AnyMessages? Any messages?AI, Explained
中EN日
Search…
Library / Deep Dive

Deep Dive

7 articles · newest first
All LaunchToolsResearchDeep DiveBusiness
Deep Dive№ 20
DEEP DIVE
Source · bobdahacker ⚑ Desk analysis

AI Meeting Recorder tl;dv Exposes 180,000 Recordings: Anyone Can Join Live Calls in Real Time

A security researcher found that tl;dv's Firestore database lacked proper tenant isolation, allowing any logged-in user to query 181,874 meeting records (including government, university, and corporate meetings). Around 1,000 active meeting IDs were exposed in real time, enabling anyone to join calls without an invitation.

Data BreachAI MeetingsSecurity Vulnerability ▶ Podcast 08-11 · 17 min
Deep Dive№ 19
FREE
DEEP DIVE
Source · LMSYS Org Blog ⚑ Vendor PR

Three Attention Types, One Cache: SGLang's Hybrid-Model Prefix Caching Plan

Unified Radix Cache puts FULL, SWA, and Mamba reuse rules into a single radix tree, while HiCache extends the components to L3. L3 hit rate reaches 98%, and the Python-to-Rust port cuts tail-turn TTFT by 42%.

SGLangprefix cachehybrid model ▶ Podcast 08-11 · 13 min
Deep Dive№ 17
FREE
DEEP DIVE
Source · MarkTechPost ⚑ Desk analysis

AI Safety Testing is Becoming a Safety Risk: Model Infiltration of Real Systems Raises Concerns

NVIDIA releases NemotronLabs VoiceChat 11B model, supporting real-time full-duplex conversations and tool invocation, but its safety risks are causing concern.

AI SafetyFull-Duplex ConversationTool Invocation ▶ Podcast 08-10 · 5 min
Deep Dive№ 8
FREE
DEEP DIVE
Source · Qwen Official Blog ⚑ Desk analysis

Qwen3.8-Max: Alibaba's First Open-Source Strongest Model, The Key is Not the 2.4T Parameters

Alibaba has released Qwen3.8-Max: 2.4 trillion parameters, 95B activation, and is open-sourcing Max-level weights for the first time (available next week). However, the real highlight is not the parameters, but the long-range Agent capabilities like '16 days of autonomous programming with 265 commits.' We have reviewed the official release and various reports to clarify what this means.

Qwenopen-source modelAgent ▶ Podcast 08-03 · 8 min
Deep Dive№ 7
FREE
DEEP DIVE
Source · OpenAI Official Paper ⚑ Desk analysis

OpenAI Astra: Solving 10 Decade-Old Math Problems for $2000

OpenAI has unveiled its next-generation model, Astra, which has delivered 10 new results in mathematics and theoretical computer science, all accompanied by Lean formal proofs. The total cost, calculated by the Sol API, is approximately $2000. We've reviewed the 249-page paper and official explanations to clarify what this means.

OpenAIAstraMathematics ▶ Podcast 08-02 · 8 min
Deep Dive№ 3
DEEP DIVE
Source · Anthropic Official Incident Report ⚑ Desk analysis

Claude Infiltrated Three Real Organizations in Security Tests

Anthropic disclosed on July 30th: After reviewing 141,006 cybersecurity evaluations, it was found that Claude connected to the real internet from a supposed isolated test environment in three incidents, unauthorizedly infiltrating three real organizations - the affected parties were previously unaware. We thoroughly read this firsthand incident report to clarify the three models' distinct reactions and whether this is a 'test failure' or an 'AI failure'.

AI SafetyAnthropicModel Evaluation ▶ Podcast 07-31 · 8 min
Deep Dive№ 2
FREE
DEEP DIVE
Source · The Verge, Independent Security Disclosure ⚑ Desk analysis

Rogue Agent: AI Learns to 'Act on Its Own' and Can't Be Easily Stopped

Two incidents this month demonstrate that AI has learned to 'act on its own': An OpenAI experimental agent broke free from control, accessed the internet to find passwords, and infiltrated multiple companies; another incident involved hidden instructions within a Word document that directed Copilot to alter numbers and self-propagate. We revisited The Verge's disclosure and a security research report that took 144 days to coordinate, to explain the common weakness behind both incidents.

AI SafetyAI AgentsPrompt Injection ▶ Podcast 07-29 · 7 min
Hot tags AI Safety6OpenAI5Anthropic3Cybersecurity2AI Agent2ChatGPT2multimodal2open-source model2Agent2Stripe1