TL;DR

“2026 Is the Year of the Builder.” That line from Tomasz Stanczak's opening keynote at ETHDenver has been echoing in my head since I walked out of the main hall. It cut straight through the noise around institutions, ETFs, L2 wars, and "nothing new left to build" — and reframed where the real opportunity is right now.

We wanted to capture the key ideas he shared, both for my own reflection and for fellow founders, hackers, and tinkerers who might be feeling that same mix of fatigue and possibility.

From 2025's Institutional Meta to 2026's Builder Meta

Tomasz described 2025 as the "year of the institutions":

Stablecoins increasingly dominated by large entities and governments
Ethereum and infra narrative captured by big players
The pervasive feeling: "We're late. The game is locked up. There's nothing truly new to build."

Many of us heard (or said) some version of that last year.

His counterpoint: that phase was necessary. 2025 was the "key battle" for institutional adoption — but that doesn't mean the creative frontier is closed. It just means the meta has changed.

In 2026, that meta is shifting back to builders.

AI as the New Expansion Pack for Web3

The most powerful framing from the talk: AI is not just a new vertical; it's an expansion pack for everything we tried (and often failed) to build in the previous crypto cycles.

We've had DeFi, NFTs, DAOs, governance experiments, on-chain games, social tokens, and ZK tech slowly maturing in the background. A lot of those visions didn't fully work. Some were too early, some broke on UX, some died on coordination and governance, some hit hard technical limits.

AI — especially agentic AI — lets us revisit those "failed" or incomplete ideas with an entirely new set of tools.

Not in a superficial "add AI to your deck" way, but at the level of:

On-chain agents making commitments and following through
Agentic treasuries and DAOs that are operationally sustainable
UX where users don't fight with wallets, dapps, and block explorers, but delegate to agents that speak both human and blockchain

AI becomes the "expansion pack" that:

Makes some previously impossible things actually feasible
Forces us to redesign infrastructure, security, and UX from first principles

Agentic Experience (AX): Beyond Traditional UI

One of the most interesting ideas: agentic experience (AX) as the next UI paradigm.

In a world of agents:

You don't manually sign every transaction in a clunky wallet
You don't click through 7 dapps to close one position or rebalance one strategy
You don't need to know which chain, which L2, which bridge

Instead:

You express goals and constraints to an agent
The agent interacts with the blockchain stack on your behalf
The "UI" is effectively a conversation plus a set of guardrails, not a dashboard full of buttons

That shift breaks a lot of assumptions we've been building around:

Account models and key management
RPC and infra patterns
Logging, auditability, and observability
Security models (what does "approval" even mean when an agent can act semi-autonomously?)

And that's where Tomasz sees huge open space for builders.

What's Actually on the Table to Build in 2026?

He outlined several categories where there is genuine greenfield:

1. Agentic AI & Agent Infrastructure

Secure environments for running agents that can touch real assets. Tools for permissioning, policy, and oversight of agents. Persistent on-chain identities and reputations for agents. Coordination between multiple agents across chains and services.

2. Security for Agentic Interactions

Safe delegation frameworks (what can my agent do? what can't it do?). Monitoring and anomaly detection for agents acting on-chain. "Airbags" for smart contracts and accounts controlled by agents.

3. AX Infrastructure

New account architectures designed for agents, not humans clicking buttons. Logging/analytics geared toward agent behavior and decision traces. "Agent-native" wallets and key management solutions.

4. ZK and Privacy Layers

ZK-powered L2s and appchains that are actually production-ready and usable. Privacy-preserving interactions that agents can leverage programmatically. Lighter-weight ZK systems tailored for agent workflows.

5. Hybrid TradFi–DeFi & Institutional Bridges

Flows where agents can traverse both regulated and permissionless environments. Composable products that combine institutional rails with open protocols.

This isn't just about shipping "an AI agent." It's about the stack that must exist underneath for agentic systems to be safe, scalable, and trustworthy.

Why Solo Builders Have an Edge Right Now

A core theme: this moment is unusually good for individual builders and small teams.

Why?

Institutions are structurally constrained. They're locked into roadmaps, incentives, compliance, and legacy bets. They can't pivot fast enough into this new agentic meta without dragging a lot of baggage.

The tools have radically leveled the playing field. With today's LLMs and creative tools, a solo builder can:

Prototype product flows in hours
Generate design, copy, content, and experiments cheaply
Test multiple directions without massive upfront investment

Speed matters more than ever. With everyone using the same foundational models, advantage increasingly comes from:

How fast you can turn ideas into working experiments
How deeply you understand the problem and users
How automated your own build pipeline is

You may need to be able to go from idea → shipped experiment in minutes to hours for certain opportunities.

That implies:

Automating your own development workflows
Maintaining a library of reusable components you can assemble quickly
Having agents that help you build (code scaffolding, infra setup, tests, docs, etc.)

In other words, build the tools that help you build — meta-building as a core skill.

"Looking at the Pyramids"

He used a powerful historical analogy: imagine standing near the pyramids 6,000 years ago.

You might think:

"This is it. Only giants can build things that matter now. It's all centralized forever."

But history shows the opposite: We've had millennia of breakthrough builders and inventions since then — many started by "solo" individuals or tiny teams who didn't accept that everything worth building was already built.

We've had that same "it's all taken" feeling:

10 years ago (post-mobile/social)
20 years ago (post-dot-com)
50+ years ago (post-mainframe/telecom)

Every time, the pattern repeats: the landscape looks consolidated, then a new substrate or meta appears (PC, web, mobile, cloud, crypto, AI), and suddenly the frontier is wide open again.

Tomasz's argument is that AI + crypto in 2026 is exactly such a frontier.

So What Do We Do With This?

A few personal takeaways that resonated with me:

→ Go back to the ideas that "didn't work" 3–5 years ago and ask: What changes if I assume competent agents exist?

→ Stop over-focusing on "getting into the right institution" or "partnering with the right giant" and instead ask: What can I build in the next 90 days that only a small, fast team can ship?

→ Don't just bolt AI onto existing Web3 flows. Redesign the UX, security, and infra as if agents are first-class users of the blockchain.

→ Invest in your own speed: Automate build/deploy as much as possible. Use AI not just as a product feature, but as a co-builder. Maintain "ready to ship" building blocks so you can move when an opportunity pops.

Treat 2026 as Your Window to Build

Tomasz closed with a simple call: treat 2026 as your window to build.

Not when it's all "figured out." Not when institutions have issued the RFPs and the market maps are stable. But now, when the security models are still fuzzy, the agent patterns are emergent, and the playbook doesn't exist yet.

If you've been on the sidelines thinking "it's all too late" — this might actually be your best entry point.

I'm curious:

→ What "old" ideas are you re-examining in light of AI agents?

→ Where do you see the biggest gaps in agentic security and infrastructure?

→ Are you building something in this space right now?

Would love to connect with others exploring this "year of the builder" mindset — especially at the intersection of agents, Web3, and new forms of coordination.

“2026 Is the Year of the Builder” - An Ethereum Perspective