Cohere's Aidan Gomez on AI's pricing divide, plus Big Tech earnings — 4/30/2026
Speakers
Summary
CNBC's Deirdre Bosa covers Big Tech earnings with a focus on AI capex and free cash flow pressure. She interviews Cohere CEO Aidan Gomez on the enterprise AI pricing divide and the shift toward efficient models. Analyst Daniel Newman argues that compute constraints benefit chip makers like Nvidia, Qualcomm, and Google, while Apple's AI strategy is lagging.
- Microsoft, Google, Amazon, and Meta reported earnings with combined capex commitments of ~$700 billion.
- Free cash flow is collapsing across the group, with Amazon down 95% year-over-year.
- Chinese AI models like DeepSeek V4 are nearly as capable at a fraction of the cost, but are not an option for regulated industries.
- Cohere CEO Aidan Gomez emphasizes security and efficiency as key enterprise AI buying criteria.
- Cloud growth is booming: Google Cloud grew 63%, AWS had its best quarter in three years.
- Daniel Newman highlights Nvidia, Qualcomm, ARM, and Intel as beneficiaries of agentic compute demand.
- Google's full-stack advantage is rewarded by the market, while Apple is seen as weak on AI.
- The demand for AI is still in early stages, with only 2% of consumers and 20% of businesses paying for it.
Trade Ideas
Google benefits from cloud growth and full stack
Google's 63% cloud growth justifies its huge capex, and its full-stack advantage (cloud, TPU, models) is being rewarded by the market, as seen in the 10% stock price rise after earnings.
Chip adders benefit from compute constraints
Compute constraints are so severe that any company adding capacity, like Qualcomm with its custom XPU CPU, ARM, and Intel, will rip and run forward as beneficiaries of the agentic compute trend.
Nvidia will solve open source AI
Nvidia is going to solve the open source AI problem and is well positioned to benefit from the compute-constrained environment and massive AI infrastructure buildout.
Apple's AI strategy is weak and needs innovation
Apple has fumbled its AI strategy, lacks a compelling AI story, and needs more innovation; the installed base may not be enough as new agentic devices emerge, making it a stock to avoid.
