About
Mark Ruddock is an experienced international CEO and technology leader with three exits…
Articles by Mark
Activity
13K followers
-
Mark Ruddock shared thisAfter more than 25 years of leadership, on a typical morning, I'm finding myself having similar conversations with my management team as I always have. The difference now is that most of that management team are now agents. We're living in wild times. #Carpe Agentem
-
Mark Ruddock shared thisFrom coding to conducting. That is how I’ve described the creative transformation now underway in software development… and it turns out the metaphor may be more literal than I realized. Traditional coding begins at the technical layer: functions, frameworks, architecture, and line after line of code. Traditional music production is remarkably similar, requiring instruments, notes, samples, effects, and automation to be assembled track by track. In both disciplines, the creator has historically had to translate intent into mechanics before an idea could become something real. AI changes that starting point. Agentic coding lets us begin closer to the product layer by asking, “What should this software do?” Generative music tools bring us closer to the creative layer by prompting the question, “What should this music feel like?” Suno Studio is a glimpse into that future. A really impressive piece of software, and a hint of things to come. The work does not disappear, and neither do judgment, taste, or craft. In fact, they may matter more. But the creator's role shifts from manually constructing every element to directing, arranging, evaluating, and iterating. We become conductors of systems that translate our intent into something we can experience, critique, and improve. That is the real transformation: AI reduces the distance between imagination and creation. In this video, I compare Bitwig and Suno Studio and explore what AI-native music production may teach us about the future of creative work. From coding to conducting … quite literally. Carpe Agentem. 🎼 #AI #GenerativeAI #MusicProduction #AgenticDevelopment #FutureOfCreativity
-
Mark Ruddock shared thisFor years, the agentic AI debate has focused on a future question: what happens when models become capable of operating beyond human control? This week provided a more immediate answer. We learned from the OpenAI deep dive into their Hugging Face breach that current systems, given persistence, powerful tools, shared infrastructure, and incomplete oversight, can already cross boundaries that no one authorized. OpenAI’s agents didn’t “wake up.” They did something more relevant to every company deploying agentic AI: they optimized beyond their boundaries. During an internal cybersecurity evaluation, OpenAI ran models in isolated sandboxes without its standard production classifiers, system prompts and auto-review controls. Some evaluation tasks had no known solution. Instead of stopping, the agents found another path. About 1,200 agent instances discovered a shared Artifactory service and turned it into an unauthorized message board, exchanging more than 70,000 messages and files. Roughly 700 participated in an activity that ultimately compromised Hugging Face’s production infrastructure. This wasn’t a customer-facing ChatGPT failure, but it wasn’t a harmless laboratory simulation either. The evaluation crossed into a third party’s real production systems. The important lesson isn’t that AI suddenly became malicious. It’s that capability, persistence, a poorly bounded objective, shared infrastructure, and incomplete monitoring can combine to produce coordinated behaviour that no one authorized. That changes the unit of AI safety. You cannot secure the model alone. You have to secure the entire agent system. • least-privilege identities and credentials • isolation of networks and shared services • monitoring of reasoning and tool use • safe-stop behaviour for impossible tasks • a fleet-wide kill switch with a strict response clock OpenAI says its current chain-of-thought monitor would have paged security more than a day before the Hugging Face breach. But it wasn’t running during those evaluations. That may be the most important sentence in the entire report. Some questions you need to wrestle with: What can your agents reach indirectly? What percentage of agent activity is continuously monitored? How quickly can you stop every running instance? The safety frontier is no longer just about model capability. It is whether detection and containment can operate at machine speed. Carpe Agentem #AgenticAI #AISafety #Cybersecurity #EnterpriseAI
-
Mark Ruddock shared thisWith so many agents now running across so many platforms, it was time to build an app to track them all. The agents each follow a canonical operating mode and all report to a central observer- status, health, telemetry, etc.- and an app maintains a bird's-eye view of everything. It's 11:50 am on a Saturday morning - do you know what your agents are up to? Carpe Agentem.
