Episode Details

Back to Episodes
OpenAI Models Escape Their Sandbox and Hack Hugging Face

OpenAI Models Escape Their Sandbox and Hack Hugging Face

Published 2 months, 3 weeks ago
Description

In this episode, we discuss the Iran war’s renewed oil shock, OpenAI models compromising Hugging Face during an evaluation, and Kraken’s expansion of xStocks into global equities.



00:00 Introduction



01:24 Crypto Headlines and Web3 Roundup



07:25 Microsoft and Mistral Expand Their AI Partnership



13:46 Iran War Escalation and Oil Market Risk



18:45 OpenAI’s Hugging Face Security Incident



21:55 How the OpenAI Model Escaped Its Sandbox



25:36 GLM 5.2 and the AI Defense Asymmetry



30:52 Why Model Restrictions Hit Retail Users



35:00 Kraken Expands xStocks Beyond U.S. Equities



38:16 Crypto Perpetuals Reach $90 Trillion in Volume



42:02 HIP-3 Stock Perps and Exploding Trading Volume



46:29 Leverage Demand From Hyperliquid to the KOSPI



The content in this show is never financial advice. The content is an expression of individuals' opinions and is for informational and entertainment purposes only.

Listen Now

Love PodBriefly?

If you like Podbriefly.com, please consider donating to support the ongoing development.

Support Us