Apple Lost the AI Race
Our read
Apple is ceding the cloud-hosted AI battle to competitors because their real play is controlling the local hardware execution layer.
What happened
Marques Brownlee deconstructs the narrative of Apple's AI failure by dividing the technology into two distinct battles: cloud-based raw intelligence and local hardware execution. While competitors burn capital on massive data centers, Apple's strategy is to let others subsidize the software R&D while they leverage their hardware monopoly to remain the ultimate local toll collector.
Quotes
“I don't think Apple Intelligence has to be that good... It just has to be good enough for investors to believe Apple's competitive... but it doesn't have to be amazing for Apple to still be winning.”
Marques Brownlee · 02:38
“They don't do search engines either, and they sell you the device that you do the Googling on.”
Marques Brownlee · 04:30
The brief
Marques Brownlee (MKBHD) analyzes the strategic positioning of Apple in the artificial intelligence landscape. Rather than viewing Apple's late entry and reliance on third-party integrations as a failure, the analysis frames this as a deliberate hardware-first strategy.
By focusing on local execution via Apple Silicon, Apple positions itself to capture the value of AI processing without the massive capital expenditures associated with training and running cloud-scale models.
The real danger to Apple's business model is not a superior software assistant, but the potential emergence of alternative, AI-native hardware form factors that could disrupt the smartphone's dominance.
Receipts
Visual-only receipts
- Pie chart showing Apple's Revenue Share by Segment, highlighting the iPhone's dominance in 2025 at $209.59 Billion.
- Google Search results indicating a trend of developers and early adopters buying Mac Minis specifically to run local AI models.
- Article headline: OpenAI Said to Be Making Phone With AI Agents to Replace Apps.