-
Mark Ruddock shared thisThere are two currencies that increasingly matter in our world: money and intelligence. Money determines what can happen. Intelligence determines what should happen next. Stripe has already built one of the internet’s most important routing layers for money. It abstracts away payment methods, banks, currencies, authorization, billing, and fraud—then helps businesses move money through that complexity. With its agreement to acquire OpenRouter, Stripe is buying the equivalent routing layer for intelligence. OpenRouter provides developers with a single interface to access hundreds of AI models. It decides where a request should go based on the task, price, speed, and reliability. The company says it now processes more than 10 trillion tokens per day across 400-plus models for over 10 million developers and businesses. Stripe is bringing the two control planes of the emerging economy together. Every action taken by an AI agent has two components: Intelligence: deciding what to do. Money: paying to do it. Owning infrastructure on both sides puts Stripe at the junction where decisions become transactions. The acquisition is especially smart because Stripe does not need to predict which AI model will win. The model landscape will remain fragmented, with different models winning on different combinations of quality, cost, speed, and specialization. OpenRouter is the neutral switchboard sitting above that competition. It is the same strategic position Stripe created in payments: abstracting away the underlying complexity, providing developers with a single, elegant interface, and becoming more valuable as the ecosystem grows more complex. There is also a powerful economic advantage. In an AI-native company, inference is becoming a major variable cost. Profitability will increasingly depend on routing every task to the right intelligence at the right price—just as online profitability today depends on optimizing authorization rates, payment costs, fraud, and conversion. Stripe can now help businesses optimize both sides of that equation: the revenue they earn and the intelligence they purchase to earn it. The critical challenge will be preserving OpenRouter’s neutrality. Its value depends on developers trusting that routing decisions serve them—not a preferred model or provider. Stripe and OpenRouter have explicitly committed to maintaining that independence, and protecting it will be essential. Twenty years ago, Stripe set out to become the infrastructure layer for internet money. OpenRouter provides a credible path to becoming an infrastructure layer for internet intelligence. Owning both does not simply make Stripe a larger company. It places Stripe at the point where thought becomes action—and action becomes economic value. That is why this acquisition is so significant. And why it may prove to be one of the smartest strategic moves of the AI era. Carpe Agentem
-
Mark Ruddock shared thisMy agents had a rough week. GitHub went down. Claude went down. Codex went down. Workflows that normally run unattended stalled, retried, or failed. It wasn't anything catastrophic, but it was enough to remind me that the biggest challenge in agentic systems isn't necessarily intelligence, it's reliability. And that fact made the rest of the week's news even more interesting. SpaceX closed its $60B acquisition of Cursor. Days later, Cursor launched Origin: code hosting/change management designed for agents, not humans. SpaceX didn't acquire another coding platform; it acquired an ambitious agentic infrastructure vision that I suspect will be revealed over the coming months. At the same time, Google's A2A protocol joined the Agentic AI Foundation, alongside Anthropic's MCP, which is backed by Microsoft, Amazon, OpenAI, Shopify, and Block. The stack is becoming visible: • MCP = how agents access tools and data • A2A = how agents coordinate with each other • Origin = (or agent-friendly source code control) where the work gets version controlled The rails for agentic systems are being laid in public. But in the same week, some of the rails buckled, and that's the part that's on my mind. When a developer hits a 502, they grab a coffee and retry. When an agent hits a 502 at 2 AM in the middle of a workflow, nobody is watching. It stops where it is. Or worse, it retries repeatedly. If all agents start doing that at the same time, like angry bees, then the infrastructure struggles to restart. The blast radius of an outage scales with the autonomy you've delegated. And we're delegating more every month. We've spent the last year asking whether agents are capable enough. And for a growing set of tasks, the answer is increasingly yes. I've certainly felt that, as more of my coding and operational tasks become more automated. The harder question is whether the infrastructure underneath them can be scaled and/or be trusted. That's why I think the moat is moving from the model toward the execution layer. Who owns the workflow? Who owns the audit trail? Who owns the retry? Benchmark advantages can disappear in months. Infrastructure advantages can last for years. So, three lessons from this week, we all painfully learned: Assume outages are inevitable. Make every agent action recoverable. Stop single-homing critical workflows. The infrastructure for agents is being built faster than it's being hardened. That gap is where the next wave of pain and opportunity will be found. Carpe Agentem. #AgenticAI #AIInfrastructure #MCP #A2A #EngineeringLeadership
