&#34;Why has 'lobster farming' become so popular? What industries will 'AI lobster' disrupt? How can it be commercially monetized?
According to a research report on Chinese securities released by JP Morgan on March 12, analysts Xu Wentao and Yao Cheng recently communicated with the project manager of KNOWLEDGE ATLAS AutoClaw, deeply analyzing the reasons behind the popularity of products like AutoClaw and OpenClaw, as well as the paths and logic for subsequent application implementation and commercial monetization.
JP Morgan believes that &#34;the importance of products like AutoClaw and OpenClaw is not because they have suddenly made autonomous AI commercially mature, but because they have significantly lowered the threshold for non-technical users to experience intelligent agent workflows.&#34;
For the market, the most core impact is: although the adoption of intelligent agents is expected to expand model usage and infrastructure demand, short-term monetization is still in its early stages. Actual deployment will first land in relatively structured workflows, rather than widespread complete autonomous human replacement.
<h2>The Explosion of &#34;Lobster&#34; Intelligent Agents: A Victory of Product Design Rather Than Model Mutation</h2>
The recent craze for OpenClaw-like products, is it a leap in model capability or an optimization in product design? The interviewees provided a clear answer. This popularity &#34;reflects improvements in product design and usability, rather than a mutation in model intelligence.&#34;
Interviewees particularly emphasized three key factors: &#34;Integration with existing communication tools, allowing agents to build user profiles over time with persistent memory, and broader system permissions that expand the actual working scope of agents.&#34;
JP Morgan pointed out that this distinction is crucial. The current craze is driven by better productization and accessibility, which means that user engagement can be expanded before achieving true enterprise-level monetization.
<h2>Quality of Base Models: The Core Determinant of Commercial Limits</h2>
In today's world where intelligent agents are emerging one after another, will base models be commoditized? One clear point from the discussion is that &#34;the commercial limits of intelligent agents still largely depend on the quality of the base models.&#34;
Interviewees repeatedly emphasized, &#34;Intelligent agents are essentially a container or medium; the model remains the core factor determining whether tasks can be completed accurately and consistently, and whether the depth of reasoning is sufficient to function in high-value contexts.&#34;
JP Morgan believes this refutes the view that the intelligent agent layer will completely commoditize the model layer in the short term. &#34;Better models should still translate into better task completion, stronger instruction adherence, more stable long-context performance, and superior handling of open workflows.&#34; Therefore, the increase in adoption rates of intelligent agents remains a significant benefit for leading model suppliers.
<h2>Commercial Monetization: Still in Early Stages, Focused on Structured Tasks</h2>
Despite the popularity of the intelligent agent concept, the interviewees' statements imply that &#34;the intelligent agent market may still lean towards the exploratory stage without full monetization in the short term.&#34; Current products are still in the phase of helping users discover usage scenarios. To significantly expand in commercial scenarios, &#34;it may still require 6 to 12 months of model improvements, workflow training data, and product iterations.&#34;
JPMorgan Chase believes this aligns with the current state of enterprise-level AI. &#34;Coding and technical workflows remain the clearest early monetization paths, as tasks are structured, target functions are clearer, and execution trajectories are easier to define.&#34; Beyond coding, the lack of standardized &#34;trajectory&#34; data is a significant factor limiting agents' ability to execute multi-step tasks in the real world.
<h2>Market Landing: Technical Engineering and Structured Workflows Lead the Way</h2>
In the process of going to market, which areas will adopt agents first? Respondents emphasized three major categories:
<blockquote>
<ul>
<li>
Technical Engineering: &#34;Expanding from coding to testing, deployment, operation, configuration, and debugging. This seems to be the most credible category in today's business.&#34;
</li>
<li>
Information and Content Workflows: &#34;Including research, report generation, document processing, office document operations, and internal content production.&#34;
</li>
<li>
Personal Productivity: &#34;Such as email, calendar, and message management.&#34; While appealing to consumers, converting this into sustained monetization may take longer.
</li>
</ul>
</blockquote>
JPMorgan Chase advises investors to &#34;focus short-term expectations on technical and structured enterprise tasks, rather than making overly aggressive inferences from a consumer trial perspective.&#34;
<h2>Open Architecture and Moat: Rapidly Replicable Features Are Not Important</h2>
Another important point in the discussion is that the agent layer may not be a winner-takes-all proprietary model channel. AutoClaw supports multiple model providers, and management clearly supports an open architecture rather than mandating the exclusive use of KNOWLEDGE ATLAS models.
JPMorgan Chase believes this broadens the potential market for the product and enhances the opportunity for the agent platform to become a model ecosystem aggregator. However, for model providers, this means &#34;the agent interface itself may not guarantee exclusive downstream value capture unless the supplier also leads in model performance, agent tool invocation, and workflow integration.&#34;
Regarding the moat, management believes that feature comparison is less important, as many visible features can be quickly replicated.
They argue that true defensive capability is built on three aspects: &#34;speed of product insights, quality of foundational models, and accumulated agent functionalities (such as browser tools, memory systems, and workflow processing).&#34;
JPMorgan Chase agrees with this view, pointing out that investors should focus on whether suppliers can &#34;consistently improve task completion rates over time, reduce friction, and integrate data to enhance agent performance.&#34;
<h2>Restructuring the Industry Chain: Who Benefits and Who Gets Disrupted?</h2>
Wider adoption of agents will benefit multiple parts of the AI stack:
<blockquote>
<ul>
<li>
First, broader adoption of agents should benefit model suppliers, as more autonomous workflows mean more token consumption and more sustained usage.
</li>
<li>
Second, it should benefit reasoning infrastructure, cloud, and related computing providers, especially as demand continues to outstrip supply.
</li>
<li>
Third, collaborative and workflow platforms that offer open APIs and allow controlled integration can also benefit by becoming interfaces that embed agents into daily work
On the contrary, the value proposition for &#34;shallow intermediaries or low-threshold information processing companies&#34; may face risks. For roles or services with &#34;limited moats, public information, and workflows that are relatively easy to automate, AI is more likely to exert pressure.&#34;
</li>
</ul>
</blockquote>
In addition, security and regulation are practical constraints for corporate deployment. Management believes that &#34;prompt injection, permission errors, malicious third-party skills, and software vulnerabilities are actual constraints.&#34; This may slow down monetization in the short term but will increase the importance of trusted vendors and compliance-level architectures.
JPMorgan Chase has given KNOWLEDGE ATLAS an &#34;overweight&#34; rating, with a target price of HKD 800 by December 2026, based on a 30 times expected price-to-earnings ratio for 2030, and expects a compound annual growth rate of over 100% for its revenue from 2026 to 2030.
<pre><code></code></pre>

