The AI landscape is evolving at a rapid pace, with data emerging as the central driver of competition, according to Gavin Baker, a leading U.S. TMT (Technology, Media, and Telecommunications) investor. As AI models proliferate, Baker warns that those lacking access to unique datasets will face rapid depreciation in value.

Elon Musk Endorses the Insight

Baker’s analysis on “future AI competition trends” recently caught the attention of Elon Musk, who responded with a simple “Good analysis” on social media. This endorsement coincided with the launch of Musk’s xAI Grok-3 model.

Shifting Dynamics in the AI Industry

Baker predicts that OpenAI’s dominance is narrowing. While OpenAI held a strong position from mid-2022 to spring 2024, competitors like Alphabet Inc. Class C(GOOG.US) Alphabet Inc. Class A(GOOGL.US), xAI, and Deepseek have now caught up to GPT-4 in terms of technical capabilities. This shift, driven by the diminishing returns of the traditional "Scaling Law" in AI model training, signals a broader change in the industry. Even Microsoft Corporation(MSFT.US) CEO Satya Nadella has acknowledged that OpenAI’s period of unparalleled leadership is drawing to a close.

The Rise of New Competitors and the Value of Unique Data

As competition intensifies, Baker stresses that the key differentiator will be access to exclusive data. Companies like Google and xAI, with access to platforms such as YouTube, TeslaVision (Tesla Motors, Inc.(TSLA.US)), and X, are well-positioned to leverage this data advantage. Baker cautions that future AI models that fail to harness unique, valuable data sources could quickly become obsolete, offering limited return on investment.

A Shift Toward Inference-Centric AI Infrastructure

Looking ahead, Baker predicts a significant shift in AI infrastructure. The focus will move away from pre-training models toward inference-centric computing. In this new paradigm, only 2-3 mega data centers will handle most of the computing power, with inference accounting for 95% of AI computing needs. This change will be supported by advances in cheaper electricity, quantization technologies, and more cost-effective infrastructure models.

Indirect Investment Guidance

Investors may want to consider how companies are positioning themselves to leverage unique data, as this will likely become the key to maintaining a competitive edge. Furthermore, infrastructure investments that focus on inference capabilities, rather than pre-training, could represent a more sustainable growth avenue. With key players like Microsoft shifting focus toward providing inference computing power and exploring open-source models, investments in firms that adapt to these trends might hold long-term value.

Seizing the New AI Wave: Promising Stocks to Watch in Key AI Sectors

While OpenAI may retain its leadership due to its early advantages, the competition in the AI space is rapidly evolving. The companies that succeed in securing unique data and developing cost-efficient inference infrastructures will likely dominate the next phase of AI development. As Baker concludes, “Time will tell.”

Below is a list of prominent AI-related stocks that are likely to capture investor interest in the U.S. markets: