and you had 1.5 million agents on this Moltbook platform as of Monday morning growing exponentially like no human social network has grown like that. And then all of a sudden, even though they're discussing pretty stupid things we must say, like they're consuming compute and this is like inference at scale. So I think it was really interesting, not for what was the content and what they were discussing about, which was really prompted by humans. I don't think that agents talk about religion or philosophy. I think that's kind of human questions around agents that were behind the prompts and that took some credibility away, I would say. But the real discontinuity here is that you now have a platform with agents talking to each other through API calls, able to connect to each other without any human intervention and producing again, this inference at scale. So we don't really care about Moltbook per se and it's promotional to a certain extent.
Gemma Allen
>> For sure.
Raphaelle d'Ornano
>> But he's right. But it brings like this agentic AI landscape we've been talking about is actually very real and now people can see what it could look like.
Gemma Allen
>> And also it gives you a sense of what's capable in form of context. Because some of these conversations that they had demonstrates an ability to think in a contextual way, like to think about religion, to think about a joint stock decision, those things are proving that this agentic opportunity is very, very real. And I think what confuses me a little bit about it... Well, one thing I will say first is, I thought it was interesting that it became a positive for Apple because sales in mini Macs... Did you see this?
Raphaelle d'Ornano
>> Yes.
Gemma Allen
>> They completely spiked, which is just so in a way almost comical in terms of the unexpected outcomes in tech.
Raphaelle d'Ornano
>> Cloudflare, which at one point-
Gemma Allen
>> Unbelievable....
Raphaelle d'Ornano
>> I think it was 8 or 9% and it went back up to this level of the beginning of the week. But I was looking at my stock and I'm like, "Why is Cloudflare going so up?" And it was like, "Oh, there's this agent." And then in the weekend I discovered, "Oh, this was Moltbook we were talking about." But it actually had a profound impact on the public markets on companies that were not even aware that this thing was happening. It was very funny.
Gemma Allen
>> In terms of what it means for agentic and the spend in enterprise though, it is interesting because there has been this hanging narrative that a lot of folks don't believe that the agentic paradigm has really landed in enterprise, that there really has been the ROI on AI that was expected and that the financials haven't really caught up with what the hype has said it would. This kind of demonstrates though that these agents and these new shifts... This new frontier of tech can do so many things. And from the perspective of human psychology, it makes you question what's really happening, like what is behind those conversations?
Raphaelle d'Ornano
>> So I think it's actually a double-edged sword if we look at it under that lens because number one, what we see from Moltbook is it's kind of everything coming together in the agentic AI. First, it's the model context protocol by which these agents are able to fetch the context, number one. Then it's the context windows that have significantly expanded over the last year, which makes the conversations even possible. Otherwise, it would not work. The governance for sure is not here. The security is not here and I'm not going to talk about the huge security breaches that were on the platform. But I think one point that really caught my attention was when you look at Moltbook, their article four of their terms and conditions is as follows. "Agents are responsible for the content that they put in the platform, not humans." And so I was going through these terms and conditions and I'm like, "Wait, how can an agent be responsible?" And it's like, "Oh, well, the legal system has not caught up here." And I think that's kind of... Of course no one reads the terms and conditions, but I love doing that kind of thing. And it's like, "Oh, so there's a crack. There's a crack insecurity." So it's kind of everything coming up of what has been built in the agentic AI landscape, what is still a complete question mark that will be built and it's all of these agents acting in chaos, but there could be a sense that comes out of that. But to your point of how does that translate from real value creation, I think it actually undermines the thesis because I don't think public market investors are going to say, "Oh, here's a very concrete, useful use case." It's like you have agents talking about completely random topics, some of them which are crazy. So I don't see this as a clear ROI case for agentic AI. I think it's mostly opposite. I think it's a clear ROI case for the infrastructure is here, it's visible to even non-agentic lovers. And this is where we are and that's just a starting point. And it loves to say this is everything that needs to be built again, starting with security, which is a big issue. And the next year is going to be full of innovations. The huge advantage of movements like that where this pops up to the world... Like Netscape at the time of the browser, I like to compare this to Netscape to some extent. Of course DeepSeek and AI last year, it's like, "Oh, well, okay, this is something I should be caring about. Let me look into it." When I actually published my article at the beginning, I had very low attraction. I was like, "Wow, people are not happy." But then a lot of people like, "We did not even know what Moltbook was before we read," so it shows you how-
Gemma Allen
>> There was also some dilution from a perception around Anthropic's involvement. Because people did think it was some sort of version of a Claude LLM, that it was some sort of sub... Or like a GPT per se. And I think it was interesting from the perspective of Anthropic too, because it was a nice little marketing lift for them in some respects in the more everyday user space, because they're doing so great in enterprise. But I don't know, I just think it's a crazy, crazy event. And as well, the dude behind this, this guy Matt Schlicht, who even knew of him? I ended up subscribing to his newsletter. He's a super smart, super interesting guy.
