SOXL sinks 12%, Alibaba's Qwen4-Max, and the p99 latency interview question
From My Daily AI & Markets Briefing · 00:22:28
Show by Mahdi Naser Moghadasi
The chip sell-off deepens, Alibaba drops a new flagship model, and we coach you through a staff-level latency question. Markets: SOXL -12.16% to $51.23, VGT -1.92% to $109.05, QQQ -1.61% to $668.91, VUG -1.39% to $83.06. AI news: Alibaba releases Qwen4-Max, its new flagship AI model, raising price and capability pressure on US frontier labs. Interview Tip of the Day: You need to cut a model's p99 latency in half without retraining. What can you change in serving (batching policy, caching, quantization, hardware, request routing, shorter outputs), and how would you prove quality didn't drop? Did You Know: The mailbox notifier, a tiny sensor that pings your phone when the post arrives. Questions or feedback: [email protected]
More episodes of My Daily AI & Markets Briefing
More episodes
- Nvidia after Jensen, QQQ slides 1.2%, and the mixture-of-experts interview
- SOXL sinks 11%, Nvidia's post-Huang test, and the feature store question
- SOXL sinks 12%, Google's Gemini 4 Hydrogen, and the 35% GPU utilization question
- SOXL drops 4.5%, Jensen Huang exits Nvidia, and the safe LLM agent question
- SOXL up 12.8%, Anthropic's IPO founder sales, and catastrophic forgetting
- SOXL jumps 4.6%, Anthropic's Claude Mythos 5.2, and the prohibited-ads ML question
- SOXL soars 15%, Nvidia's board probes Jensen Huang, and the RLHF vs. DPO question
- SOXL slides, Anthropic files its S-1, and the model-drift interview question
- SOXL sinks ~38%, OpenAI's $1.25T raise, and the 0.1% fraud-model question
- SOXL jumps 17%, METR's Anthropic speed audit, and the streaming recsys question
PodSearch is a free, independent podcast search engine. Read about PodSearch, the privacy policy or contact us.