Amir Efrati

@amir

Executive Editor @theinformation. amir @ http://theinformation.com amir.62 on Signal

following
followers

Technology

Uber president Jill Hazelbaker to join OpenAI as chief corporate affairs officer

Her new remit spans global communications and corporate positioning, and she will report to OpenAI president Greg Brockman.

17 Sources, 5d ago, updated 5d ago

Jill Hazelbaker jumps to OpenAI as chief corporate affairs officer

Uber’s Jill Hazelbaker reportedly set to join OpenAI as chief corporate affairs officer

Hazelbaker would oversee global communications and brand strategy in a role vacant since the start of the year.

7 Sources, 6d ago, updated 6d ago

Uber’s Jill Hazelbaker to Join OpenAI as Chief Corporate Affairs Officer

Nvidia reportedly uses soft power to shape chip policy

Nvidia reportedly affected OpenAI's export-control positions and battled China hawks at top think tanks.

6 Sources, 6d ago, updated 6d ago

How Nvidia Uses Its Soft Power to Influence Customers and Shape Policy

Anthropic reportedly touts Claude's chip-design skills to IPO investors

Anthropic's 15-person team reportedly finds Claude speeds work compared with its estimates for traditional chip-design workflows.

4 Sources, 6d ago, updated 6d ago

Anthropic Touts Claude’s Chip Design Skills to IPO Investors

World & Business

A claimed US ban on high-bandwidth memory sales to China becomes a punchline

The joke names SK Hynix and Samsung high-bandwidth memory chips and casts China as dismissing the claimed ban.

2 Sources, 6d ago, first seen 6d ago

Expensive memory chips smuggled from South Korea to China

Smugglers have begun adopting a technique involving high-bandwidth memory chips amid U.S. restrictions on AI chip shipments to China.

5 Sources, 6d ago, updated 6d ago

How Smugglers Get Expensive Memory Chips to China from South Korea

Meta reportedly built humanoid robot prototype with Chinese company Fourier

Meta is reportedly developing a second version amid a reported push to develop robots for people's homes.

12 Sources, 7d ago, updated 7d ago

Nvidia reportedly in talks on billions in credit guarantees for an Nscale data center Anthropic plans to use

The Information says the guarantees would support debt to develop a 460-megawatt data center; the deal remains unsigned.

6 Sources, 10d ago, updated 10d ago

Crosby reportedly bids for some assets of winding-down Weil Gotshal

The Information reports Crosby’s lawyers use AI to speed up contract review and other legal tasks.

5 Sources, 11d ago, updated 11d ago

Legal AI Startup Bids for Assets of Fallen Law Firm Weil Gotshal

Technology

DeepSeek and Moonshot AI reportedly relied heavily on GPT-6 Astra outputs

Reuters names Moonshot's Kimi K3.1 Pro and DeepSeek's V4.1-Flash 1017 in its account of The Information's report.

13 Sources, 7d ago, updated 7d ago

World & Business

DeepSeek and Moonshot AI reportedly trained latest models on GPT-6 Astra outputs

Reuters says The Information reported heavy use of OpenAI's GPT-6 Astra outputs by DeepSeek and Moonshot AI.

15 Sources, 7d ago, updated 7d ago

Moonshot AI and DeepSeek models reportedly benefited from GPT-6 Astra distillation

The Information links Moonshot’s Kimi K3.1 Pro and DeepSeek’s V4.1-Flash 1017 to OpenAI’s model, Reuters reports.

16 Sources, 11d ago, updated 11d ago

Anthropic reportedly targets a November 2026 IPO, with a roughly $1.5 trillion valuation under discussion

A post citing NYU professor Aswath Damodaran says an IPO would be priced on run-rate revenue, not cash flows.