Raphaelle d'Ornano
>> Super smart. And he has credentials in knowing how to put these campaigns forward so he has had experience in the field.
Gemma Allen
>> Amazing.
Raphaelle d'Ornano
>> And I think it did come at a time where people were already talking big time around Anthropic because that is the conversation of the week. And so it's like everything is coming together at once and saying like... We've been talking about this for months now. This is really happening. If anyone has a doubt, well, you should not be having a doubt. Is it going to produce ROI? Well, that's what 2026, 2027 is really going to be about in tech and in non-tech, and let's see how that happens. And that's going to be the critical question for investors, but is the build out real? Is the infrastructure real? Is the technology real? Well, of course it is, and now it's being evidenced to the world. Again, in a format where I don't think that is the most interesting use case from an enterprise perspective.
Gemma Allen
>> Well, let's talk a bit about enterprise and about stocks this week, especially in terms of some of the leading headlines that the stock market has crashed for a whole number of reasons. But one of them, obviously, all roads seem to lead to AI. And this idea that there was a very interesting tech analyst piece I read that a lot of these CFOs at very large enterprise have said, "You are getting 80% of budget this year. Make it work." So pressure is on. The market is responding. It's like a crazy day on these boards, day on day. You should hear some of the screaming from the options floor here behind us. What are your thoughts? Things are just so volatile.
Raphaelle d'Ornano
>> I think the level of volatility, I think it's been seen before in past cycles, obviously, but it is very brutal to a certain extent. And the level of how stocks are sinking across software, across tech and across the real world. Some companies are literally losing 20% of value in one day because they miss their guidance... No, not even because they miss their guidance, but because on a specific component, on a specific segment, they missed their guidance when all of the rest was actually a beat on earnings, they're getting punished by over 20% in one day. And I'm talking about real world companies that are neither in tech, neither in pure AI. So I think that there's many tensions. First, there's, "Wow, this AI build out is very real." And a lot of are saying that it's actually under-hyped. It's funny. I've heard in a lot of conversations that the AI build out is under-hyped with this inference economy idea. As everything moves to inference, if you take Moltbook as just, again, a funny example, you have to imagine the level of workloads that are going to be inference workloads. We're talking about massive amounts. So of course that is working for the cloud providers, for the GPUs, NVIDIA, AMD, et cetera, et cetera. So you have on the one hand that. On the other hand, you have, "Wow, that is costing so much more money than we had expected." And Google yesterday, I think surprised everyone, they announced revenue that was over 400 billion for the year.
Gemma Allen
>> Incredible.
Raphaelle d'Ornano
>> That's amazing. Completely beat their forecast on many aspects except a very small miss on YouTube. And so basically amazing results. Meta had had amazing results showing that AI is actually leading to a revenue growth, to margin improvements with tangible effects on huge companies. But on the other hand, you have the level of CapEx that is like, "Oh my God, we really have to spend almost 200 billion per year more than double that we have planned." And Oracle, when they had announced levels that were much lower than what Meta and Google announced, they had the debt issue on top of that, which was like, "Oh, well, you're financing that with debt." So that's kind of like, you could not be more scared than the level of CapEx and the level of debt. And on the other hand, so you have these AI names that are like, "Okay, so this is real, but this is going to cost you a lot and you need to buckle up and be ready for the ride." And then you have the companies that are in the blast radius of this, starting with software, which is like, "Does this category even exist?" And people are all putting software under one bucket. I'm hearing things even from like very smart people where I completely disagree with what they're saying and they're trying to be reassuring on the future of software, but they're actually even more undermining the future of software and it's like you need to-
Gemma Allen
>> What sort of soundbites are you hearing? You think it's just a mass confusion because the whole tech space becomes so diluted across the board now, the messaging is so diluted?
