/
Navigation
C
Chronicles
Browse all articles
C
E
Explore
Semantic exploration
E
R
Research
Entity momentum
R
N
Nexus
Correlations & relationships
N
~
Story Arc
Topic evolution
S
Drift Map
Semantic trajectory animation
D
P
Posts
Analysis & commentary
P
Browse
@
Entities
Companies, people, products, technologies
Domains
Browse by publication source
Handles
Browse by social media handle
Detection
?
Concept Search
Semantic similarity search
!
High Impact Stories
Top coverage by position
+
Sentiment Analysis
Positive/negative coverage
*
Anomaly Detection
Unusual coverage patterns
Analysis
vs
Rivalry Report
Compare two entities head-to-head
/\
Semantic Pivots
Narrative discontinuities
!!
Crisis Response
Event recovery patterns
Connected
Nav: C E R N
Search: /
Command: ⌘K
Embeddings: large
VOICE ARCHIVE

Jen Zhu

@jenzhuscott
32 posts
2026-03-02
Worth a read - •Ex-OpenAI geopolitics lead: frontier AI labs' military policies r deliberately vague & changeable to preserve “optionality” •Anthropic's DoD standoff isn't the ethical win as portrayed. Dario is hardly a white knight - he's open to fully autonomous weapons if
2026-03-02 View on X
fishbowlification

Frontier AI labs' military usage policies for their AI tools are incoherent, vague, and often change, which allows company leadership to preserve “optionality”

I led the Geopolitics Team at OpenAI for approximately three years and then joined two other teams before deciding to leave in June 2025.

Worth a read - •Ex-OpenAI geopolitics lead: frontier AI labs' military policies r deliberately vague & changeable to preserve “optionality” •Anthropic's DoD standoff isn't the ethical win as portrayed. Dario is hardly a white knight - he's open to fully autonomous weapons if
2026-03-02 View on X
Hyperdimensional

The Anthropic-DOD skirmish is the first major public debate on control over frontier AI, and institutions behaved erratically, maliciously, and without clarity

On Anthropic and the Department of War  —  I.  —  A little more than a decade ago, I sat with my father and watched him die.

Worth a read - •Ex-OpenAI geopolitics lead: frontier AI labs' military policies r deliberately vague & changeable to preserve “optionality” •Anthropic's DoD standoff isn't the ethical win as portrayed. Dario is hardly a white knight - he's open to fully autonomous weapons if
2026-03-02 View on X
The Verge

Sources: OpenAI agreed to follow US laws that have allowed for mass surveillance in the past, and the DOD didn't budge from its demands over bulk analyzing data

On Friday evening, amidst fallout from a standoff between the Department of Defense and Anthropic, OpenAI CEO Sam Altman announced …

Worth a read - •Ex-OpenAI geopolitics lead: frontier AI labs' military policies r deliberately vague & changeable to preserve “optionality” •Anthropic's DoD standoff isn't the ethical win as portrayed. Dario is hardly a white knight - he's open to fully autonomous weapons if
2026-03-02 View on X
The Atlantic

A source describes the failed Pentagon-Anthropic talks: through the end, the Pentagon wanted to use Anthropic's AI to analyze bulk data collected from Americans

Right up until the moment that Pete Hegseth moved to terminate the government's relationship with the AI company Anthropic …

2026-02-12
Mustafa has not heard of liabilities, accountability, licensing, Bar qualifications in those professions he uttered. “Fully automated” won't happen in this timeline. What it will happen is, similar to coders, elevating, transforming humans role in these functions to produce
2026-02-12 View on X
Financial Times

Mustafa Suleyman says Microsoft is pursuing “true self-sufficiency” in AI by building models for enterprise and health care and reducing its reliance on OpenAI

2026-01-02
Excellent thread summarising @deepseek_ai Jan 1st mHC paper. Two quant funds' open-sourced labs in 🇨🇳 already set a dizzy pace for 2026 in its first 24 hours. This builds nicely on the ByteDance Hyper-Connections paper - mHC's manifold constraint feels like a principled [image]
2026-01-02 View on X
South China Morning Post

DeepSeek researchers detail mHC, a new architecture they used to train 3B, 9B, and 27B models, finding it scaled without adding significant computational burden

