Hello and welcome to another edition of Lateral Connections.
Started reading Stewart Brand’s “Maintenance: Of Everything, Part One” this week. Fascinating - pound for pound, his books have changed my thinking more than most (I started listing other authors who have had the same effect and soon the list became too long 😀).
But that is for another post - the opening chapter brought to mind the difference between Roald Amundsen and Ernest Shackleton 1 and that difference is a worthy reflection on leadership. I’ve found that we usually celebrate the 'heroic save.' We love a Shackleton: the leader who drags a team back from the brink of disaster through sheer force of will. But if you look at the history of Antarctic exploration, I find myself increasingly drawn to Roald Amundsen. Amundsen didn’t have a 'heroic' story because he planned the drama right out of the mission. This week, I want to explore why both the Architect and the Crisis Manager have a seat at the table, and how to know which one you need to be in a given circumstance.
The Explorer’s Matrix: When to Lead Like Amundsen vs. Shackleton
In the world of complex systems, we often mistake "activity" for "progress." By looking at how these two men managed their expeditions, we can categorize our own leadership challenges.
The Amundsen Playbook: Precision in the "Steady State"
Amundsen’s success was built on Domain Humility. He didn't assume he knew best; he looked at who was already surviving in the Arctic (the Inuit) and copied their "tech stack" (dogs and furs).
The Lesson: "Boring" is a feature, not a bug. Amundsen famously said, "Victory awaits those who have everything in order – people call that luck. Defeat is certain for those who have forgotten to take the necessary precautions in time – that is called bad luck."
Key Behavior: Over-specifying your "supply depots." If you’re migrating a database, Amundsen-style leadership means having triple-redundancy and markers so clear you could find them in a "data blizzard." Amundsen knew that finding a single flag in a white-out was a matter of luck. To remove luck from the equation, when he laid his primary supply depots, he didn't just plant one flag. He placed 20 black flags in a line extending several miles on either side of the depot.
The Shackleton Playbook: Resilience in the "Pivot"
Shackleton’s genius was Psychological Safety. When the Endurance sank, the mission was dead. He didn't cling to the original requirements document; he threw it away to save the team. ” We have been compelled to abandon the ship, which is crushed beyond all hope of ever being righted, we are alive and well, and we have stores and equipment for the task that lies before us. The task is to reach land with all the members of the Expedition. It is hard to write what I feel”. To build a cohesive unit, Shackleton eliminated the hierarchy inherent in expeditions of that type in that era.
The Lesson: In a crisis, the people matters more than the project and he gave them a new North Star immediately. This lesson is relevant during major system outages, sudden budget cuts, or "Black Swan" events that render your 5-year plan obsolete.
Key Behavior: When the Endurance was sinking and Shackleton ordered the men to abandon ship, he gave a strict limit on weight: two pounds per man. He famously threw his own gold watch and some gold coins into the snow to show that status and wealth were now "dead weight." However, he insisted that the meteorologist, Leonard Hussey, keep his heavy 12-pound banjo. Shackleton called it "vital mental medicine." He understood that in a crisis, "irrational" morale-boosters are more functional than "rational" tools. Just as Shackleton kept the grumblers in his own tent to manage their influence, a leader in a crisis stays closest to the points of highest friction.
A brief History
Amundsen who had originally planned to go to the North Pole secretly pivoted south when he heard the Americans had beat him to it. He didn’t even tell his crew (or the King of Norway) until they were already at sea.He chose a shorter but steeper route than his rival, Robert Falcon Scott. He relied entirely on Greenland Huskies (dogs) and skis - Scott and his team tried a mix of ponies, motor sledges (that failed because he overlooked maintenance 😀), and dogs. Amundsen moved with surgical efficiency. He treated the journey like a logistics problem. He used the dogs not just for transport, but as a fresh food source (a grim but practical "bio-fuel" strategy) to prevent scurvy. He reached the Pole on December 14, 1911, planted the Norwegian flag, and returned to his base camp safely. He was so well-prepared that his team actually gained weight during the trip, Scott and his team died 11 miles short of their depot which held vital supplies.
Shackleton with his Imperial Trans-Antarctic Expedition—the first land crossing of the Antarctic continent from sea to sea - wanted to claim the last worthy Antarctic exploration challenge. But before they even reached the continent, their ship, the Endurance, became trapped in pack ice in the Weddell Sea. Even though he failed in his main objective it ended being celebrated as one of the most remarkable stories of endurance.
They remained stuck for 10 months as the ice slowly crushed the hull. In October 1915, the ship sank, leaving 28 men stranded on floating ice floes with limited supplies and three small lifeboats. Shackleton pivoted to anew mission: "All men must survive."
After months of camping on drifting ice, they took the lifeboats to the uninhabited Elephant Island. From there, Shackleton and five others sailed a 22-foot boat (the James Caird) 800 miles across the world’s most dangerous ocean to South Georgia Island. After landing, Shackleton had to trek across uncharted mountains and glaciers to reach a whaling station. Every single member of the 28-man crew was eventually rescued. Though the mission failed entirely, it became the greatest survival story in history.
This week in AI & Data
I guess this is the week that Clawdbot to OpenClaw transformation reached peak hype. More on that below. But this particular news caught my eye. “In this paper, we introduce PaperBanana, an agentic framework designed to bridge this (…Among these illustration tasks, generating methodology diagrams represents a significant challenge, demanding both content fidelity and visual aesthetics) gap by automating the production of high-quality academic illustrations. Given a methodology description and diagram caption as input, PaperBanana orchestrates specialized agents powered by state-ofthe-art VLMs and image generation models (e.g. Gemini-3-Pro and Nano-Banana-Pro) to retrieve reference examples, devise detailed plans for content and style, render images, and iteratively refine via self-critique. This reference-driven collaborative workflow allows the system to effectively master the l’ogical composition and stylistic norms required for publication-ready illustrations. Beyond methodology diagrams, our framework demonstrates significant versatility by extending to statistical plots, offering a comprehensive solution for scientific visualization”. One by one steps that required substantial labor are being transformed at a frenetic pace. The examples given in the paper are extraordinary.
And the OpenClaw hype is being hailed as the “Jarvis” moment. The cybersecurity guy in me shudders at how far back we are in catching up to these advancements. I will quell my anxiety and plunge into this new era. OpenClaw made has members of the AI community hyperventilating and debating its novelty and importance. For developers, OpenClaw offers a highly customizable and powerful AI assistant (albeit with aforementioned cyber caveats) that can operate with little input from humans.
This week in Cyber.
As usual there is the expected crop of CISA directives and KEVs. Amongst them this one hopefully is not an issue for too many agencies.
These two items caught my attention.
Cyberwire daily indicated that the Trump administration is preparing a major overhaul of US cybersecurity policy, led by National Cyber Director Harry Coker Jr. with a strong emphasis on private sector collaboration and regulatory reform. Anything which targets security theater and focuses on actual risks would be welcome news.
Google has usually stayed out of the news, and so this one about Looker caught me by surprise. Researchers disclosed multiple critical vulnerabilities in Google Looker that could allow for cross-tenant data theft. With AI enabled cyber exploits on the rise, the cloud providers have to really up their game or else we are in deep waters. Mercifully there is some good news as well from Anthropic - “When we pointed Opus 4.6 at some of the most well-tested codebases (projects that have had fuzzers running against them for years, accumulating millions of hours of CPU time), Opus 4.6 found high-severity vulnerabilities, some that had gone undetected for decades.”
Till next time
Stay safe and dream.