Money Frontier 2026 countdown: 7 days left — a preview of the highlights of the summit

Money Frontier 2026 (held in Hong Kong at the Grand Hyatt Hotel from July 27 to 28) will bring together industry leaders, leading platforms, investors, policy researchers, and front-line operators, sharing real observations and hands-on practical experience from the market, products, and the industry.

The summit will not focus on the macro trends of Web3 and AI; instead, it will concentrate on specific changes that have already happened and are actively reshaping the industry—helping attendees see the new market structure clearly, understand how opportunities form and where risks are hidden, and build actionable judgment.

When public markets and the crypto market accelerate their convergence, why is the cost of capital price even more inconsistent? When traditional high-return opportunities gradually disappear, where can investors still find opportunities with manageable risk? When AI leaders are already difficult to participate in or are overvalued, have ordinary investors truly missed this round of opportunities?

A G E N D A H I G H L I G H T S — F I R S T L O O K A T T H E K E Y P O I N T S

Highlight 1: Latest regulatory policy updates

U.S. strategic Bitcoin reserves—what happens next?

Bitcoin Policy Institute, which has long participated in U.S. Bitcoin policy research and advocacy, will share the latest developments on the U.S. strategic Bitcoin reserve and related policies. The institute has long conducted Bitcoin research and public policy advocacy for policymakers, and has continued to publish research related to strategic reserves.

What’s worth关注 is not only whether the Bitcoin bill will pass, but also the institutional pathway through which strategic reserves will be advanced, where the assets will come from, how Congress and the executive branch will coordinate, and how these policy changes may affect institutional allocation, market liquidity, and digital asset policies in other countries. This session will help participants see through headlines and understand the real pace of policy advancement, key obstacles, and their potential market impacts.

Hong Kong lawmakers will also share views on the future direction of digital assets in Hong Kong.

Highlight 2: How to understand the new “new market structure”

With DeFi and CeFi continuing to develop, digital asset treasury companies (DAT), RWA, asset tokenization, and new asset issuance platforms are accelerating the connection between public markets and the crypto market.

But the connection is not unifying the markets; it is instead amplifying liquidity fragmentation. Because different platforms differ in funding costs, access requirements, collateral rules, redemption mechanisms, trading hours, and jurisdictions, the same asset may have different interest rates, prices, and liquidity across different platforms—creating differences between digital assets and traditional financial markets.

The Bank for International Settlements notes that RWA and tokenisation may exacerbate market fragmentation and raise financing costs; research from the Federal Reserve also shows that cross-market pricing frictions and market segmentation still exist between digital currency markets and traditional financial markets.

Therefore, capital allocation is no longer just about choosing a platform and comparing visible returns; it is about identifying frictions and boundaries between different markets, understanding how different product structures are formed, judging whether the spread comes from market access, liquidity, credit, leverage, and subsidies—or from risks that have not been clearly labeled yet—then managing funding costs, redemptions, custody, counterparty risk, smart contract risk, and regulatory risk accordingly.

On July 27, founders and CEOs of multiple leading protocols such as Ethena and Spark (MakerDAO) will deeply break down their product structures, sources of yield, and platform operating mechanisms. Strive CEO and Chairman will also share, around STRC, SATA, and other new Bitcoin-based credit products, the product structures, yield logic, and potential risks.

Senior executives from multiple traditional financial institutions and the digital asset industry will also bring front-line observations and unique judgments from different market and business perspectives.

Highlight 3: When AI makes everyone a target worth attacking

As AI moves from concept to reality, security problems no longer belong only to large institutions, trading platforms, or high-net-worth groups.

Identity, accounts, assets, communication records, and social relationships can all become entry points for attacks. With more low-cost, scalable, and highly personalized attack methods, do traditional security habits still work?

The summit will discuss how AI changes attackers’ ways of selecting targets, obtaining information, and carrying out attacks, and how individuals and institutions should re-examine identity verification, asset custody, device permissions, and internal security processes.

The core issue is not just “whether AI is dangerous,” but: when the cost of attack drops rapidly, how should we raise the attacker’s cost?

Highlight 4: How to find opportunities and growth paths amid the AI wave?

Worried you might have already missed the AI wave? Actually, maybe not. Better AI investment entry points may not be before technological breakthroughs, but when technology has been validated by the market and rapid growth begins to expose industry bottlenecks.

For most investors, opportunities may not come from placing big bets on the next AI application in advance. Spending a large amount of time and capital to judge an unverified technology often means taking risks where you don’t have an advantage.

A more realistic path is to wait until the technology or trend completes market validation and enters a phase of rapid expansion, then look for the industry bottlenecks exposed during growth. As demand rises quickly, key resources such as GPUs, memory, data centers, and electricity often cannot expand in sync, leading to supply-demand imbalance.

Compared with predicting who will become the next winner, those links that have already been validated by real demand but are constrained by supply capacity may provide ordinary investors with clearer bases for judgment and ways to participate.

Which shortages are merely temporary cyclical mismatches? Which bottlenecks might last for years? Which assets sit in the AI industry chain but cannot truly share in the industry’s growth? And how do we identify the key links that have pricing power, expansion barriers, and real customer demand?

You’ll get answers to all of these on the July 27 agenda.

