Watch / Artificial Intelligence (AI)
Why the AI-trade sceptics might be wrong
Why the AI-trade sceptics might be wrong
About this video
Key points:
AI trade seen as a healthy correction, not a collapse
Chinese open‑source models like Kimi viewed as credible but constrained
Energy, regulation and data‑centre scepticism cited as primary AI risks
Hatem Dhiab from Gerber Kawasaki portrays the AI trade as tired rather than broken, characterising the current pullback as a healthy correction after a three‑year surge. Dhiab states that investors are now pressing harder on questions of sustainability, margins and capital expenditure, which he views as a constructive shift away from euphoria. He maintains that AI infrastructure demand is only in the early stages, with inference workloads set to multiply as businesses embed AI more deeply.
On China’s new “Kimi” model, Dhiab notes that open‑source Chinese AI regularly surprises those who assume it lags the US, suggesting Kimi may outperform some Western systems. He flags, however, that unknown model parameters, high hardware requirements and trust issues around Chinese technology limit its immediate impact on US leaders such as OpenAI, Anthropic and Google’s models.
Dhiab remains bullish on Nvidia (NASDAQ:NVDA), arguing it sits at the centre of AI training and inference, and questions whether hyperscalers can sustain soaring capex. He highlights upcoming results from Alphabet (NASDAQ: GOOGL), Microsoft (NASDAQ: MSFT), Meta Platforms (NASDAQ: META), Apple (NASDAQ: AAPL) and Amazon (NASDAQ: AMZN), focusing on capex, cloud growth and margin pressure from rising energy costs. He views regulatory and societal pushback on data centres and energy use as the main risks to the AI build‑out.