DeepSeek has published a technical paper co-authored by founder Liang Wenfeng proposing a rethink of its core deep learning architecture

2026-01-01
Excellent thread summarising @deepseek_ai Jan 1st mHC paper. Two quant funds' open-sourced labs in 🇨🇳 already set a dizzy pace for 2026 in its first 24 hours. This builds nicely on the ByteDance Hyper-Connections paper - mHC's manifold constraint feels like a principled [image]
2026-01-01 View on X
South China Morning Post

DeepSeek researchers detail a new mHC architecture they used to train 3B, 9B, and 27B models, finding it scaled without adding significant computational burden

DeepSeek has published a technical paper co-authored by founder Liang Wenfeng proposing a rethink of its core deep learning architecture

2025-11-30
The whale 🐋 saves the world from AI Feudalism. The fact that it's a Chinese lab and team truly challenge people's prejudices. But those who know, know. Thankful for @deepseek_ai and each and everyone who's working and contributing on open-sourced AI 💙
2025-11-30 View on X
The Decoder

DeepSeek says its new DeepSeekMath-V2 model got gold-medal level status on the International Mathematical Olympiad 2025 and Chinese Mathematical Olympiad 2024

where models prove formal mathematical theorems—GPT-5 scores 20%.  Gemini Deep Think IMO Gold hits 65.7%.  DeepSeek Math V2 (Heavy) scores 61.9%.  That's second place—but Gemini is...

The whales 🐋 is back! DeepSeek-Math-V2: 685B-parameter math monster built on V3.2-Exp-Base, fully open under Apache 2.0 - 1st model to use a generator-verifier loop in training: writes proofs → verifier scores them → RL closes the loop for self-verifiable reasoning. - Focuses
2025-11-30 View on X
The Decoder

DeepSeek says its new DeepSeekMath-V2 model got gold-medal level status on the International Mathematical Olympiad 2025 and Chinese Mathematical Olympiad 2024

where models prove formal mathematical theorems—GPT-5 scores 20%.  Gemini Deep Think IMO Gold hits 65.7%.  DeepSeek Math V2 (Heavy) scores 61.9%.  That's second place—but Gemini is...

It just dawned on me why Wenfeng named his lab ‘DeepSeek’ w 🐋 as symbol: It rarely surfaces, but each time it does, it makes a huge splash. Then, it just goes deep into the ocean again & seek deeply. It carries heavy weight w quiet persistence. It's intelligent. It's rare.
2025-11-30 View on X
The Decoder

DeepSeek says its new DeepSeekMath-V2 model got gold-medal level status on the International Mathematical Olympiad 2025 and Chinese Mathematical Olympiad 2024

where models prove formal mathematical theorems—GPT-5 scores 20%.  Gemini Deep Think IMO Gold hits 65.7%.  DeepSeek Math V2 (Heavy) scores 61.9%.  That's second place—but Gemini is...

2025-11-29
It just dawned on me why Wenfeng named his lab ‘DeepSeek’ w 🐋 as symbol: It rarely surfaces, but each time it does, it makes a huge splash. Then, it just goes deep into the ocean again & seek deeply. It carries heavy weight w quiet persistence. It's intelligent. It's rare.
2025-11-29 View on X
The Decoder

DeepSeek says its new DeepSeekMath-V2 model got gold-medal level status on the International Mathematical Olympiad 2025 and Chinese Mathematical Olympiad 2024

Chinese startup Deepseek reports its new DeepseekMath-V2 model has reached gold medal status at the Math Olympiad …

The whale 🐋 saves the world from AI Feudalism. The fact that it's a Chinese lab and team truly challenge people's prejudices. But those who know, know. Thankful for @deepseek_ai and each and everyone who's working and contributing on open-sourced AI 💙
2025-11-29 View on X
The Decoder

DeepSeek says its new DeepSeekMath-V2 model got gold-medal level status on the International Mathematical Olympiad 2025 and Chinese Mathematical Olympiad 2024

Chinese startup Deepseek reports its new DeepseekMath-V2 model has reached gold medal status at the Math Olympiad …