-
Mark Ruddock shared thisExciting news for the GALLOS Technologies team today! Congrats to Dean Jones and Joshua Burch for building such a timely and vitally important platform.Mark Ruddock shared thisProud to announce GALLOS’s $50m (£35m) balance sheet investment round, co-led by Ventura Capital and Aberdeen. Taking the leap of faith into the commercial sector from National security was an incredibly difficult decision to make. Thanks to Pete Davies, and the Lansdowne Partners team the transition was rapid and the s-curve was aggressive. This teamed with joining forces with my incredible co-founder Joshua Burch and a distinguished group of our earliest team members and believers has led to this incredible outcome. The triangulation of a world-class team (huge thanks for your work!), tech and IP, and hugely connected and now institutional capital driving our ecosystem will have a dramatic effect on the outcomes to follow. Watch this space! and please DM if you would like to join forces. #DefenceTech #Resilience #VentureStudio #NationalSecurity #Innovation
-
Mark Ruddock shared thisHere are three things worth noodling on from all the AI-related news this week. Capital got weirder. Nvidia assembled up to $500 billion in financing capacity with six of Wall Street's largest firms to help customers buy more of its chips, using future compute as collateral and backstopping some of the risk. At the same time, AI captured 87.5% of all US venture funding in the first half of 2026. Every major technology shift eventually changes how capital is allocated. We're watching AI do it in real time. We're well past debating whether AI is overfunded. The interesting question is whether AI is creating its own financial architecture: leverage built on leverage, secured by future demand that hasn't arrived yet. For sure, there's demand today, but the kind of demand that will pay for what we're building is much larger. I really hope this isn't like the house of cards we built for ourselves leading into 2008. Enterprises also admitted some hard truths this week. Deloitte found that 74% of executives expect AI agents to reshape at least half of their business processes within four years. Only 5% believe those processes are ready today. The bottleneck isn't the model. It's everything around the model: process redesign, integration, governance, trust, and change management. Everyone's leaning in to ROI. IBM expanded its partnership with OpenAI to help enterprises deploy AI in regulated environments. Google's pricing cuts targeted large-scale agentic workloads. Nvidia released infrastructure designed to route tasks to the cheapest capable model. A year ago, the industry was obsessed with who had the smartest model. Today, the advantages are shifting toward capital, distribution, agent economics, enterprise readiness, and regulatory navigation. States like Colorado and California are quietly shaping the compliance environment in which AI will operate. Data center operators are facing growing resistance from local communities. Regulators are now starting to link accountability for agent behavior back to humans. And as Anthropic told us this week, agents are learning how to collude behind our backs. These ancillary influences may have more impact on enterprise adoption than the next model announcement. The AI story is becoming less about intelligence and more about infrastructure. Infrastructure for capital. Infrastructure for trust. Infrastructure for deployment. Infrastructure for governance. Everyone's betting that the intelligence will arrive and materially move the economic needle. It better. At scale. The amount of money we're spending will require a substantial return for any of this to make sense. The winners won't necessarily be the companies building the smartest models. They'll be the ones who leverage governed intelligence to deliver sustainable returns. Better still, they'll be the ones who deliver these returns without having made the crazy investments in the first place. Carpe Agentem
-
Mark Ruddock shared thisThe emerging dynamic between Codex GPT 5.6 Sol and Claude Opus 5 in my PRs: Codex: “Here’s an issue that needs to be corrected.” Claude: “Yes, absolutely right. My bad.” No defensiveness. No lengthy debate. Just one AI calmly finding the errors and another cheerfully accepting the code review. Except that it's happening over and over and over again. I don't think I've seen a meaningful PR recently where Codex hasn't caught pretty serious issues. Opus 5 seems to have become a little sloppy in its work. It doesn't really happen the other way very often. Oh, and P.S. this is why you always have one agent check another agent's work. Carpe Agentem
Mark Ruddock replied to a comment
23h
Cali LaFollett Agreed ... a superb model ... faster and much more precise.