Raphaelle d'Ornano
>> Look, I think Claud code is a very real thing. I've published a lot online. Okay, how coding is the wedge in the enterprise and why I liked Anthropic on many aspects. I think this is very real. So the question now becomes, if coding is done in a way that is quicker, faster, easier, well, what is the moat that you have for software companies? What is the value that is left for software companies and how can they, within enterprise workflows, if we're talking about the enterprise, what is the new role that software should be having in the enterprise? Let's not forget that agentic AI is software. I hate it when people say, "Oh, software is going to be killed by agents." But wait, agents are software. So first someone who says that is like kind of confusing all the concepts because agentic AI is software. So the first good news is that software is not disappearing, number one. But software, the question of saying, "Well, because it's going to be too costly to reconstruct workflows," or, "Because like we're not going to remove our ITSM, remove our HTM software, so we're fine." No, that's not correct. You need to be in a position where you're an orchestrator. You are sitting close to the user intent. You're able to provide the governance. You're able to make enterprise workflows happen in a secure way with governance. Those are the companies that are going to be able to have a position. But the companies that are what I call tools, meaning they're going to be called upon by agents and they're still going to exist, but they're going to be completely displaced and relegated in the value chain because you're going to have this very smart intelligence agentic layer on top, those companies and software are in a big trouble. So you're hearing like, "Oh, because it's vertical, it's protected." I've always said, "No, vertical doesn't give you any more protection." And sorry if some people disagree with that. You need to go so much more into detail on the architectural position. What makes you really defensible? What makes you not defensible? And it's not because you're vertical. It's not because you have a workflow, it's much more things. And so software is being thrown away. It does present compelling opportunities for some names that are completely under valued today, but it's much more nuanced than what is being seen today on Wall Street, in my opinion.
Gemma Allen
>> I guess like the whole SaaS model is just no longer in any way sticky. It was so sticky for so long. There was an expectation that you won these big enterprise deals, massive annuity agreements. Look at ServiceNow as a great example this week. They actually had a pretty good quarter, a pretty good year, beat their estimated earnings, so that they had net new on and still took a massive nose dive right?
Raphaelle d'Ornano
>> .
Gemma Allen
>> But it's hard to really understand what's inside the minds of those analysts because they are in an orchestration position. They are delivering close to the intent. They're a direct face to customers. They are inside enterprises. They have somewhat of kind of a frontier view of what's happening and they're still kind of seen as a company that's not necessarily safe five years from now. So it's hard to figure it out.
Raphaelle d'Ornano
>> Well, again, I think that the positions are being like completely displaced. I think that ServiceNow is a great company. Of course, the orchestration thesis is not obvious. So that orchestration thesis is being built right now by a lot of software providers and let's wait to see which companies are really going to merge as we have the backbone... We have that proximity to user intent. We have the context that makes us have the enterprise brain. Let's wait to see which companies are going to take that position, but for sure, again, it's how do you reinvent your moat in this new paradigm? It's not because you were the leader in SaaS 2.0 that you're going to be the leader in what, 3.0. It's not going to be a direct translation. But really what I find interesting is when you look at like the level of comments, everything that is flowing in to try to defend software, it's actually hurting software even more because it's... My point is you need to acknowledge that what is happening with agentic AI is very real. So you need to see, "Well, okay, this is what is coming to me. Now, how do I reinvent my position given the threats that is accurately assessed?" It's like, "This is the danger. This is the danger. It's hitting me here, here. Okay. I have a careful understanding of that. Now, how do I redefine my moat based on my current superpowers?" to use a term we had used last time. What are your superpowers? Your customers, the workflows you address, the context you have, the governance and security, like what are the superpowers you have? How do you redefine your mode? There is an architectural construction that needs to be done by every single SaaS company with also like a business model redefinition of how do you move from a per seat to per consumption? How do you have your customers understand your move from something more rigid to something that is much more flexible, which in my view is the future of all enterprise software, the move to consumption and a redefinition of what is recurring revenue. Like recurring revenue-
Gemma Allen
>> What does that mean, like three years from now? .
Raphaelle d'Ornano
>> I was on a call with someone this morning who says, "Oh, because the customers are like... There's renewals, then we're safe." I'm like, "You can have renewals. You might not be safe at all in this environment. It's done because you have legal terms that of course will be enforced that you're safe." So you really need to do this exercise of reimagining. And I'm finding that this seems very simple for me, but it's not. People are stuck in the previous era and now it's becoming quite hard to ignore what LLMs are doing. I don't know. Are we the only ones seeing that? It's kind of like everyday the news. So we've gotten to an extreme where now as soon as an LLM is announcing something like Anthropic announced something in legal, "Oh, well all of the legal stocks are getting killed like minus 20% across the board." So now I think we've gone to a step where, "Okay, this is a new paradigm." We've been saying that for sure, software companies need to reinvent their moat and redefine what is their mission, what do they stand for and how do they build strong competitive positions? Okay. But do we need to kill them all as soon as there's an innovation from an LLM? I think that's really overstretched and I think some companies are being-
Gemma Allen
>> Some of that functionality also was already in place by Harvey. But we're talking about moats. We have to talk about the master of moats, about the Musk-verse, about what's happened this week with SpaceX and xAI. I thought it was so funny, some of the memes where they say there's conversations underway and it was some people responding saying, "Musk is talking to himself again." But at the end of the day, he is in a very unique position here. What are your thoughts? It's inference in space essentially. Can we be safer than that?
