ETHDenver 2026 Observation: Side Events Cut Short, AI Boom, Builders Searching for New Paradigms

Author | Kyle Wu on Blockchain

Side Events Cut Short

Let’s start with the most straightforward data: the number of side events this year has significantly dropped. According to data from Luma, compared to 668 side events in 2025, this year’s conference (February 18–21, 2026) has only about 215, a 68% decrease.

There are several reasons behind this. One is the industry cycle itself; another is that the event dates are close to the Lunar New Year, which is not friendly for Asian project teams and attendees; additionally, external events like the Lakehouse Manor WLFI Forum have also diverted industry attention, drawing away some OGs and core participants.

So, attendees still feel that ETH Denver has many North American Web3 folks gathering, but the number of people coming from outside the US has dropped significantly, weakening the overall spillover influence of the conference.

From the activity level within specific ecosystems, it’s clear that the industry has entered a phase focused more on efficiency and cost-saving.

Monad and X Layer have a strong presence—Monad hosted three events, and X Layer sponsored the main stage (though their team only attended the first day of the expo). Aptos and Sui also had some exposure; Solana hosted only one small-scale event with good quality. Overall, many public chain ecosystems are no longer pursuing the previous approach of numerous activities and overwhelming presence, but instead are maintaining a few side events, hosting one or two activities just for show, emphasizing cost-effectiveness.

AI Everywhere

One of the biggest changes this time is that AI is almost everywhere.

Besides the main stage, ETH Denver split the main venue into five stages—ETHERSpace (applications and user ownership), Devtopia (infrastructure and developer tools), New France Village (finance/compliance/ institutionalization), Futurllama (AI / DePIN and other frontier trends), and Prosperia (privacy, public goods, and community). After three days of exploring, Futurllama had the highest foot traffic.

There was also a summit called Claws Out for the trending “Lobster” OpenClaw project, and small roadshows for AI project founders. Outside the venue, Sentient hosted the Open AGI Summit, focused on AI, which even drew more people than some main areas. This ETH Denver no longer feels like a traditional crypto conference; it’s more like a combined AI × Crypto expo.

Even the project forms have changed. Seeing robots and robotic arms at the expo gives a “CES” vibe. PrismaX, Gensyn, and other projects focusing on embodied intelligence have their own booths. Many projects, although still tagged Web3, are no longer solely about public chains, DeFi, or wallet infrastructure. Instead, they are moving toward agents, chatbots, and other AI product applications.

A strategic manager at an exchange told me that at this conference, they’re not focusing on mature projects but on very early-stage or even just embryonic ideas. Regarding AI + Crypto, they’re not looking for immediate revenue but whether the combination can create new production relationships, distribution methods, workflows, and on-chain collaboration mechanisms.

He also shared an internal strategic view: everyone already expects AI to bring huge productivity gains, but in Web3, the real value isn’t in large models but in embedding AI tech into exchanges to improve current product experiences. They are developing large language models integrated into exchanges that can give trading recommendations based on real-time market news and enable direct trading within chat interfaces.

ETH Denver Still Builder-Oriented

Despite AI being the hottest topic, ETH Denver remains one of the most builder-focused crypto conferences.

This is evident from the agenda. On the last day, aside from the closing ceremony, the entire event is dedicated to hackathons and builder workshops. Side events during the main chain events, like those for Base, are also developer-centric.

Notably, during the conference, Base showcased a small product called Braindate. They set aside a dedicated area in their booth for social chatting. Unlike traditional casual chat zones, you can join or create a “session” on Braindate to connect with like-minded people.

Inside the venue, this atmosphere is even more apparent: compared to discussions about market trends, most attendees are talking about “what they’re building lately.” Especially students and builders, who care more about “can I make this” or “is this product interesting,” rather than short-term market fluctuations. Because of this, although ETH Denver’s overall scale has shrunk, it hasn’t become a purely narrative-driven, non-product conference. At its core, genuine builders still matter.

The best reflection of this is the hackathon, BUIDLathon. This year, it has a clear change: the competition format is more front-loaded.