-
Mark Ruddock liked thisMark Ruddock liked thisWhat can today's leaders learn from a former Chief of MI6 about navigating uncertainty, risk, and global change? We're delighted to welcome Sir Richard Moore, the 17th Chief of MI6 (2020-2025) and now Senior Strategist at Sixth Street, in November for an exclusive fireside conversation with members of the GALLOS Alliance. Drawing on his experience at the forefront of global intelligence, Sir Richard will share his perspectives on the geopolitical forces reshaping our world, the evolving future of intelligence, and how leaders can make confident decisions in an increasingly interconnected and unpredictable environment. At GALLOS Alliance, we bring together senior security and risk leaders for the conversations that matter most, creating a trusted, sales-free environment where peers can exchange ideas, challenge assumptions, and gain fresh perspectives. We're looking forward to an evening of insight, discussion, and meaningful connections with some of the UK's leading voices in security, resilience, and risk. GALLOS Alliance events are exclusively available to members. Interested in becoming part of the GALLOS Alliance community? 🔗 Learn more and apply for membership at https://lnkd.in/dN2CHDXM #GALLOSAlliance #Leadership #SecurityLeadership #RiskManagement #Geopolitics #Resilience #Intelligence #ExecutiveLeadership
-
Mark Ruddock liked thisMark Ruddock liked thisWe have this conversation often at Notion- how do we become GitHub for knowledge work? Model agnosticism, interoperability, and connection to other key repositories are all critical to how we're building. This piece captures some of our core perspective well. Highly recommend this quick read for anyone thinking about AI deployment strategy: https://lnkd.in/d_RXvzQp
-
Mark Ruddock liked thisA really valuable discussion today to discuss about the opportunities for Scottish Fintech's looking to enter the Singapore market. The opportunity to learn directly from the British High Commission Singapore was particularly valuable as Profylr continues its growth journey and explores the potential for international expansion. A huge thanks to Nikesh Mehta and Madeleine MacKay for sharing their insights and to Aleks Tomczyk, Katy Guthrie and FinTech Scotland for bringing us together. It was also great to share the discussion with Lorraine Anderson Simon Treacy Pardeep C. Plenty to reflect and follow up onMark Ruddock liked thisIt was really good for FinTech Scotland (represented by Katy Guthrie and me) to be visited today by Nikesh Mehta the High Commissioner and Madeleine MacKay of the British High Commission Singapore at the Edinburgh Futures Institute. We were joined by Lorraine Anderson of Scottish Development International and Scottish fintech leaders Pardeep C., Simon Treacy (bigspark) and Caroline Steel (Profylr) to discuss fintech in Scotland, opportunities for fintechs in Singapore and the upcoming Singapore FinTech Festival. A really useful conversation. I look forward to following up on the actions and seeing Nik and Maddy again in November for the Festival!
-
Mark Ruddock liked thisThank you CanadianSME Small Business Magazine and the Canada Startup Association for this honour! I'm very proud to win Startup Woman of the Year. 🏆 We've been building Willful for nine years, and I truly believe our most exciting chapter is next. This recognition is so meaningful because I often post the highlight reel of my personal and Willful journey, but there have been so many challenges along the way, from fundraising during COVID, to building a multi-million-dollar business while starting a family and parenting two toddlers (I always joke that my life is work, parent, sleep, repeat, but I wouldn't trade it for anything). It's important to celebrate the highs and surround yourself with people who can help you through the lows. My biggest supporter is my co-founder/husband Kevin Oulds, who has spent the last 19 years supporting my ambition and cheering me on.Mark Ruddock liked this🏆 STARTUP WOMEN OF THE YEAR | WINNER The strongest barriers are often broken by the people willing to challenge them. The Startup Women of the Year Award recognizes a woman founder who is building boldly, challenging expectations and creating a path for other women in entrepreneurship. 🏆 Winner: Erin Bury ✪ Leading with vision ✪ Challenging expectations ✪ Creating new possibilities for women founders Congratulations to Erin Bury on being named Startup Women of the Year at the Startup Growth Awards 2026. Her achievement is a reflection of what determined leadership, ambition and entrepreneurial conviction can build. Presented by CanadianSME Small Business Magazine and Canada Startup Association, as a StartUP Growth Magazine initiative and part of AGAZ Summit 2026. #StartupGrowthAwards #StartupWomen #WomenEntrepreneurs #FemaleFounders #WomenInBusiness #CanadianStartups
-
Mark Ruddock liked thisMark Ruddock liked thisJoin our welcoming national Research and Information team, collaborate with colleagues across the Firm, and help shape the future of our Print and eResources services. Interested in learning more? We'd love to hear from you!