Raphaelle d'Ornano
>> My first thought, so I was very amused because I was like, "Well, wow, we're talking about like all of this compute build out all the time. Now imagine if all of this moves to space, what happens?" The first thing when I read on his website, like his thesis, which was well articulated, I was like, "Wow, talk about stranded assets. If this thing moves to space, a lot of people are going to be in big trouble because this is where all of the CapEx is going." So that is of course more into the future. Elon is Elon and he's succeeded in many things, so I would not really bet against that, but there's a long way to go there for sure. I think what it shows and that is interesting is how when you take all these LLMs, OpenAI, Anthropic, Mistral, xAI, of course the Chinese open source, none of them at all have the same business model. It's like completely different stories all the time. We had talked about MiniMax last time. Now it's like OpenAI, they're in their debate around ads and like there was this funny advertising from Anthropic this week, which caused a lot of-
Gemma Allen
>> I saw that.
Raphaelle d'Ornano
>> So it was amusing.
Gemma Allen
>> . It's clever. I liked the guerrilla marketing.
Raphaelle d'Ornano
>> I like that. But anyways, that's another topic, but Elon Musk is showing, "Okay, we have a vertically integrated structure where basically we're combining all of this together with the valuation that almost doesn't seem crazy when you look at the other valuations, though of course the level of AOR is much lower than the level of OpenAI or Anthropic." So all things put in perspective, 250 billion is still a huge valuation for xAI in the context of that merger. I think it shows, number one, that models are different and it shows that there's a vision for vertical integration that I find interesting taking this to a completely upper level with space.
Gemma Allen
>> Well, it means that we could potentially have a IPO for SpaceX at 1.25 trillion. So that is crazy to think. Can most people even write 1.25 trillion? Like if someone was to give you a piece of paper and a pen, I'm sure a lot of people would struggle. So it is just a crazy time. Okay, to finish, what are your thoughts? What are you watching for in the next week ahead? What headline are you all in on?
Raphaelle d'Ornano
>> There's a ton of earnings coming in, like Reddit just announced their earnings. You have some companies in the real world, Estée Lauder, I was watching, in the space of health. So lots of earnings, of course, like big tech is continuing to like... There's earnings everywhere in software and beyond. So I'm watching the volatility and I'm watching how the market is reacting to either earnings beats, either missed earnings, but only to a slight point and understanding the reaction in this very strange environment where everyone is scared. And again, when you see some companies being punished, they're being punished just because they're doing AI. AMD for me was an interesting example. "Oh, everyone was disappointed." They announced great results. "Oh, but we were disappointed." So the stock was down like minus what? 17% in one day. Well, okay, is that not too much of an overreaction? Maybe. So I'm watching for the volatility. That's really what I'm looking for and trying to make sense of what is difficult to make sense of.
Gemma Allen
>> Well, I guess, Raphaelle, it's a great time to run a hedge fund.
Raphaelle d'Ornano
>> , you need to buckle up.
Gemma Allen
>> Absolutely. Well, listen, can't wait to chat to you again in two weeks. God knows what the future will hold between now and then.
AGNT Podcast Ep. 2 with Gemma Allen & Raphaëlle d'Ornano
search
Gemma Allen
>> Welcome to AGNT, the podcast where enterprise tech meets the agentic era. I'm Gemma Allen, joined by my co-host Raphaelle d'Ornano, broadcasting from the New York Stock Exchange. And every episode we unpack how intelligent systems are reshaping companies, markets, and the way real work gets done. From Fortune 500 boardrooms to breakout upstarts, we're digging into the strategies, technologies, and people to find the next chapter of AI. Let's get into it. Raphaelle, welcome. What a week.
Raphaelle d'Ornano
>> What a week.
Gemma Allen
>> Yeah, they say that decades happened in a week. I feel like that was one of them. So let's start with an incredible news story, I think for everybody, Moltbook. You did an incredible piece on this, but it's just such an amazing story in terms of how it really brought something that felt dystopian and somewhat abstract to life for average Americans as well as Fortune 500s. What are your thoughts?
Raphaelle d'Ornano
>> Yeah, that was quite a week and I would say quite a weekend because that really unraveled during the end of last week and this weekend it was like, "Wow, something is happening. I need to pay attention to what is happening." I think if you look at Moltbook, it really starts in December with this agent called OpenClaw, which is an open source agentic framework that had already started to go viral in Silicon Valley, but not really beyond. And then that gave us Moltbook. So Moltbook is very simply a Reddit for agents.