512820

512880

513750

The dialogue between both parties indicates that the recent explosion of the "Lobster" AI intelligence is due to the optimization of product design and tool integration, rather than a mutation in the underlying model intelligence. Currently, commercialization is still in its early stages, and the most certain money in the short term lies in structured technical tasks such as "coding," rather than C-end personal assistants. In this reshaping, the underlying models and computing infrastructure will reap the benefits, while "shallow information intermediaries" lacking data barriers are on the brink of being the first to be reshuffled

- J.P. Morgan's research highlights the success of AI products like AutoClaw and OpenClaw, emphasizing their ability to lower barriers for non-technical users.  
- The report notes that the impact of agent adoption is still in early monetization stages, with initial deployments focused on structured workflows.  
- Investors are advised to concentrate on technical and structured enterprise tasks while acknowledging potential risks in low-barrier information processing sectors.

Wallstreetcn

WP CSI Banks ETF

Guotai CSI All Share Investment Banking & Brokerage ETF

GF Fund CSI HK Connect Financials(ex Banks) Thematic ETF

Yinhua CSI All Share Investment Banking & Brokerage ETF

Hwabao WP CSI All Share Investment Banking & Brokerage ETF

Hwabao WP CSI Banks ETF

ChinaAMC CSI Banks ETF

KNOWLEDGE ATLAS

Fullgoal CSI 800 Banks ETF

Huaan CSI Banks ETF

JPMorgan Chase

JPMorgan Chase Dialogue with KNOWLEDGE ATLAS AutoClaw: Why did the intelligent agents explode, is model quality still important, and how to achieve commercial monetization?