Based on participant feedback, the BUIDLathon now includes an online hacking phase, with topics announced a week in advance, moving away from the traditional “start building on-site” approach.

This is an interesting innovation. It aligns better with how many builders work today and can improve project completion rates. However, it also shortens in-person interaction time—from last year’s 8 days down to 4—requiring much of the preparation to be done beforehand. Offline is no longer the starting point for building but more like the final phase of project polishing.

This change in format may simply reflect the organizers’ organizational preferences, but the prize pool’s drastic reduction reveals the industry’s current cycle.

After interviewing multiple participants, they all complained that sponsors’ numbers and funding力度 are much lower than previous years. Last year’s total prize pool was $1.03 million, but this year it shrank to $132,000. Sponsor budgets are more concentrated and focused on their main interests: AI.

One participant mentioned that a winning project in the hackathon was themed “AI Girlfriend.” Users can tip the AI girlfriend with crypto, creating a behavioral incentive mechanism through frequent interactions, thus bringing the two closer. It sounds like a joke, but the fact that it won shows that judges are no longer only looking at wallet and settlement scenarios but want projects to depict more relatable, mainstream use cases.

I also spoke with Justin, who won third place with Base sponsorship. His project is an open protocol for AI agent advertising monetization.

The idea is that many future free AI products for the public might need to be monetized via ads rather than subscriptions. Advertisers pay for promotion tasks, AI agents recommend and drive conversions based on user context, and earn revenue after completing tasks.

The biggest challenge is proving that AI truly completes the ad tasks.

Justin’s solution involves on-chain validators to verify ad results. In their demo, they mainly validated verifiable on-chain actions like purchases: only when the validator confirms the delivery as genuine does the protocol trigger settlement, and the advertiser’s budget is paid to the agent. This means the project isn’t just an “AI ad platform,” but an infrastructure for future AI-native advertising markets, providing verification and settlement.

Another notable point about this hackathon is the diversity of participants.

It includes university students, seasoned builders, industry veterans, and innovative creators. Projects span AI, DeFi, and also interesting GameFi ideas.

ETH Denver’s hackathon hasn’t become a game exclusive to crypto insiders. Many newcomers are participating, unburdened by the need to focus solely on traditional crypto scenarios, and are more willing to combine AI, gaming, advertising, and social scenes with on-chain tech naturally.

Rough and Ready Judging Process

Of course, this hackathon isn’t perfect, and the judging process feels somewhat “rough.”

Participants reported that the judging is divided into main track and sponsor track. The main track is clearer: queues start at 9 am, usually facing 2–3 judges, and the pitch and demo are not very stressful. But the five-minute presentation time favors teams that can quickly explain their project and showcase a demo. Under time constraints, humor and memorability often matter more than technical polish, and bugs don’t necessarily disqualify winners.

After the main track, teams must also present to sponsors like Base, which is less organized—first-come, first-served, with less clear management, testing teams’ improvisation and execution.

Market Predictions

Besides AI, another theme worth highlighting is prediction markets.

During the Frontier Markets session hosted by Monad, I heard some key points from the discussion. First, liquidity remains the biggest barrier. Second, since each market has an end time, liquidity migrates and doesn’t accumulate like perpetual contracts. Third, attracting liquidity providers for long-tail markets is still a tough problem.

Meanwhile, prediction markets differ from traditional trading. Traditional markets tend to fluctuate gradually, but prediction markets can suddenly zero out at a specific point, making leverage, market-making, and risk management more complex. Many large traditional market makers are hesitant to enter.

However, prediction markets have their own uniqueness: popular markets often attract retail liquidity providers. In other words, success depends on continuously creating engaging markets that draw retail participation and reorganizing liquidity around these markets to improve user experience.

Crypto Resilience

To conclude on ETH Denver, I’d say it reflects the industry’s state during a bear market and transition phase. It’s not a hot, trend-driven, hype-fueled event anymore. The excitement and market buzz are diminished, and the so-called “economic boom” feels absent. But at the same time, a core group of builders are still exploring seriously, some investors are still watching early-stage projects with potential, and some projects, despite imperfections and bugs, are beginning to reveal the next viable business models and outlines for the crypto space.