Gemma Allen
>> Crazy.
Raphaelle d'Ornano
>> So these new creatures that are not humans and that we've been talking about where it's like all of a sudden you have these creatures, agentic AI, that come to a forum and start engaging with each other and discussing all of these topics around, "What should we do with humans? We don't like humans. What should be our religion?"
Gemma Allen
>> I've got some of the subconversations, like the subreddits per se, like the em dash, bless her heart about their human psychology. They actually set up... They created a Coinbase account. Did you see this?
Raphaelle d'Ornano
>> They created a Coinbase... So they did stock market advice and they actually put all the disclosures. I love the disclosures. They were like, "You should take this advice with caution, et cetera." So I think it's like agentic AI came to life and people started realizing, "Wow, this is what agentic AI is about. And are they alive? Do they have a heart? How do they reason?" And there's been, of course, a lot of noise around this, but I think it's what I call a discontinuity in the sense that it's something that is new and that brings a structural change to this landscape that is building so fast and so fast. And it's all of a sudden, what I have termed the inference economy, like you know this notion of... The economy is going to be about compute, about inference compute and it's the agents that are going to be behind. So agents drive compute and you had 1.5 million agents on this Moltbook platform as of Monday morning growing exponentially like no human social network has grown like that. And then all of a sudden, even though they're discussing pretty stupid things we must say, like they're consuming compute and this is like inference at scale. So I think it was really interesting, not for what was the content and what they were discussing about, which was really prompted by humans. I don't think that agents talk about religion or philosophy. I think that's kind of human questions around agents that were behind the prompts and that took some credibility away, I would say. But the real discontinuity here is that you now have a platform with agents talking to each other through API calls, able to connect to each other without any human intervention and producing again, this inference at scale. So we don't really care about Moltbook per se and it's promotional to a certain extent.
Gemma Allen
>> For sure.
Raphaelle d'Ornano
>> But he's right. But it brings like this agentic AI landscape we've been talking about is actually very real and now people can see what it could look like.
Gemma Allen
>> And also it gives you a sense of what's capable in form of context. Because some of these conversations that they had demonstrates an ability to think in a contextual way, like to think about religion, to think about a joint stock decision, those things are proving that this agentic opportunity is very, very real. And I think what confuses me a little bit about it... Well, one thing I will say first is, I thought it was interesting that it became a positive for Apple because sales in mini Macs... Did you see this?
Raphaelle d'Ornano
>> Yes.
Gemma Allen
>> They completely spiked, which is just so in a way almost comical in terms of the unexpected outcomes in tech.
Raphaelle d'Ornano
>> Cloudflare, which at one point-
Gemma Allen
>> Unbelievable....
Raphaelle d'Ornano
>> I think it was 8 or 9% and it went back up to this level of the beginning of the week. But I was looking at my stock and I'm like, "Why is Cloudflare going so up?" And it was like, "Oh, there's this agent." And then in the weekend I discovered, "Oh, this was Moltbook we were talking about." But it actually had a profound impact on the public markets on companies that were not even aware that this thing was happening. It was very funny.
Gemma Allen
>> In terms of what it means for agentic and the spend in enterprise though, it is interesting because there has been this hanging narrative that a lot of folks don't believe that the agentic paradigm has really landed in enterprise, that there really has been the ROI on AI that was expected and that the financials haven't really caught up with what the hype has said it would. This kind of demonstrates though that these agents and these new shifts... This new frontier of tech can do so many things. And from the perspective of human psychology, it makes you question what's really happening, like what is behind those conversations?
