Which AI Stocks Likely To Benefit From Rise Of Large Concept Model (LCM)?

LLMs have revolutionized the field of artificial intelligence and have emerged as the de facto tool for many tasks. The current established technology of LLMs is to process input and generate output at the token level. This contrasts sharply with humans who operate at multiple levels of abstraction, well beyond single words, to analyze information and generate creative content.

Large Concept Model (LCM), substantially differs from current LLMs in two aspects: 1) all modeling is performed in a high-dimensional embedding space instead of on a discrete token representation, and 2) modeling is not instantiated in a particular language or modality, but at a higher semantic and abstract level.

Large Concept Models (LCMs) represent an exciting new frontier in AI, moving beyond the token-by-token prediction of Large Language Models (LLMs) to understand and reason with higher-level semantic "concepts." This shift promises more coherent, logical, and adaptable AI, with potential benefits across various industries.

In this article I would like to share the breakdown that I have worked on for the types of AI stocks likely to benefit from the rise of LCMs:

Core AI Research & Development (Companies actively developing LCMs)

$Meta Platforms, Inc.(META)$ : Meta AI is a pioneer in LCM research, having published significant papers on the topic and developing models like those based on their SONAR embedding space. They are at the forefront of this architectural shift, and their advancements will directly benefit their internal products and potentially be licensed or open-sourced.

Meta have shown that it is highly reliant on machine learning for its ads product, and it has reaped the obvious benefit of applying generative AI to advertising, an advertiser can buy ads based on desired outcomes, whether that be an app install or a purchase, and leave everything else up to Meta.

So with LCM, this would help Meta which will work across their vast troves of data that make the using machine learning-derived algorithms faster to find the right targets for an ad and deliver exactly the business goals requested.

Other Major AI Labs (Google/ $Alphabet(GOOGL)$, $Microsoft(MSFT)$, OpenAI, Anthropic): While Meta has been prominent in publicizing LCM research, it's highly probable that other leading AI research labs are also investing heavily in similar concept-based AI architectures. Their ability to develop and deploy these models will be key. Keep an eye on their research papers and product announcements.

Companies Providing Infrastructure for Advanced AI (Hardware & Cloud)

LCMs, like LLMs, will require significant computational resources.

$NVIDIA(NVDA)$ : As the dominant provider of GPUs essential for training and running large AI models, NVIDIA stands to benefit immensely. The increased complexity and potential for deeper reasoning in LCMs might even drive demand for more powerful or specialized AI hardware.

Generative AI is clearly a big deal, but the biggest winner so far is Nvidia, as we have heard about Nvidia plan on the Generative AI from its CEO Jensen Huang, I believe the LCM would be a perfect model to run on the Nvidia AI platform, and hence the big question weighing on investors’ minds is when all of this GPU spend will generate a return, might soon be answered.

Cloud Service Providers ($Amazon.com(AMZN)$ - AWS, Microsoft (MSFT) - Azure, Google (GOOGL) - Google Cloud): These companies provide the computing infrastructure that powers the training and deployment of large AI models. As LCMs become more prevalent, demand for their cloud computing services, particularly for AI workloads, will grow.

Specialized AI Hardware Developers: Companies developing custom AI chips (ASICs) or accelerators designed for more efficient AI computation could also see a boost, though this is a more niche and higher-risk area.

Companies Developing AI Applications that Benefit from Enhanced Reasoning & Coherence

LCMs' strengths lie in improved contextual understanding, coherent long-form generation, and logical reasoning across multiple languages and modalities. This makes them ideal for:

Content Generation Platforms: Companies that leverage AI for writing articles, reports, marketing copy, creative works, or even code will benefit from LCMs' ability to produce more coherent, contextually relevant, and less repetitive output, potentially reducing the need for extensive human editing. This could include platforms serving:

Journalism and Media: For generating news summaries, articles, or even scripts.