Several well-known investors, including Xingkong, will share their investment experience and judgment frameworks in early-stage markets for frontier technologies; XianDao will also bring《Second Life Practical Handbook》, sharing how to turn judgments about new trends into executable personal choices and practices.

Day 2: From GPU to electricity—dissecting the end-to-end AI compute infrastructure industry chain

Highlight 1: Domestic chips accelerate into intelligent computing data centers—new industry windows are opening

As large-model inference and industry AI applications accelerate into real deployment, competition in domestic computing power is no longer limited to a single technical route. How can China’s domestic general-purpose GPUs further move into intelligent computing data centers and real business scenarios? How can China’s advantages across the industrial chain enable a “curveball takeover” that reshapes the market landscape dominated by overseas GPU giants? For ASICs optimized for specific AI tasks, can they achieve new breakthroughs in performance, efficiency, cost, and scalable deployments? The summit will invite representatives from Moore Threads, as well as Dr. Yang Zuoxing, founder of Engineered Extreme Electronics, who has long focused on domestic AI ASIC R&D and launched its in-house “Shenmu” brand. Starting from different technical paths, they will share domestic intelligent computing chip R&D progress, deployment practices, and directions for industrialization, jointly observing the new opportunities emerging as the local computing power ecosystem opens up.

Highlight 2: Domestic open-source large models lower the innovation barrier—Agents move toward real applications

With the release of high-performance open-source large models such as ChatGLM 5.2 and Kimi K3, more and more enterprises can directly obtain model capabilities close to cutting-edge levels. Models are no longer just proprietary resources of a few top companies; competition in the AI industry is also shifting from “who can train bigger models” to “who can build truly usable products and systems based on models.”

This also opens a new development window for Agents. When model capability becomes a foundational capability that can be called upon, how can enterprises further connect data, tools, and business processes? Which Agents have moved beyond concept demos and entered real production scenarios? Which applications have the possibility of sustained demand, paid capability, and scalable replication? And how can a closed loop form among model capability, compute costs, and business models?

The summit will discuss topics such as “Agent wave: comprehensive reshaping from technological evolution to a new industrial era” and “From model to Agent: deployment of AI-native applications, paths for compute and capitalization,” inviting guests from Tencent Cloud, BytePlus Hong Kong, KUAI.CLOUD, as well as AI startups and investment institutions. They will share real cases and development directions of Agents, observing the new AI application ecosystem being fostered by domestic open-source models from multiple dimensions including technological evolution, product deployment, compute support, and commercialization paths.

Highlight 3: From data center going overseas to AI Factory—engineering capability becomes the core moat

An AI data center is not simply a matter of adding GPUs in traditional server rooms. As power density per rack continues to rise, the power supply and distribution, cooling, network interconnection, equipment deployment, and operations and maintenance systems all need to be redesigned. Whether land, electricity, equipment, and operational capabilities can be organized into a stable, efficient, and sustainably scalable system is becoming the true engineering barrier for the compute industry.

Going overseas with data centers also brings more complex real-world issues: how to choose parks and power conditions suitable for AI workloads? How to control construction cycles and delivery costs? What differences exist across markets in infrastructure, supply chains, and operating environments? What upgrades must traditional data centers complete in order to truly evolve into an “AI Factory” for AI training and inference?

Around topics such as “Data center going overseas” and “Beyond traditional data centers: the rise of AI Factory and intelligent compute,” companies including canaan, Chincode Intelligence, Skyward Digital, JDK Capital, and Goodvision AI, along with industry guests such as the head of a special task force under the Korean Presidential AI National Strategy Commission, will use front-line project experience to break down the key know-how in planning, building, going overseas, and operating data centers, and discuss which experiences can be replicated and which technologies and construction risks are easiest to overlook. The agenda will also focus on core areas such as power reliability, rack power density, cooling systems, network interconnection, and operational capabilities.

Highlight 4: From POW to AI—value boundaries of electricity resources are being redefined

If GPUs determine compute performance and data centers determine how compute is carried, then electricity constitutes the lowest-level resource constraint of the entire compute system. As AI compute demand continues to grow, AI data centers and POW are competing for the same batch of scarce resources—stable electricity with pricing advantages, locations suitable for deploying high-density equipment, and power contracts that can lock in long-term costs.”

In this context, electricity is no longer just an operating cost for data centers; it may also become a strategic asset that needs to be independently allocated, operated, and re-valued. Can different types of compute workloads switch flexibly based on market demand? Can existing POW infrastructure further support AI computation? How can the compute value created per megawatt of electricity resources be improved? What new forms of assets and business models might emerge around electricity, park-and-load scheduling?

The summit will extend from existing compute industry experience through topics such as “From POW to AI” and “What problems does POW solve? What is the next problem?” to further discuss efficiency in allocating electricity resources, revenue elasticity, and future opportunities. Whether existing AI compute service or token distribution mechanisms can organize resources efficiently, transparently, and at scale—like mining pools allocating compute—still needs to be further validated by the industry. KuPool will also start from the problems already solved by POW to explore the next set of new propositions the compute industry must face: when chips and models keep iterating, what truly determines the industry’s expansion boundary may be not only electricity resources themselves, but also the capabilities of organization, scheduling, and compute trading.

For more details, please visit:

For attendance and partnership inquiries, please contact the summit staff.