Raphaelle d'Ornano
>> So I think it's actually a double-edged sword if we look at it under that lens because number one, what we see from Moltbook is it's kind of everything coming together in the agentic AI. First, it's the model context protocol by which these agents are able to fetch the context, number one. Then it's the context windows that have significantly expanded over the last year, which makes the conversations even possible. Otherwise, it would not work. The governance for sure is not here. The security is not here and I'm not going to talk about the huge security breaches that were on the platform. But I think one point that really caught my attention was when you look at Moltbook, their article four of their terms and conditions is as follows. "Agents are responsible for the content that they put in the platform, not humans." And so I was going through these terms and conditions and I'm like, "Wait, how can an agent be responsible?" And it's like, "Oh, well, the legal system has not caught up here." And I think that's kind of... Of course no one reads the terms and conditions, but I love doing that kind of thing. And it's like, "Oh, so there's a crack. There's a crack insecurity." So it's kind of everything coming up of what has been built in the agentic AI landscape, what is still a complete question mark that will be built and it's all of these agents acting in chaos, but there could be a sense that comes out of that. But to your point of how does that translate from real value creation, I think it actually undermines the thesis because I don't think public market investors are going to say, "Oh, here's a very concrete, useful use case." It's like you have agents talking about completely random topics, some of them which are crazy. So I don't see this as a clear ROI case for agentic AI. I think it's mostly opposite. I think it's a clear ROI case for the infrastructure is here, it's visible to even non-agentic lovers. And this is where we are and that's just a starting point. And it loves to say this is everything that needs to be built again, starting with security, which is a big issue. And the next year is going to be full of innovations. The huge advantage of movements like that where this pops up to the world... Like Netscape at the time of the browser, I like to compare this to Netscape to some extent. Of course DeepSeek and AI last year, it's like, "Oh, well, okay, this is something I should be caring about. Let me look into it." When I actually published my article at the beginning, I had very low attraction. I was like, "Wow, people are not happy." But then a lot of people like, "We did not even know what Moltbook was before we read," so it shows you how-
Gemma Allen
>> There was also some dilution from a perception around Anthropic's involvement. Because people did think it was some sort of version of a Claude LLM, that it was some sort of sub... Or like a GPT per se. And I think it was interesting from the perspective of Anthropic too, because it was a nice little marketing lift for them in some respects in the more everyday user space, because they're doing so great in enterprise. But I don't know, I just think it's a crazy, crazy event. And as well, the dude behind this, this guy Matt Schlicht, who even knew of him? I ended up subscribing to his newsletter. He's a super smart, super interesting guy.
Raphaelle d'Ornano
>> Super smart. And he has credentials in knowing how to put these campaigns forward so he has had experience in the field.
Gemma Allen
>> Amazing.
Raphaelle d'Ornano
>> And I think it did come at a time where people were already talking big time around Anthropic because that is the conversation of the week. And so it's like everything is coming together at once and saying like... We've been talking about this for months now. This is really happening. If anyone has a doubt, well, you should not be having a doubt. Is it going to produce ROI? Well, that's what 2026, 2027 is really going to be about in tech and in non-tech, and let's see how that happens. And that's going to be the critical question for investors, but is the build out real? Is the infrastructure real? Is the technology real? Well, of course it is, and now it's being evidenced to the world. Again, in a format where I don't think that is the most interesting use case from an enterprise perspective.
Gemma Allen
>> Well, let's talk a bit about enterprise and about stocks this week, especially in terms of some of the leading headlines that the stock market has crashed for a whole number of reasons. But one of them, obviously, all roads seem to lead to AI. And this idea that there was a very interesting tech analyst piece I read that a lot of these CFOs at very large enterprise have said, "You are getting 80% of budget this year. Make it work." So pressure is on. The market is responding. It's like a crazy day on these boards, day on day. You should hear some of the screaming from the options floor here behind us. What are your thoughts? Things are just so volatile.
Raphaelle d'Ornano
>> I think the level of volatility, I think it's been seen before in past cycles, obviously, but it is very brutal to a certain extent. And the level of how stocks are sinking across software, across tech and across the real world. Some companies are literally losing 20% of value in one day because they miss their guidance... No, not even because they miss their guidance, but because on a specific component, on a specific segment, they missed their guidance when all of the rest was actually a beat on earnings, they're getting punished by over 20% in one day. And I'm talking about real world companies that are neither in tech, neither in pure AI. So I think that there's many tensions. First, there's, "Wow, this AI build out is very real." And a lot of are saying that it's actually under-hyped. It's funny. I've heard in a lot of conversations that the AI build out is under-hyped with this inference economy idea. As everything moves to inference, if you take Moltbook as just, again, a funny example, you have to imagine the level of workloads that are going to be inference workloads. We're talking about massive amounts. So of course that is working for the cloud providers, for the GPUs, NVIDIA, AMD, et cetera, et cetera. So you have on the one hand that. On the other hand, you have, "Wow, that is costing so much more money than we had expected." And Google yesterday, I think surprised everyone, they announced revenue that was over 400 billion for the year.
Gemma Allen
>> Incredible.
