Le savoir des derniers podcasts.

De nouveaux épisodes traités par Podwise, prêts à chercher, parcourir et continuer dans l'app.

Tout juste résumés

Lisez ce que couvre chaque épisode avant de lancer la lecture.

The AI Daily Brief: Artificial Intelligence News and Analysis9 nov.

Nvidia CEO Admits Affair With Suspected Chinese Spy

Résumé IA Nvidia CEO Jensen Huang’s admission of an affair with a woman suspected of being a Chinese intelligence agent has triggered significant geopolitical and market instability. This scandal threatens to undermine Nvidia’s lobbying efforts against stringent AI chip export controls…

Nvidia
Anthropic
Megaspeed

00:26:41EnglishRésumé · Transcription

The AI Daily Brief: Artificial Intelligence News and Analysis8 nov.

Agentic Video Is Real and Here’s What You Should Try

Résumé IA The latest generation of AI models, particularly Opus 5.5, has triggered a significant shift in video editing, moving it from a specialized skill to a core capability for knowledge workers. This transformation stems from AI's ability to treat editing as a conversational process,…

Opus 4.5 / 5.5
Agentic coding
Vibe coding

00:30:58EnglishRésumé · Transcription

The Rollup12 nov.

Jeff Dorman: Cash-Flow Assets Are Taking Over Digital Markets (Cash-Flow Era)

Résumé IA The digital asset market is undergoing a fundamental transition from speculative social-value assets toward sustainable on-chain businesses. Investors should prioritize three high-growth sectors: stablecoins and payments, real-world asset (RWA) tokenization, and decentralized…

Hyperliquid
RWA (Real World Assets)
DeFi (Decentralized Finance)

00:45:09EnglishRésumé · Transcription

InRecord2 nov.

How Modders Unlocked a $5,000 Chip Inside Normal Nvidia GPUs

Résumé IA Nvidia intentionally throttles the high-precision AI performance of its consumer GeForce RTX graphics cards to prevent them from competing with its $5,000 enterprise workstation cards, despite both using the same underlying silicon. A new open-source tool called GPUUnlock…

GPUUnlock
Tensor Cores
RTX 4090

EnglishRésumé · Transcription

Junk GPU Lab29 oct.

Qwen 3.8 27B at 67 tok/s on Two $78 Nvidia Tesla P100s

Résumé IA Running a high-performance 27B parameter LLM at home is possible without expensive consumer hardware by repurposing budget-friendly, older server components. By utilizing two $78 Nvidia Tesla P100 cards, it is possible to achieve 67 tokens per second, significantly exceeding the…

Qwen 3.8 27B
VRAM
Quantization

EnglishRésumé · Transcription

Planet Money13 nov.

Are wildfires making the future uninsurable?

Résumé IA The insurance industry’s traditional risk-spreading model is fracturing as climate-driven megafires transform once-rare disasters into annual occurrences. In Paradise, California, residents face a dual crisis: skyrocketing premiums and widespread non-renewals that threaten the…