Editor’s note: Besides tracking technological advancements and innovations, our author is a Juilliard-trained musical composer. Listen to “Accelerated Capabilities”, an original improvisation by Howard Lieberman, composed for this column.
This is my 100th column in The Berkshire Edge. For the last ninety-nine Mondays, I have written about innovation as it unfolds in real time. Not as a buzzword or a strategy deck abstraction, but as a lived process shaped by tools, incentives, institutions, and human behavior. Looking back across 2024 to 2025, one pattern stands out clearly. Innovation did not slow down. It fractured.
Capabilities accelerated. Power consolidated. Barriers to entry fell. Permission structures hardened. Innovation became simultaneously more available and more constrained, depending on where you stood and what you were trying to do.
This is not a contradiction. It is the defining tension of the moment.
What got better
The most important change of the last two years is not that artificial intelligence arrived. It is that high-level capability became cheap, fast, and widely accessible.
Writers can draft, revise, and explore ideas at a pace that once required a newsroom. Designers can prototype without a team. Musicians can test arrangements, orchestrations, and production ideas without booking studio time. Inventors can model, simulate, and communicate concepts before committing serious capital. The tools did not replace thinking. They amplified it. They shortened feedback loops and lowered the cost of exploration. That matters because innovation is rarely about having the right answer immediately. It is about being able to try more paths before resources run out.
There are early signs that productivity is responding. Recent data from the U.S. Bureau of Labor Statistics show labor productivity rising through 2024 and into 2025 as output grew faster than hours worked, reversing much of the stagnation seen earlier in the decade.
International comparisons tell a similar story. The Organisation for Economic Co-operation and Development (OECD), a consortium of advanced economies that tracks productivity and growth trends, shows the United States outperforming many peer nations in labor productivity growth, even as global growth remains uneven. See OECD productivity indicators here.
At a more human scale, this shift has been visible to me locally. Over the past two years, I helped found iiCoCo, an international composers’ collaborative built around short works, peer feedback, and rapid iteration rather than institutional commissioning or gatekeeping. It exists largely because the coordination, communication, and creative tooling now make that kind of distributed collaboration practical.
Research spending in the United States also remained massive in absolute terms, with business leading much of the investment. But equally important is how much creative and intellectual work now happens outside traditional research institutions, simply because the tools allow it.
Most importantly for creative outliers, the barrier to entry continued to fall. Publishing, distribution, collaboration, and audience access all became more direct. The gatekeeping layers thinned. You no longer need permission to begin. You only need the discipline to continue.
What got worse
If capability democratized, decision-making did not.
Capital became cautious and concentrated. Venture funding did not disappear, but it grew far more selective. Investment flowed into fewer companies and larger rounds, particularly in AI-related sectors, while early-stage access tightened.
That concentration is even clearer in artificial intelligence. Reporting across venture markets shows AI startups capturing a majority share of venture capital investment in early 2025, reinforcing winner-take-most dynamics.
This tension between individual capability and institutional consolidation also shows up outside venture capital. I rebooted the Silicon Valley International Innovation Institute (SVIII) during this same period, not as a physical institution but as a distributed network. The move was less philosophical than practical. The old assumptions about where innovation had to live no longer held.
Consolidation pressures also remained real. Antitrust enforcement moved from theory to practice. Heightened scrutiny of serial acquisitions and market concentration is reshaping exit expectations across multiple industries.
Research funding showed strain more quietly. While total R&D spending stayed high, federal basic research funding fluctuated year to year, even as competitors increased long-term investment. Multiple science and technology indicators show slower U.S. growth than China across several key research areas.
Add to this a political environment that treats innovation as rhetoric rather than infrastructure, and the picture sharpens. Innovation thrives on continuity. Uncertainty taxes it.
Howard Lieberman created this image with AI assistance.
The forecast
Looking ahead, the weather looks mixed but navigable.
Artificial intelligence will remain central, but the conversation will shift from novelty to governance. Questions about training data, creator rights, attribution, and compensation are moving from theory into practice. Markets are beginning to formalize how creative work is used, tracked, and compensated in AI systems. Those mechanisms will increasingly shape who benefits from the next phase of AI adoption.
At the same time, innovation is becoming more local and more relational. In the Berkshires, I helped launch the Berkshires Culture Salon, a monthly gathering that brings writers, musicians, artists, and technologists together to share unfinished work. Nothing about it depends on scale. Everything about it depends on participation.
Venture capital will remain active but disciplined. Fewer narrative-driven bets. More demand for real traction. This favors founders who understand customers deeply and can articulate value without theatrics.
Meanwhile, creative leverage will continue to spread. Even as institutions consolidate, individuals gain power through tools. That is the paradox of the moment. Centralization at the top. Decentralization at the edges.
Innovation will increasingly come from people and small groups who do not wait to be chosen. Writers who publish directly. Musicians who distribute globally. Inventors who prototype before asking for permission. These are not rebels. They are adapters.
The lesson of 2024 to 2025 is not that innovation is in crisis. It is that innovation has changed shape. The terrain is rougher. The tools are better. The skies are unsettled.
Those who mistake friction for failure will retreat. Those who understand friction as information will adjust their course.
Innovation has always favored the latter.
