A new artificial intelligence (AI) model from Chinese startup DeepSeek has sparked a divide among innovators in machine learning, exposing a similar tension between two visions for the technology’s future.
On one side are those seeking “artificial general intelligence” (AGI), where machines match or surpass human capabilities. They envision AI as a path to creating thinking machines. Their champions include chatbots like OpenAI’s ChatGPT and Anthropic’s Claude, which use large language models to predict the next word in a string of text.
On the other side are those focused on solving specific problems efficiently using machine learning algorithms. They see AI as a tool, not an end goal. Google DeepMind’s AlphaFold2 is their icon, a model that predicted protein structures with remarkable accuracy using tailor-made deep learning algorithms and limited computing power.
DeepSeek’s latest model has raised doubts about the return on investment in AGI, causing shares of Nvidia, Microsoft, and Alphabet to drop. However, investors may find value in the more pragmatic approach, which can deliver concrete economic benefits without requiring vast investments in high-end hardware.
As the AI landscape evolves, it’s essential for investors to understand the divide between AGI enthusiasts and those who prioritize practical applications. The latter approach has a compelling logic: take aim at well-defined predictive challenges and choose inputs to achieve at least cost. With companies like DeepMind already achieving significant success using this strategy, it may be time for investors to reassess their stance on AI’s future.
Source: https://www.reuters.com/breakingviews/ais-civil-war-will-force-investors-pick-sides-2025-02-07