Each year, we publish our annual "Juxtapositions"—observable tensions that sharpen our thinking and fuel inspiration for the year ahead. As we surveyed the team at the end of last year, we kept returning to the architectural concept of "tensegrity" as an analog to what we are seeing in the company-building landscape today.
In the mid-20th century, sculptor Kenneth Snelson created works exploring what his mentor Buckminster Fuller would name "tensegrity." "Tensegrity" refers to a quality of structures that derive strength not from mass or rigidity, but from the balanced interplay of tension and compression. Snelson's Needle Tower in Washington, D.C., exemplifies this concept: a form held in equilibrium by opposing forces. The concept revolutionized architecture, enabling designers to build large structures around omnidirectional tensions rather than simple stacking, leveraging the inherent strength of core materials.
As our firm prepared for the start of this year, we began to see tensions that exemplified this “tensegrity” at work, not just in the rapidly evolving markets and shifting technology paradigms, but in the fundamental craft of company creation itself. This year's Juxtapositions shift our lens inward, examining the structural tensions within our own business model: the how, what, and why behind Juxtapose.
Building companies in 2026 demands designing organizations that not only tackle and withstand the tensions of rapid change, but also grow stronger because of it. This year's list outlines the structural dynamics we believe will meaningfully shape company building across capital markets, AI transformation, ecosystem partnerships, tech-enabled services, and regulation.
We hope you enjoy!
Dynamic #1: More Money Meets Less Need
Companies today can raise more capital than ever while also operating more efficiently than ever.
Founders have access to unprecedented funding, yet AI-powered automation and operational transformation allows teams to achieve more with less. The tension lies in knowing where to place bets: building capital-efficient companies that reach defensible scale without relying on massive infusions, while knowing how to leverage real capital when it can capture market leadership in a high return-on-capital manner.
AI can replace human tasks with greater speed, at lower cost, and often with higher quality, but making AI work inside an organization increasingly depends on people.
AI automates work, but not adoption. The bottleneck for impact is shifting from technical capability to organizational capability: aligning incentives, redesigning workflows, retraining teams, and restructuring decision rights. Companies treating AI as a piecemeal "tool upgrade" will struggle, while those treating it as a behavioral and cultural shift will accelerate.
Dynamic #2: Technical Velocity Meets Organizational Reality
The speed of AI transformation rewards rapid bets where adoption is measurable, while some sectors still demand patience and rigor.
Many organizations overestimate how quickly AI can be integrated into complex operations and underestimate the friction of change. But evidence is emerging that AI can deliver outsized impact in targeted domains like systems integration and managed services, where AI is already improving scheduling, maintenance, workflow optimization, and predictive analytics. When AI is applied in data-rich, workflow-integrated, and domain-informed contexts, the returns can be real and scalable.
AI tempts teams to build from scratch, yet the biggest wins often come from wrapping intelligence around the machine you already have.
Some workflows demand a full reset of data models, decisions, and interfaces. Others benefit from lighter-touch improvements, like summaries, recommendations, or automation layered onto what already works. Being early doesn't mean AI must always lead. Innovators will win by designing product and data architecture so that when AI takes center stage, the distribution, workflows, and behavioral norms AI runs on already work.
Dynamic #3: Company Creation Meets Ecosystem Transformation
Building independently lets you chase the moonshot, but the greatest leverage often comes from embedding into a customer’s operations.
Standalone company creation offers strategic freedom, but true leverage is unlocked once a product integrates into live workflows and gains access to proprietary data, daily usage, trust, and continuous feedback. That embedded position may limit flexibility and broaden commitments in the short-term, but it sharply accelerates learning for the longer-term. You quickly discover whether AI creates measurable economic value or just theoretical performance. The tensions are between optionality and proof, and between the freedom to imagine at scale and the discipline of proving impact in the field.
AI companies are building unprecedented capabilities in-house, but they are increasingly dependent on a web of specialized vendors to reach the next wave of adoption.
As well-funded AI leaders mature, they're forming technical and commercial partnerships with specialized vendors, often in unexpected combinations. Over the next year, the AI landscape will be defined less by standalone companies and more by the vendor networks that enable them, creating an ecosystem where collaboration is as critical as raw capability. Specialized demand is too great for any single vendor to satisfy—don’t be surprised by strange bedfellows in 2026!
Dynamic #4: Challenger Velocity Meets Incumbent Advantage
All companies are now technology companies, but a key way to monetize tech-enablement is through full-stack service delivery—actually delivering the product or service to customers, not just providing the technology.
Software and AI have become table stakes rather than differentiators. As tools grow cheaper and easier to replicate, standalone tech margins will continue to compress. Economic value will increasingly concentrate at the point of execution, where the actual service reaches the customer. This drives a shift from "tech that enables work" to "companies that do the work." Full-stack models assume more operational complexity and risk, but they also unlock control over outcomes, data, and unit economics. In today's market, technology is the entry ticket; operating the service is where enduring value is created.
As barriers to innovation decline, the value of being an incumbent increases while the pace of change simultaneously advantage new companies.
Intelligent systems feed on real-world data, operational scale, and hard-to-access environments, which favors established players. But AI also automates away the routines incumbents have long relied on as barriers, like complex coordination, expert judgment, manual processes, and proprietary know-how. In this new AI economy, incumbents are becoming essential nodes while challengers are becoming the essential innovators. The big winners will be the ones who turn incumbency from an anchor into acceleration.
Dynamic #5: Regulation Tailwinds Meet Regulation Headwinds
Regulatory uncertainty poses real risk for AI builders while simultaneously creating a rare window for rapid experimentation and scale.
AI-native companies are scaling at an unprecedented pace as the regulatory ground continues to shift beneath us. AI remains lightly regulated federally, but that won't last. Over the next two to three years, expect rapid layering of state and federal laws and regulations that will shape what can be built, how it's deployed, and where risk sits. Even as recent federal actions signal attempts to centralize or preempt state oversight, several states are moving ahead with targeted AI laws focused on high-risk use cases, including discrimination in hiring, consumer harm, and mental health applications. The result is a growing mismatch between the velocity of company creation and the pace and fragmentation of legal clarity. For founders and builders, the question isn't whether regulation is coming, but whether companies being created today will be structurally nimble enough to adapt when it does.
AI is producing content faster than humans can manage, and the only way to keep up is with AI itself.
Generative models produce text, images, code, and video at scales no human team could review. The solution is emerging in real time: "LLMs as judges," or AI systems trained to evaluate, filter, and moderate other AI outputs. These AI judges will span industries, threat types, and modalities to enforce standards, ensure safety, and flag risks. The tension is clear: the force overwhelming human oversight is simultaneously becoming the only tool capable of controlling it.
