Silver Rises Over 120% YTD 
DeepSeek’s breakthrough in cost-efficient AI model training has sent shockwaves through financial markets, particularly affecting semiconductor and tech stocks.
The Chinese company’s ability to train a competitive AI model for just $6 million, compared to the much higher costs of industry leaders like OpenAI ($78 million) and Google ($191 million), has raised fundamental questions about the sector’s economics.
While some analysts remain skeptical of DeepSeek’s claims, their significantly lower end-user pricing – charging just 14 cents per million input tokens versus OpenAI’s $15 – has caught market attention.
The implications extend beyond chip manufacturers to recent infrastructure investments, including Trump’s announced $500 billion AI deal with SoftBank, Oracle, and OpenAI. Furthermore, the development could dramatically impact utility companies and power infrastructure planning, as Morgan Stanley’s projection of AI consuming 10% of U.S. electricity by decade’s end may need revision if DeepSeek’s efficient training methods become industry standard.
The open-source nature of DeepSeek’s model means these efficiency gains could quickly spread throughout the industry, potentially transforming the entire AI infrastructure landscape.