The whales 🐋 is back! DeepSeek-Math-V2: 685B-parameter math monster built on V3.2-Exp-Base, fully open under Apache 2.0 - 1st model to use a generator-verifier loop in training: writes proofs → verifier scores them → RL closes the loop for self-verifiable reasoning. - Focuses
2025-11-29 View on X
The Decoder

DeepSeek says its new DeepSeekMath-V2 model got gold-medal level status on the International Mathematical Olympiad 2025 and Chinese Mathematical Olympiad 2024

Chinese startup Deepseek reports its new DeepseekMath-V2 model has reached gold medal status at the Math Olympiad …

2025-11-20
Memo for near-term home robots vs Figure is utility-first over human mimicry philosophy: - Wheeled base + lift column = passive stability, 5-10x longer battery life, no falling, far cheaper/safer than bipedal balancing act - Simple dual parallel grippers + elite software =
2025-11-20 View on X
Wired

Sunday Robotics unveils Memo, a fully autonomous home robot capable of tasks like making espresso and loading a dishwasher, and plans to launch in beta in 2026

Sunday Robotics has a new way to train robots to do common household tasks.  The startup plans to put its fully autonomous robots in homes next year.

2025-11-19
Wait, Gemini 3 was trained on TPUs?! [image]
2025-11-19 View on X
matt shumer

Gemini 3 hands-on: a fundamental improvement on daily use, extremely fast, Antigravity IDE is a powerful launch product, and its personality is terse and direct

Gemini 3 is a fundamental improvement on daily use, not just on benchmarks.  It feels more consistent and less “spiky” than previous models.

Wait, Gemini 3 was trained on TPUs?! [image]
2025-11-19 View on X
The Information

Gemini co-lead Oriol Vinyals says Gemini 3's gains come from better pre-training and post-training, contradicting the idea that pre-training gains are falling

which we discussed in our NeurIPS '25 talk with @ilyasut and @quocleix—the team delivered a drastic jump. The delta between 2.5 and 3.0 is [image] Andrej Karpathy / @karpathy : I p...

2025-10-21
Is it just me or DeepSeek OCR paper feels like something they've done a while ago but only decided to release now? What are they working on?? [image]
2025-10-21 View on X
The Decoder

DeepSeek releases DeepSeek-OCR, a vision language model designed for efficient vision-text compression, enabling longer contexts with less compute

the new frontier of OCR from @deepseek_ai , exploring optical context compression for LLMs, is running blazingly fast on vLLM ⚡ (~2500 tokens/s on A100-40G) — powered by vllm==0.8....

2025-03-10
I think Manus is here to destroy the $20,000 / month subscription. And the $200/month subscription. 🤯 [video]
2025-03-10 View on X
Forbes

A look at Manus, which its Chinese creators claim is the world's first fully autonomous AI agent, as some say it might be China's second DeepSeek moment

Manus is a general AI agent that bridges minds and actions: it doesn't just think, it delivers results. Luiza Jarovsky / Luiza's Newsletter : ✋ Manus AI: Why Everyone Should Worry ...

Can I say something w/out everyone getting upset? Manus is an exciting opp, especially as healthy competition against the monopoly wannabes. But it's young & needs time. If we want it to grow, we should give the team space to keep working on it. Overhyping can be detrimental.
2025-03-10 View on X
TechCrunch

Some early Manus users say the agentic AI is no panacea, with long waits, errors, unsatisfying answers, and endless loops often plaguing the experience

Manus, an “agentic” AI platform that launched in preview last week, is generating more hype than a Taylor Swift concert.

The Manus team counting down to the launch. Another mostly domestically educated AI team making global waves - Manus CEO Xiao Hong, born after 90's, educated in Huazhong Univ of Science & Technology - again not Tsinghua or Peking U. Something is shifting w this gen in China. [video]
2025-03-10 View on X
Forbes

A look at Manus, which its Chinese creators claim is the world's first fully autonomous AI agent, as some say it might be China's second DeepSeek moment

Manus is a general AI agent that bridges minds and actions: it doesn't just think, it delivers results. Luiza Jarovsky / Luiza's Newsletter : ✋ Manus AI: Why Everyone Should Worry ...