Marketing and Advertising: For producing compelling ad copy, campaign narratives, or personalized content. Academic and Research Tools: For summarizing complex papers, generating research outlines, or assisting in hypothesis testing.

Enterprise AI & Business Automation: Customer Service & Support: AI chatbots and virtual assistants powered by LCMs could provide more intelligent, context-aware, and consistent interactions across various channels (text, voice, video). This would benefit companies in customer relationship management (CRM) and contact center solutions.

Data Analysis & Decision Making: LCMs can analyze complex datasets, identify underlying relationships, and explain their reasoning, making them valuable for business intelligence, risk assessment, and strategic planning. Companies in sectors like finance, healthcare, and supply chain management could integrate LCMs for better insights.

Cybersecurity: Enhanced reasoning capabilities could improve threat detection and analysis by understanding complex attack patterns and predicting risks more accurately.

Multilingual & Multimodal AI Solutions: Translation and Localization Services: LCMs' language-agnostic nature and ability to understand concepts across languages could revolutionize real-time translation, cross-lingual communication tools, and localization efforts for global businesses. Accessibility Technologies: Companies working on tools for individuals with disabilities, such as sign language interpretation or audio-visual summarization, could see significant advancements with LCMs.

Drug Discovery & Scientific Research: LCMs could accelerate scientific discovery by helping researchers analyze vast amounts of data, test hypotheses, identify patterns in complex biological systems, and even suggest new experimental directions. Companies in biotech and pharma that are heavily investing in AI could see benefits.

Key Considerations for Investors

Early Stage: LCMs are still a relatively new concept compared to LLMs. While promising, their widespread commercial adoption and impact are still developing.

Integration with LLMs: Many experts believe LCMs will not entirely replace LLMs but will rather complement them. Companies that can effectively integrate both architectures for superior performance will likely be winners.

Data and Compute: Access to vast, diverse, and high-quality data, as well as significant computing power, will remain crucial for training and deploying effective LCMs.

Talent: Companies with leading AI research teams and the ability to attract top talent in this specialized field will have a competitive edge.

Given the early stage of LCMs, investing directly in companies primarily focused solely on LCMs might be highly speculative. A safer approach might be to consider:

Major tech giants with established AI research divisions (Meta, Google, Microsoft) that are already investing in this area.

Enablers like NVIDIA and cloud providers who will benefit regardless of the specific AI model architecture that gains prominence.

Application-layer companies that are known for rapidly adopting and integrating cutting-edge AI into their products and services.

Examples of How LLM and LCM Process Text Token

Imagine reading a novel. An LLM would process it token by token, focusing on individual words and their immediate neighbors. With this approach, it could generate a summary by predicting the most likely next word. But it may miss the broader narrative and underlying themes.

LCMs, however, analyze larger sections of text to extract the underlying ideas. This approach helps them understand the broader concepts: overall story arc, character development, and themes. This approach can not only help them generate a more complete summary, but it can help them expand on the story in a more meaningful way.

The below diagram is a simplified explanation of how each of the three modular components of an LCM works. The encoder turns language into abstract concepts. Here, these abstract concepts are represented as images. In the model, these concepts are represented mathematically. The core runs inference on those concepts. Then the decoder turns those abstractions into human-readable language. For this figure, Copilot provided the first draft of the animal drawings.

Summary

As LCMs mature, more specific beneficiaries will emerge, but the general trends point towards companies that can leverage more sophisticated AI reasoning for complex, nuanced tasks.

The applications of LCMs overlap with those of LLMs, but because of their focus on concepts and deeper understanding, they have the potential to create a more profound impact on industries that require deeper thought.

Appreciate if you could share your thoughts in the comment section whether you think rise of LCMs will speed up the AI race and forces more companies to innovate fast.

@TigerStars @Daily_Discussion @Tiger_Earnings @TigerWire appreciate if you could feature this article so that fellow tiger would benefit from my investing and trading thoughts.

Disclaimer: The analysis and result presented does not recommend or suggest any investing in the said stock. This is purely for Analysis.