DeepSeek's Economy Shifts: The Modern Financial Strategy in AI
The stunning impact of Chinese AI DeepSeek this January, which caused a staggering $1 trillion loss in the stock market, serves as a stark reminder of the rapid evolution of artificial intelligence (AI). Beyond the headlines, this event symbolizes a transformative shift in how businesses operate.
DeepSeek may not have rivaled industry giants like OpenAI in terms of capabilities, but its significant achievement lies in offering comparable results at a fraction of the expense. For business leaders, this development signifies AI transitioning from an intriguing novelty to a universally accessible tool.
This cost-oriented revolution, a well-established pattern in technology development, should come as no surprise to those familiar with its historical trajectory. Just like numerous advancements before it, AI has consistently become more efficient on a quality-adjusted basis. The key difference is how dramatically DeepSeek highlighted this trend for the global market.
In its early stages, large language models (LLMs) focused primarily on augmenting quality, piping up with interesting but not always practical results. Enhancing these models required substantial resources, including high costs associated with training them on vast data sets.
However, as quality has grown and costs have subsided, new objectives have arisen. Reducing expenses and boosting performance are now priorities. By curbing attention to minor details and omitting underutilized parameters, developers can achieve a target level of quality while performing fewer calculations.
Another major development underway is minimizing hallucinations – a tendency of LLMs to fabricate information – particularly as these models enter real-world applications. Advancements in this area will increasingly understand and cater to human needs, potentially revolutionizing various industries.
This shift in priorities mirrors the evolution of automobiles. In the late 1800s, finding a way to move vehicles without horses was a groundbreaking achievement. Then, as the automotive industry progressed, priorities evolved from cost-cutting to enhancing features and, ultimately, reliability.
This historical analogy holds true for AI as well. Many concerns are currently being addressed by developers, eventually resulting in the widest application of AI across diverse industries.
Economist John Cochrane astutely noted that while AI producers may not be the immediate beneficiaries, it is the businesses that grasp how to effectively and creatively leverage increasingly affordable AI that will truly reap the rewards[1]. As a result of lower costs, the total spending on AI is expected to grow as businesses identify new uses for this technology, often at higher expense to maximize gain.
In conclusion, the introduction of DeepSeek symbolizes an exciting new era for AI, driving innovation in various sectors worldwide. The transformation from expensive to accessible AI will empower businesses and reduce barriers, leading to increased adoption and improved ROI[2][3].
DeepSeek's achievement of providing comparable results at a lower cost is a testament to the Jevons paradox, where technological advancement leads to increased demand and use, resulting in a potential paradoxical increase in the amount of the input being used.
This cost-effective AI development not only challenges industry giants but also contributes to the realization of the Jevons paradox in the field of artificial intelligence, making AI a more universally accessible tool for businesses.
