The State of AI x Web3 from ETH SF

Where Web2 and Web3 AI Converge -– Abdul Al Ali, Venture Investor at Decasonic

ETH SF 2025 reinforced one thing:reinforced one thing: innovation at the intersection of AI and Web3 isn’t just continuing—it’s accelerating at an unprecedented pace. The differentiator among projects isn’t just vision; it’s execution and the speed at which teams bring ideas to life. My conversations with builders were centered around utility, adaptability, adoption, and real-world applications. It seems that the evolution of AI x Web3 is widely accepted, and the core builders of this intersection are accelerating its adoption.

San Francisco remains the epicenter of this movement, bringing together Web2 and Web3 AI founders in a way that no other city does. The energy here is unparalleled—it’s a builder’s city, fueled by ambition, resilience, and an unrelenting drive to deliver utility and innovation. It’s why founders move here, why top talent flocks to the Bay, and why some of the most transformative breakthroughs in AI and Web3 are happening right here—including the home bases of many AI x Web3 companies.

Beyond ETH SF itself, hackathons and side events hosted by Near Protocol, Sui, Base, and others were packed with developers, builders, and teams. The focus wasn’t just on experimentation; it was on building products designed for longevity and impact. Nearly every conversation, panel, and late-night discussion revolved around how AI is reshaping Web3 and how founders can move from concept to widespread adoption.

Here are my 12 key takeaways from ETH SF 2025:

1. Builder Energy is Unmatched

ETH SF was all about building with intent. Founders weren’t just experimenting—they were solving real problems and pushing technical boundaries. The energy wasn’t just about innovation but about making AI x Web3 applications scalable, widely adopted, and enterprise-ready. The era of simple prototypes is behind us—this was a showcase of products built to last.

2. AI x Web3 is King

AI wasn’t just a theme at ETH SF—it was the driving force. Nearly every conversation, project, and panel revolved around use cases for AI. AI is no longer just an “emerging trend” in crypto—it’s the catalyst to mainstream adoption, unlocking entirely new applications and economic models at the intersection.

3. Web3 AI Founders Are Thinking Beyond Web3

Rather than just porting Web2 AI models onto blockchain rails, founders are redefining what’s possible. Many are building platforms that don’t just mimic Web2 but actively onboard Web2 AI users and enterprises into crypto. The opportunity is bigger than Web3 alone—founders are thinking globally and cross-industry, with AI as the bridge to crypto adoption.

4. SF is Cementing Itself as the AI x Web3 Capital

San Francisco and Silicon Valley continue to be the gravitational center for AI x Web3 innovation. The density of top-tier AI talent, blockchain expertise, and venture capital makes it a relatively unbeatable launchpad to be in. Founders are relocating here not just for the network, but for the relentless work ethic and the sheer pace of innovation. Some of the founders I spoke with are making SF their new home.

5. Blockchains Are Competing for AI Founders

Major blockchain ecosystems are in an arms race to attract AI-native founders. Some are more aggressive than others, but AI-focused hackathons, grants, and incentive programs are a clear priority. Expect continued competition among chains as they refine their tooling, incentives, and go-to-market strategies to secure AI builders on their chains.

6. Hot Take: Compute’s Relevance Sustains

AI’s insatiable demand for compute remains a defining factor, regardless of the increasing model cost efficiency. The demand for compute will persist—and likely accelerate. Cost efficiencies won’t slow innovation; they’ll fuel even greater experimentation and adoption.

7. Physical AI Infrastructure is Entering the Conversation

Inspired by Jensen Huang’s January 2025 keynote, the conversation at ETH SF extended beyond code—toward hardware, robotics, and decentralized infrastructure. AI-powered robotics, edge computing, and autonomous systems are emerging as frontier opportunities for Web3. The future of AI in Web3 isn’t just about digital agents, it's about bringing those agents to life in the real-world through the next phase of AI evolution, Physical AI.

8. AI Agents Need a Hybrid Web2/Web3 Workload

For AI agents in Web3 to be truly effective, they can’t be confined to on-chain execution. Seamless integration across Web2 and Web3 environments is key. Innovations from Manus and Open AI's Agent Kit, and other frameworks are making this possible—enabling AI agents to interact via APIs, blockchain transactions, and hybrid workloads. At Decasonic, we expect AI Agent launchpads and frameworks to increasingly integrate the latest Web2 advancements. The ultimate differentiator will be the quality of these agents—with potential moats forming around personalized data, optimized workflows, and cost efficiency.

9. Model Layer is Increasingly Commoditized

AI models are getting cheaper, faster, and more efficient. As costs continue to fall, value accrual in AI x Web3 will shift towards the application and interface layers. The biggest winners won’t be those deploying generic AI models, but those building sticky AI-driven applications.

10. Hackathons Demonstrate Passion

Beyond the Stanford events, hackathons from NEAR Protocol, Sui Network, Base, and other L1s were buzzing with AI x Web3 builders. The sheer volume of early-stage projects emerging from these events underscores the competition among chains to attract AI-native founders. One standout event was with the Celo and Base community hackathons—watching it live, the energy around AI adoption was energizing.

11. Increasing Web2/Web3 Collaboration

Rather than positioning Web2 AI companies competitors, Web3 x AI founders are collaborating with legacy AI companies. This is why many are relocating to SF and Silicon Valley—to be at the epicenter of AI innovation, learn from Web2 leaders, and apply key learnings in their execution.

12. The Web3 AI Supercycle Accelerates

The shift from infrastructure to applications and interfaces is accelerating, and AI will be the catalyst for mainstream adoption. Crypto-native AI applications—whether in gaming, social, finance, or infrastructure—are driving the next future crypto cycles.

For deeper insights into our 2025 strategic outlook and exclusive market maps showcasing the top projects shaping AI x Web3, explore our latest published articles below:

Where Does AI x Web3 Go from Here?

ETH SF reinforced that the AI x Web3 wave is accelerating - it’s not going anywhere. Founders are pushing past speculation, focusing on usability, and solving problems that extend beyond crypto’s current user base. The focus of the user base for many of the founders is not exclusively limited to Web3 users and projects - but extends beyond that to Web2 users, companies, and institutions. The future of AI is on blockchain rails.

Decasonic’s Role in AI x Web3

Decasonic is at the forefront of investing in the AI x Web3 frontier, backing founders who are actively building and advancing innovation at this intersection.

If you're building at the convergence of AI and Web3 reach out to us. We’re actively partnering with founders to help advance your vision and turn it into reality.

The content of this material is strictly for informational and educational purposes only. It is not intended to constitute investment advice, nor should it be considered a recommendation or a solicitation to buy, sell, or hold any asset. Decasonic does not endorse investments in any specific tokens, and nothing in these blog posts should be construed as legal, tax, or financial advice. Please consult with a qualified professional advisor before making any financial decisions. Decasonic provides no warranties, whether expressed or implied, on the content provided in these blog posts, including its accuracy, completeness, or correctness. The opinions expressed here are those of the authors and do not necessarily reflect the views of Decasonic. Please note that Decasonic may hold a position in some of the tokens mentioned, including Virtuals. Decasonic is not liable for any errors or omissions in the content of this material or for any actions taken based on the information provided herein.