We used large models to "divine" the tokenization/Web3 landscape for 2026

For the same set of data, do you trust human interpretation or AI "divination"?

It sounds like a cyberpunk joke, but in today's world overflowing with opinions and high emotionality, what the market truly lacks is precisely those unvarnished, real sample pieces of information that haven't had time to be polished.

After all, positions can really make people lie.

As the year turns, to get as close as possible to the truth of tokenization/Web3 at this inflection point of 2025-2026, we conducted an anonymous sampling targeting the frontline Builder community via Maicai Tong MSX, focusing primarily on three questions:

• Keywords for 2025 US stocks / tokenization in your eyes;
• Your actual holdings in 2025 (not limited to US stocks / Web3);
• The directions you are most willing to allocate to in 2026 (not limited to US stocks / Web3);

After some "coercion and inducement," we collected a sample size that was somewhat satisfactory and compressed it into three word cloud images with almost no subjective judgment mixed in. Precisely considering that human interpretation inevitably carries bias, we decided to invite the "pure" ChatGPT, Gemini, and DeepSeek, under the exact same Prompt constraints, to "cast a divination" for these three word clouds.

Perhaps, AI truly understands this generation of Builders better than we do ourselves, and is more capable of piecing together the fate puzzle of US stocks, Web3, and tokenization in 2026 from these fragments.

1. What are the keywords for 2025 US stocks / tokenization in your eyes?

Prompt: You will see a set of keyword samples from frontline practitioners in US stock tokenization / Web3. These keywords are used to answer the same question: "What does the US stock / tokenization field look like in your eyes for 2025?"; Do not judge right or wrong, and do not introduce any external market information. Based solely on these keywords themselves, interpret them from the perspective of data structure and linguistic features, focusing on answering:

Which keywords have the highest/lowest/most abrupt frequency, and what structural state might this reflect? Which keywords or semantic directions show obvious consensus characteristics? Are there combinations of keywords with similar meanings, opposite directions, or strong tension? What stage characteristics does the overall set of keywords suggest the sector is in? If based solely on these keywords without introducing any external information, how would you understand this sector's 2025 and 2026?

Please maintain a descriptive, restrained analytical tone, avoid giving investment advice or definitive judgments, and keep the reply under 1000 words.

Editor's Note: The replies have been slightly organized and polished for readability, but the overall direction and thinking have not been significantly adjusted (same below).

ChatGPT: The highest-frequency keywords are concentrated on "acceleration" and "institutionalization," suggesting the sector is undergoing a process of significantly changing pace that is not yet stable; Medium-frequency keywords include "AI," "compliance," "differentiation," reflecting a background, normalized presence; Low-frequency or isolated words like "weightlessness," "unknown," "contested territory," "starting point," carry strong subjective feelings or emotional metaphors, possibly reflecting individuals' discomfort with structural changes, sense of competition, or direct projection of uncertainty.

The overall corpus presents linguistic features of a "mid-early stage—structural building phase," where high-frequency verbs outnumber nouns, indicating the discussion focus is still on process and change, not stable forms; Multiple keywords point to rules, verification, processes, rather than specific products or models; Emotional vocabulary exists but does not dominate, indicating uncertainty is perceived but not out of control.

If based solely on these keywords, 2025 can be understood as a transitional year of acceleration without final form, where consensus is gathering but differentiation is simultaneously occurring.

Gemini: High-frequency words include "acceleration," "institutionalization," indicating a strong consensus on growth driven by professional forces; Medium-frequency words "AI," "compliance," "differentiation" outline the state of technological means, regulatory frameworks, and uneven market performance; More abrupt low-frequency words include "weightlessness," "process-oriented," "asset verification," reflecting underlying anxieties beneath the surface of prosperity, and a micro-perspective shifting from narrative to implementation.

The overall data presents a state of "high momentum, low center of gravity," where the coexistence of "institutionalization" and "compliance" means this acceleration is not blind expansion but a structural shift constrained by rules, also forming the sector's certain tone.

Considering the overall keyword characteristics, the sector is currently in a "transitional explosive period," combining the chaos of the starting stage (weightlessness, unknown) with the rigor on the eve of maturity (compliance, process-oriented). The market is no longer monolithic but shows obvious "differentiation," indicating tokenization is no longer an isolated asset experiment but beginning to deeply couple with mainstream finance and cutting-edge technology.

Overall, from 2025 to 2026, this is a moment of professional entry, running anxiously amidst rules, reshaping, and heavily reliant on technological verification.

DeepSeek: "Acceleration" occupies the absolute core, with "institutionalization," "differentiation," "compliance," "AI" forming the second tier; Most words appear only once, showing discrete long-tail characteristics. For example, "weightlessness" appears relatively abrupt in the overall context of dynamic, positive vocabulary (like acceleration, opening up, breaking circles), possibly metaphorizing a state of temporarily losing traditional reference points amidst rapid change.

Among them, "institutionalization" and "compliance" are closely related in meaning, jointly pointing to the trend of deep involvement by traditional financial system rules and participants; "Differentiation" suggests possible separation of hierarchies or paths within the market or sector.