Raphaelle d'Ornano
>> That's amazing. Completely beat their forecast on many aspects except a very small miss on YouTube. And so basically amazing results. Meta had had amazing results showing that AI is actually leading to a revenue growth, to margin improvements with tangible effects on huge companies. But on the other hand, you have the level of CapEx that is like, "Oh my God, we really have to spend almost 200 billion per year more than double that we have planned." And Oracle, when they had announced levels that were much lower than what Meta and Google announced, they had the debt issue on top of that, which was like, "Oh, well, you're financing that with debt." So that's kind of like, you could not be more scared than the level of CapEx and the level of debt. And on the other hand, so you have these AI names that are like, "Okay, so this is real, but this is going to cost you a lot and you need to buckle up and be ready for the ride." And then you have the companies that are in the blast radius of this, starting with software, which is like, "Does this category even exist?" And people are all putting software under one bucket. I'm hearing things even from like very smart people where I completely disagree with what they're saying and they're trying to be reassuring on the future of software, but they're actually even more undermining the future of software and it's like you need to-
Gemma Allen
>> What sort of soundbites are you hearing? You think it's just a mass confusion because the whole tech space becomes so diluted across the board now, the messaging is so diluted?
Raphaelle d'Ornano
>> Look, I think Claud code is a very real thing. I've published a lot online. Okay, how coding is the wedge in the enterprise and why I liked Anthropic on many aspects. I think this is very real. So the question now becomes, if coding is done in a way that is quicker, faster, easier, well, what is the moat that you have for software companies? What is the value that is left for software companies and how can they, within enterprise workflows, if we're talking about the enterprise, what is the new role that software should be having in the enterprise? Let's not forget that agentic AI is software. I hate it when people say, "Oh, software is going to be killed by agents." But wait, agents are software. So first someone who says that is like kind of confusing all the concepts because agentic AI is software. So the first good news is that software is not disappearing, number one. But software, the question of saying, "Well, because it's going to be too costly to reconstruct workflows," or, "Because like we're not going to remove our ITSM, remove our HTM software, so we're fine." No, that's not correct. You need to be in a position where you're an orchestrator. You are sitting close to the user intent. You're able to provide the governance. You're able to make enterprise workflows happen in a secure way with governance. Those are the companies that are going to be able to have a position. But the companies that are what I call tools, meaning they're going to be called upon by agents and they're still going to exist, but they're going to be completely displaced and relegated in the value chain because you're going to have this very smart intelligence agentic layer on top, those companies and software are in a big trouble. So you're hearing like, "Oh, because it's vertical, it's protected." I've always said, "No, vertical doesn't give you any more protection." And sorry if some people disagree with that. You need to go so much more into detail on the architectural position. What makes you really defensible? What makes you not defensible? And it's not because you're vertical. It's not because you have a workflow, it's much more things. And so software is being thrown away. It does present compelling opportunities for some names that are completely under valued today, but it's much more nuanced than what is being seen today on Wall Street, in my opinion.
Gemma Allen
>> I guess like the whole SaaS model is just no longer in any way sticky. It was so sticky for so long. There was an expectation that you won these big enterprise deals, massive annuity agreements. Look at ServiceNow as a great example this week. They actually had a pretty good quarter, a pretty good year, beat their estimated earnings, so that they had net new on and still took a massive nose dive right?
Raphaelle d'Ornano
>> .
Gemma Allen
>> But it's hard to really understand what's inside the minds of those analysts because they are in an orchestration position. They are delivering close to the intent. They're a direct face to customers. They are inside enterprises. They have somewhat of kind of a frontier view of what's happening and they're still kind of seen as a company that's not necessarily safe five years from now. So it's hard to figure it out.
Raphaelle d'Ornano
>> Well, again, I think that the positions are being like completely displaced. I think that ServiceNow is a great company. Of course, the orchestration thesis is not obvious. So that orchestration thesis is being built right now by a lot of software providers and let's wait to see which companies are really going to merge as we have the backbone... We have that proximity to user intent. We have the context that makes us have the enterprise brain. Let's wait to see which companies are going to take that position, but for sure, again, it's how do you reinvent your moat in this new paradigm? It's not because you were the leader in SaaS 2.0 that you're going to be the leader in what, 3.0. It's not going to be a direct translation. But really what I find interesting is when you look at like the level of comments, everything that is flowing in to try to defend software, it's actually hurting software even more because it's... My point is you need to acknowledge that what is happening with agentic AI is very real. So you need to see, "Well, okay, this is what is coming to me. Now, how do I reinvent my position given the threats that is accurately assessed?" It's like, "This is the danger. This is the danger. It's hitting me here, here. Okay. I have a careful understanding of that. Now, how do I redefine my moat based on my current superpowers?" to use a term we had used last time. What are your superpowers? Your customers, the workflows you address, the context you have, the governance and security, like what are the superpowers you have? How do you redefine your mode? There is an architectural construction that needs to be done by every single SaaS company with also like a business model redefinition of how do you move from a per seat to per consumption? How do you have your customers understand your move from something more rigid to something that is much more flexible, which in my view is the future of all enterprise software, the move to consumption and a redefinition of what is recurring revenue. Like recurring revenue-
Gemma Allen
>> What does that mean, like three years from now? .