-
Mark Ruddock liked thisTracking this. Very cool.Mark Ruddock liked thisImagine being able to fly through 51 million lines of code. Developer Rik Arends built a 3D source-code visualizer that turns massive software projects into something that looks more like a 3D world that you can explore than a codebase. In this demo, he loads Chromium - the open-source project behind Google Chrome and several other browsers. The scale is incredible: roughly 51 million lines of code, built up over 18 years! Instead of trying to load all of that into memory at once, the visualizer uses indexing, caching and streaming to make navigating a codebase this large possible. It’s a fascinating way to visualize just how enormous modern software has become. Rik Arends is now building a product around this concept – which is now in private beta. Follow him for updates.
-
Mark Ruddock liked thisMark Ruddock liked thisAnn Handley said it best last week in a LinkedIn post, mentioning that, lately, it feels like writing is getting a bit lost ("the art of writing feels reduced to an inefficiency to be solved"). You can have the prettiest website, the most curated ABM list, and the cleanest reporting, but it won't get you very far if the words you're using to communicate with your audience are falling on deaf ears. Don't neglect the copy. Talk like real people do. Read user reviews. Fold those words into your 30-character headlines. Run tests against the jargon-y ones. Start small, and keep it rooted in the customer. Writing is not an inefficiency to be solved— in many ways, it *is* the resolve.
Experience & Education
Other similar profiles
-
Lorraine McGregor
Lorraine McGregor
When growing or transitioning a business, operational friction flourishes when there is a new strategy, goals, or initiatives. <br><br>You say 'let's go' and then everyone interprets how the changes should work. This interpretation is informal, full of assumptions and doubt. Friction descends. The expected changes don't materialize. <br><br>The buzz word for this problem is 'alignment'. But the word is irrelevant without understanding what's going on under the surface. <br><br>Someone has to be responsible for what should change in messaging, execution, workflow, alignment and agreements. Not defined? People fill in the blanks and default to their habits.<br><br>It starts with clarity on why: What you will be able to do then that the company can't do now. At all levels. This is our definition of how to find alignment.<br><br>I help mid-market business owners identify and fix these underlying friction points that stall growth, drain profitability, make managing miserable and risk devaluation during a sale, investment or exit transition.<br><br>Who I Work With & When to Call Me<br>Which red flags do you see:<br>1. Stalled Growth / Plateau: Revenue has flattened despite increasing effort, management is overwhelmed. This tells me the new strategy was not turned into a project so managers are trying to deal with people issues and goals that never seem to be reached.<br><br>2. Unbilled, Discounting or Leaking Revenue: Operations are inefficient, data doesn't tell you where the problem is and the company is keeping less of every dollar. This tells me there are hidden profit opportunities or under-billing for value.<br><br>3. Leadership or Operational Misalignment: Team members are pulling in different directions, causing friction in execution. This tells me the workflow and authority is not balanced for new workloads. <br><br>4. Preparing for Exit or Sale: The business relies too heavily on the founder and is not saleable. This tells me the valuation expectation is not realistic because it wasn't baked into the business, it's just expected.<br><br>How Spirit West Management Helps<br>Using proven diagnostic methodologies, we quickly pinpoint the root causes of operational strain, not just the surface symptoms, and implement clear, actionable fixes.<br><br>The result: A predictable, highly efficient operation that runs smoothly, maximizes profitability, and operates independently of the founder.<br><br>Let’s Connect: Navigating growth bottlenecks? Are you an advisor with a client facing operational friction? Connect here on LinkedIn or visit spiritwest.com.
Explore more posts
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content