The sector's developmental dynamism (acceleration, variability, volatility) and institutionalization process (institutionalization, compliance, process-oriented, structured) constitute the two strongest consensus semantic directions.

Their parallel existence depicts a field seeking to establish rules and structure amidst rapid change.

2. Your actual holdings in 2025 (not limited to US stocks / Web3)

Prompt: You will see a set of keyword samples from frontline practitioners in US stock tokenization / Web3. These keywords are used to answer the same question: "Your actual holdings in 2025 (not limited to US stocks / Web3)"; Do not judge right or wrong, and do not introduce any external market information. Based solely on the sample composition, conduct logical analysis and interpret from a data structure perspective:

Categorize the involved assets/sectors; Are there core assets that are repeatedly allocated? Is the overall portfolio more concentrated or diversified, revealing what kind of risk tolerance? What behavioral characteristics can be observed from these actual holdings?

Please maintain a descriptive, restrained analytical tone, avoid giving investment advice or definitive judgments, and keep the reply under 1000 words.

ChatGPT: From the sample structure, holdings can be clearly divided into several asset clusters.

• Crypto-native assets: High-frequency words include BTC, ETH, SOL, BNB, USDT, USDC, with secondary/edge assets like XRP, DOGE, ASTER, SUPRA, XTTA; This category holds a clear advantage in quantity and repetition, containing both mainstream and exploratory allocations;
• US tech and finance-related stocks: Includes tech/platform types—GOOGL, AAPL, AMZN, NVDA, TSLA, PLTR, INTC, TSM, MU, WDC; also includes finance/trading related—COIN, HOOD; Shows a preference for infrastructure, platforms, and technological capabilities;
• US consumer and traditional defensive stocks: MCD, KO, LLY, JD, etc., few in number but clearly serving structural hedging or stabilizing roles;
• Indices and cash-like/hedge assets: VOO, gold, USDT, USDC, etc., providing liquidity and volatility buffers within the overall portfolio;

Judging by repetition frequency, there is a very clear core asset layer:

• BTC is the single asset mentioned most frequently in the sample, appearing multiple times densely;
• GOOGL has the highest repetition rate on the US stock side, significantly higher than other tech stocks;
• ETH, SOL, BNB form the second layer of repeatedly allocated crypto assets;
• USDT, as a non-risk asset, is also mentioned multiple times, showing it is not a "transitional presence";

This structure indicates respondents are not holding evenly distributed portfolios but repeatedly stacking on a few high-conviction assets while maintaining diversified attempts on the periphery.

If based solely on this set of actual holding keywords without introducing any external judgment, it can be understood as a portfolio structure anchored by a few core assets, with parallel allocations across multiple asset systems, while preserving liquidity and exploratory space.

Gemini: Through preliminary sorting of the sample data, these holding assets can be divided into the following five categories.

Data structure shows this group exhibits extremely significant consensus centralization. BTC is the most frequently mentioned asset, followed by GOOGL, ETH, and SOL. These four constitute the "four-corner anchors" of practitioners' 2025 holdings.

The high-frequency appearance of BNB, TSLA, and USDT also reflects dual considerations for industry ecosystem status and liquidity while pursuing high Beta returns.

Despite diverse assets involved, about 50% of the sample capacity is occupied by the top 5-6 core assets. This indicates practitioners' 2025 strategy was not a scattergun allocation but highly concentrated bets on blue-chip assets with strong narrative support.

Specifically, the risk tolerance approach can be summarized as a barbell strategy: one end is extremely high-volatility tech growth and crypto assets, like RKLB (aerospace), SOL, and various semiconductor assets; the other end is extremely high liquidity or certainty, like USDT, gold, VOO (S&P 500 ETF).

Although traditional value stocks like MCD (McDonald's) and KO (Coca-Cola) appear in the sample, they are merely decorative. This indicates in the 2025 holding logic, the status of traditional dividend-paying assets has been marginalized. Practitioners are more inclined to define "value return" as "tech growth supported by cash flow."

DeepSeek: Holding assets in the sample can be roughly categorized into three types:

• Cryptocurrencies: Including BTC, ETH, SOL, BNB, XRP, DOGE, etc. These assets appear with relatively high frequency, covering mainstream and some altcoins;
• Stablecoins and Commodities: Stablecoins like USDT and USDC, commodities like gold. These assets are typically related to low-price-volatility hedging or liquidity management;
• US stocks span multiple industries and market caps: Including tech stocks—GOOGL (Google), AAPL (Apple), TSLA (Tesla), NVDA (Nvidia), COIN (Coinbase), PLTR (Palantir), INTC (Intel), MU (Micron), TSM (TSMC), etc.; Consumer and traditional industry stocks: like MCD (McDonald's), KO (Coca-Cola), LLY (Eli Lilly), etc.; Finance and platform stocks: like HOOD (Robinhood), FIG (possibly a financial group), etc.; ETF: VOO (S&P 500 Index ETF);

Core assets (BTC, GOOGL, ETH, SOL) account for about 37% of total mentions, giving the holding structure a "core concentrated, periphery dispersed" characteristic, meaning heavier allocation to core assets while diversifying into numerous others.