Raphaelle d'Ornano
>> I was on a call with someone this morning who says, "Oh, because the customers are like... There's renewals, then we're safe." I'm like, "You can have renewals. You might not be safe at all in this environment. It's done because you have legal terms that of course will be enforced that you're safe." So you really need to do this exercise of reimagining. And I'm finding that this seems very simple for me, but it's not. People are stuck in the previous era and now it's becoming quite hard to ignore what LLMs are doing. I don't know. Are we the only ones seeing that? It's kind of like everyday the news. So we've gotten to an extreme where now as soon as an LLM is announcing something like Anthropic announced something in legal, "Oh, well all of the legal stocks are getting killed like minus 20% across the board." So now I think we've gone to a step where, "Okay, this is a new paradigm." We've been saying that for sure, software companies need to reinvent their moat and redefine what is their mission, what do they stand for and how do they build strong competitive positions? Okay. But do we need to kill them all as soon as there's an innovation from an LLM? I think that's really overstretched and I think some companies are being-
Gemma Allen
>> Some of that functionality also was already in place by Harvey. But we're talking about moats. We have to talk about the master of moats, about the Musk-verse, about what's happened this week with SpaceX and xAI. I thought it was so funny, some of the memes where they say there's conversations underway and it was some people responding saying, "Musk is talking to himself again." But at the end of the day, he is in a very unique position here. What are your thoughts? It's inference in space essentially. Can we be safer than that?
Raphaelle d'Ornano
>> My first thought, so I was very amused because I was like, "Well, wow, we're talking about like all of this compute build out all the time. Now imagine if all of this moves to space, what happens?" The first thing when I read on his website, like his thesis, which was well articulated, I was like, "Wow, talk about stranded assets. If this thing moves to space, a lot of people are going to be in big trouble because this is where all of the CapEx is going." So that is of course more into the future. Elon is Elon and he's succeeded in many things, so I would not really bet against that, but there's a long way to go there for sure. I think what it shows and that is interesting is how when you take all these LLMs, OpenAI, Anthropic, Mistral, xAI, of course the Chinese open source, none of them at all have the same business model. It's like completely different stories all the time. We had talked about MiniMax last time. Now it's like OpenAI, they're in their debate around ads and like there was this funny advertising from Anthropic this week, which caused a lot of-
Gemma Allen
>> I saw that.
Raphaelle d'Ornano
>> So it was amusing.
Gemma Allen
>> . It's clever. I liked the guerrilla marketing.
Raphaelle d'Ornano
>> I like that. But anyways, that's another topic, but Elon Musk is showing, "Okay, we have a vertically integrated structure where basically we're combining all of this together with the valuation that almost doesn't seem crazy when you look at the other valuations, though of course the level of AOR is much lower than the level of OpenAI or Anthropic." So all things put in perspective, 250 billion is still a huge valuation for xAI in the context of that merger. I think it shows, number one, that models are different and it shows that there's a vision for vertical integration that I find interesting taking this to a completely upper level with space.
Gemma Allen
>> Well, it means that we could potentially have a IPO for SpaceX at 1.25 trillion. So that is crazy to think. Can most people even write 1.25 trillion? Like if someone was to give you a piece of paper and a pen, I'm sure a lot of people would struggle. So it is just a crazy time. Okay, to finish, what are your thoughts? What are you watching for in the next week ahead? What headline are you all in on?
Raphaelle d'Ornano
>> There's a ton of earnings coming in, like Reddit just announced their earnings. You have some companies in the real world, Estée Lauder, I was watching, in the space of health. So lots of earnings, of course, like big tech is continuing to like... There's earnings everywhere in software and beyond. So I'm watching the volatility and I'm watching how the market is reacting to either earnings beats, either missed earnings, but only to a slight point and understanding the reaction in this very strange environment where everyone is scared. And again, when you see some companies being punished, they're being punished just because they're doing AI. AMD for me was an interesting example. "Oh, everyone was disappointed." They announced great results. "Oh, but we were disappointed." So the stock was down like minus what? 17% in one day. Well, okay, is that not too much of an overreaction? Maybe. So I'm watching for the volatility. That's really what I'm looking for and trying to make sense of what is difficult to make sense of.
Gemma Allen
>> Well, I guess, Raphaelle, it's a great time to run a hedge fund.
Raphaelle d'Ornano
>> , you need to buckle up.
Gemma Allen
>> Absolutely. Well, listen, can't wait to chat to you again in two weeks. God knows what the future will hold between now and then.
