Capital shortage looms as Broadcom's Tan says AI investors lack economic clarity
Broadcom chief warns that many investors lack clarity on AI economics, comparing their strategies to gambling in a casino environment.
Key points
- Hock Tan stated that many investors have no clear understanding of what they are funding in the AI sector.
- The executive compared current enterprise AI spending to a child wasting money at an arcade without learning anything.
- Both Tan and Intel CEO Lip-Bu Tan identified capital availability as a major constraint for future AI development.
Broadcom CEO Hock Tan criticized speculative investment in artificial intelligence, comparing uninformed investors to casino gamblers during a panel discussion in Singapore. Speaking at the DBS Asia Future Forward Conference, Tan argued that many investors set extreme expectations without understanding the underlying technology or economics.
According to Bloomberg, Tan stated that a significant number of investors have no clue what they are investing in. He observed that these individuals set very high expectations only to realize their situation resembles a casino. This criticism emerged during a conversation about where the next trillion dollars in AI value will originate, highlighting concerns about sustainable business models.
Economic Realities and Constraints
Tan noted that most companies have not yet identified AI applications that generate a positive return on investment. He described current enterprise spending as wasteful, likening it to a child spending ten dollars in ten minutes at an arcade without gaining any knowledge. This behavior, he said, characterizes what is happening in businesses today.
Regarding specific valuations, Tan addressed OpenAI’s annualized revenue of $50 billion reported in September. He stated that while this figure is not bad for a startup, it does not indicate long-term profitability because the company is still losing money due to heavy training expenses. The revenue number alone says little about sustainable profits.
Capital and Infrastructure Outlook
Both Hock Tan and Intel CEO Lip-Bu Tan agreed that money could become a larger constraint on AI progress than chips, memory, or power. Tan identified capital and the cost of capital as the biggest challenges ahead. He emphasized that investors will eventually demand returns on their investments, regardless of current hype.
Despite these financial concerns, demand for computing infrastructure continues to outstrip supply, a trend Tan expects to last through 2029 or possibly into 2030. He anticipates wafer shortages easing by 2028 or 2029. Intel’s Tan said he sees no significant slowdown in the next two to three years. Broadcom is raising $60 billion in debt to help customers like Anthropic pay for AI infrastructure.
Defender's checklist
Small-business owners and IT teams should monitor how AI spending impacts their organization's budget and operational stability. Understanding the difference between experimental AI use and revenue-generating applications is critical for long-term planning. Teams should also watch for shifts in capital availability that could affect vendor stability and support.
What to do and how to stay safe: Broadcom
- Audit current AI expenditures to ensure they align with measurable business outcomes rather than speculative trends.
- Review vendor financial health and capital requirements to assess long-term support stability for critical infrastructure.
- Evaluate AI projects for clear return on investment before scaling spending on tokens or compute resources.
- Monitor industry reports on capital constraints and infrastructure supply to anticipate potential disruptions in service.
Step-by-step guide: Implementing AI Usage Policies at Work: A Step-by-Step Framework
General security guidance from the InfoSec Ledger newsroom. It is not confirmed advice from the organisations named in this story.
Frequently asked questions
Why did Hock Tan compare AI investors to casino gamblers?
Tan stated that many investors have no idea what they are investing in and set extreme expectations, similar to gambling behavior.
What did Tan say about OpenAI's revenue?
He noted that OpenAI's $50 billion annualized revenue is decent for a startup but does not prove long-term profitability due to heavy losses.
What is the biggest constraint for AI growth according to the CEOs?
Both Hock Tan and Lip-Bu Tan identified capital and the cost of capital as the primary constraints, potentially bigger than hardware supply.




