The article argues that success in AI development will be determined by practical applications that meet everyday needs, not just by having the most advanced models. It highlights significant public distrust of AI in the US compared to high trust in China, and concerns that market hype and valuations may be outpacing real-world adoption.
Financial volatility is a major concern, as AI stocks are deeply tied to the US economy and retirement funds, yet their long-term profitability is questioned. The US is advised to shift focus from just leading in large language models to integrating AI into industrial processes, leveraging massive planned infrastructure investments to fuel re-industrialization.
The main topics covered are the US-China AI competition, public trust in AI, financial market risks related to AI stocks, and the strategic need for industrial AI integration.
Opinion | China is leading in holistic AI development. Can the US catch up?
The country that will emerge victorious may not have the best models, but rather the applications which best meet people’s everyday needs
After a series of roller-coaster rides in financial markets, serious concerns remain about the long-term profitability of major artificial intelligence (AI) companies that mostly rely on circular financing – investing in each other to prop up demand.
This volatility is worrisome when many Americans’ retirement pensions are closely tied to AI stocks, which have contributed to 80 per cent of the US stock market’s rise and 40 per cent of US GDP growth last year.
Amid fears of a bubble, polls suggest that only 32 per cent of Americans trust AI, in stark contrast with 87 per cent of Chinese. This lack of public support in the US has raised concerns about whether hype has outpaced actual future adoptions that justify sky-high valuations of US big tech. Many Americans blame AI data centres for higher energy bills when energy grids struggle to meet the data centre demands.
To win the AI race, both sides are pushing for the next phase of growth, albeit in different ways. The US must go beyond its obsession with dominance in large language models (LLMs) and AI-generated images and shift towards industrial integration. It needs to connect its robust AI infrastructure through its Stargate Project – which aims to invest up to US$500 billion in AI infrastructure by 2029 – to boost its re-industrialisation efforts after decades of deindustrialisation.