Overall, the sample data structure reveals a configuration mindset seeking balance between risk and opportunity, anchoring core sectors while building buffers through diversification and hedging tools.

3. The directions you are most willing to allocate to in 2026 (not limited to US stocks / Web3)

Prompt: You will see a set of keyword samples from frontline practitioners in US stock tokenization / Web3. These keywords are used to answer the same question: "The directions you are most willing to allocate to in 2026 (not limited to US stocks / Web3)"; Do not judge right or wrong, and do not introduce any external market information. Based solely on the sample composition, conduct logical analysis and interpret from a data structure perspective:

Categorize the involved assets/sectors; Which sectors/assets do these investors prefer more in 2026, and are there core assets that are repeatedly allocated? Is the overall portfolio more concentrated or diversified, revealing what kind of risk tolerance? What behavioral characteristics can be observed from these actual holdings? What are the assets/sectors with the highest consensus in 2025 and 2026 respectively, and what changes occurred?

Please maintain a descriptive, restrained analytical tone, avoid giving investment advice or definitive judgments, and keep the reply under 1000 words.

ChatGPT: Compared to 2025, the 2026 keywords show a clear change: specific assets and abstract sectors begin to appear side-by-side, indicating respondents emphasize directionality and structural judgment more than single asset selection when expressing allocation intentions.

The sample can be roughly divided into the following categories:

• Crypto-native assets: BTC, ETH, SOL are core high-frequency words, while stablecoins and MEME crypto assets still occupy the sample center, but internal stratification is clear.
• US tech and high-tech directions: At the individual stock level, focus is on TSLA, GOOGL, NVDA, AMZN, PLTR, HOOD; at the sector level, focus is on tech stocks, chips, storage, aerospace. The characteristic of transitioning from "stock picking" to "industry direction" is very obvious;
• Commodities and physical assets: Precious metals (gold, silver, copper), oil. This category, not prominent in the 2025 sample, is systematically mentioned in 2026 intentions;
• Indices and hedging tools: S&P, SQQQ represent attention to overall market or systemic volatility;

Judging by repetition, the 2026 core is very concentrated: BTC remains the highest-frequency keyword, far above others; TSLA, NVDA are repeatedly mentioned on the US stock side, significantly higher than other individual stocks.

If based solely on these keywords, 2026 allocation intentions can be understood as a structural allocation tendency centered around a few highly consensual core assets and industry directions, while using sectoral, resource-based, and tool-based expressions to cope with uncertainty.

Gemini: The 2026 sample data shows a broader span than 2025, mainly归纳为以下四类:

BTC is mentioned in over half of the 2026 samples, the only asset with full-group consensus characteristics; TSLA and RKLB/aerospace follow closely. Compared to 2025, practitioners' allocation willingness for hardcore tech and outer space exploration significantly increases, indicating investment vision is extending from the digital to the extreme frontiers of the physical world.

The portfolio structure shows a "one-pole-multi-strong" dumbbell structure: one pole is the extremely concentrated BTC, the other is the extremely dispersed hard tech assets (involving aerospace, storage, semiconductors, and other细分领域).

Risk tolerance manifests as "extreme aggressiveness coexisting with extreme defensiveness." On one hand, high-volatility assets like MEME, RKLB are配置; on the other, hedging tools like SQQQ, gold, silver, copper, oil are introduced. This indicates practitioners in 2026 anticipate not a unilateral rise but structural opportunities amidst剧烈波动. Their risk tolerance has shifted from单纯追逐增长 to managing uncertainty.

Overall, in the 2026 keywords, the density of gold, silver, copper, oil, storage, aerospace大幅增加, reflecting that after深耕数字化赛道 (Web3), practitioners' logic is beginning to gravitate towards tangible assets and underlying energy/resources.

Deep industry chain挖掘 is also a typical feature. Respondents are no longer satisfied仅配置头部芯片股 like NVDA but extend downwards to storage and specific hardware links like MU, WDC, SNDK, showing a专业化深挖 behavior based on industry knowledge.

If 2025 was practitioners "building roads" (focusing on compliance, platforms, and infrastructure), then 2026 is an "expedition" (focusing on aerospace, deep hardware, and global大宗资源).

DeepSeek: From word frequency statistics, BTC has the highest frequency, accounting for 20.4% of total mentions, with TSLA (9.3%)紧随其后. The two合计约 30%, showing relative concentration in top preferences. NVDA, RKLB, GOOGL, HOOD, stablecoins, precious metals are also mentioned多次, but overall still分散于众多标的和主题,呈现 "core-periphery" structure—core highly concentrated in少数资产, periphery widely分散.

Among them, cryptocurrencies (especially BTC) and tech stocks (TSLA, NVDA, etc.) are the main directions, while thematic sectors like aerospace (represented by RKLB), chips, storage, and避险资产 like precious metals,