Broadcom and Intel CEO warn of AI financing costs and uncertainty over returns
BIBIBI
AT A GLANCE
Broadcom CEO Hock Tan said that capital is becoming the AI industry's greatest constraint and warned of a risk that demand for computing power could slow.
Article
Broadcom CEO Warning: Capital is replacing chips and power as the AI largest bottleneck to development
Dongcha Beating AI Flash News: Broadcom CEO Hock Tan stated at the DBS Asia Future Forward conference in Singapore that capital is replacing chips, memory, and power supply as the greatest constraint on the development of the AI industry. He warned that many AI investors do not understand the technology they are betting on, yet expect high returns, making such investment behavior akin to gambling in a casino.
Using OpenAI as an example, Tan noted that the company's September annualized revenue has reached 50000000000 USD, but high model-training costs still leave it operating at a loss; revenue growth does not mean that its business model has been validated. He also expects the wafer shortage to improve between 2028 and 2029 years, while power supply is also expected to gradually ease as new infrastructure comes online.
Intel CEO Lip-Bu Tan also stated that rising financing costs and uncertainty over investment returns are becoming core challenges for the AI industry. Although both believe that demand for computing power will remain strong over the next to 3 years, Tan warned that demand for computing infrastructure could slow significantly if adoption of advanced AI models falls short of expectations or improvements in model capabilities encounter bottlenecks.
Original link https://m.theblockbeats.info/flash/371414
Key points
01
Broadcom CEO Hock Tan stated at the DBS Asia Future Forward conference in Singapore that capital is replacing chips, memory, and power supply as the greatest constraint on the development of the AI industry.
02
Hock Tan said that many AI investors do not understand the technology they are betting on but expect high returns, comparing such investment behavior to gambling in a casino.
03
Using OpenAI as an example, Hock Tan said that the company's September annualized revenue has reached 50000000000 USD, but model-training costs still leave it operating at a loss.
04
Hock Tan expects the wafer shortage to improve between 2028 and 2029 years, while power supply may also gradually ease as new infrastructure comes online.
05
Intel CEO Lip-Bu Tan stated that rising financing costs and uncertainty over investment returns are becoming core challenges for the AI industry.
06
Hock Tan and Lip-Bu Tan both believe that demand for computing power will remain strong over the next to 3 years.
07
Hock Tan warned that, if adoption of advanced AI models falls short of expectations or improvements in model capabilities encounter bottlenecks, demand for computing infrastructure could slow significantly.
AI-assisted interpretation
The following is analysis, separate from reported facts. Verify important claims independently.
The core view of the company CEO leaders is that:AI substantial computing power may still be needed in the near term, but the capital required for construction and training, financing costs, and the ability to generate returns are becoming more prominent constraints. Hock Tan also believes that high revenue alone does not prove that a business model has been established, because model-training costs may also be high.
Why it matters to readers
These views shift the focus of the AI industry from supply issues such as chips and power to funding and investment returns. If progress in applications or model capabilities falls short of expectations, demand for the computing infrastructure described in the material could slow.
Beginners should distinguish between “strong demand” and “an investment is certain to be profitable”: the material states that demand for computing power may remain strong over the next to 3 years, but financing costs, training costs, and uncertainty over returns remain.
Risks and unknowns
The material is a media account and does not provide original conference video, transcripts, or company documents, so the full statements and context of the CEO leaders cannot be verified.
OpenAI“September The claim that annualized revenue has reached 50000000000 USD and that the company remains at a loss is not supported by original data sources in the material.
The improvement in the wafer shortage between 2028 and 2029 years, the gradual easing of power supply, and strong demand for computing power over the next to 3 years are all forecasts and have not yet occurred.
Related Developments
Loading event timeline…
Whether adoption of advanced AI models, improvements in their capabilities, and demand for computing infrastructure will slow remains uncertain.
Related concepts
annualized revenue
This term is not in the glossary yet. Browse related concepts in the glossary.