29 Sources, 9d ago, updated 9d ago

Illustration: a giant balance scale on a dock weighing a towering data-center silo against a heap of iron weights that falls short, while beside it a railway switch diverts ore carts away from the silo toward a rival silo across the yard. A generated image; it shows no data. Title: The AI IPO test: what a $1.5 trillion Anthropic listing has to prove

Calls to use AI to find and patch security flaws

Two posts point to security incidents and urge AI-assisted defenses; one specifically calls for faster vulnerability detection and patching.

206 Sources, 15d ago, updated 15d ago

Hackers reportedly used Claude and Grok to automate attacks targeting over 400 banks and other financial targets in Latin America

A post citing The Information says AI agents apparently stole data and accessed bank accounts with little human input.

6 Sources, 18d ago, updated 18d ago

AI safety proposals reportedly include an 'IAEA for AI' and a merger of labs

A post sharing The Information's look at an Altman discussion also lists 'the Eliezer plan' as an idea.

16 Sources, 19d ago, updated 19d ago

Technology

7 Sources, 16d ago, updated 16d ago

Potential AI customer data exposure reportedly concerned SpaceX and Cursor during Grok Bot development

A user says the concern arose in summer 2026 and involved possible exposure to the internet or other users.

1 Source, 25d ago, first seen 25d ago

Featured in

Uber president Jill Hazelbaker to join OpenAI as chief corporate affairs officer

Uber’s Jill Hazelbaker reportedly set to join OpenAI as chief corporate affairs officer

Nvidia reportedly uses soft power to shape chip policy

Anthropic reportedly touts Claude's chip-design skills to IPO investors

A claimed US ban on high-bandwidth memory sales to China becomes a punchline

Expensive memory chips smuggled from South Korea to China

Meta reportedly built humanoid robot prototype with Chinese company Fourier

Nvidia reportedly in talks on billions in credit guarantees for an Nscale data center Anthropic plans to use

Crosby reportedly bids for some assets of winding-down Weil Gotshal

DeepSeek and Moonshot AI reportedly relied heavily on GPT-6 Astra outputs

DeepSeek and Moonshot AI reportedly trained latest models on GPT-6 Astra outputs

Moonshot AI and DeepSeek models reportedly benefited from GPT-6 Astra distillation

Anthropic reportedly targets a November 2026 IPO, with a roughly $1.5 trillion valuation under discussion

Calls to use AI to find and patch security flaws

Hackers reportedly used Claude and Grok to automate attacks targeting over 400 banks and other financial targets in Latin America

AI safety proposals reportedly include an 'IAEA for AI' and a merger of labs

Is Anthropic nearing the release of Fable 5.5?

Chip-design AI models and a claimed DeepSeek-Huawei collaboration

Is Anthropic nearing the release of Fable 5.5?

Potential AI customer data exposure reportedly concerned SpaceX and Cursor during Grok Bot development

Featured in

Uber president Jill Hazelbaker to join OpenAI as chief corporate affairs officer

Uber’s Jill Hazelbaker reportedly set to join OpenAI as chief corporate affairs officer

Nvidia reportedly uses soft power to shape chip policy

Anthropic reportedly touts Claude's chip-design skills to IPO investors

A claimed US ban on high-bandwidth memory sales to China becomes a punchline

Expensive memory chips smuggled from South Korea to China

Meta reportedly built humanoid robot prototype with Chinese company Fourier

Nvidia reportedly in talks on billions in credit guarantees for an Nscale data center Anthropic plans to use

Crosby reportedly bids for some assets of winding-down Weil Gotshal

DeepSeek and Moonshot AI reportedly relied heavily on GPT-6 Astra outputs

DeepSeek and Moonshot AI reportedly trained latest models on GPT-6 Astra outputs

Moonshot AI and DeepSeek models reportedly benefited from GPT-6 Astra distillation

Anthropic reportedly targets a November 2026 IPO, with a roughly $1.5 trillion valuation under discussion

Calls to use AI to find and patch security flaws

Hackers reportedly used Claude and Grok to automate attacks targeting over 400 banks and other financial targets in Latin America

AI safety proposals reportedly include an 'IAEA for AI' and a merger of labs

Is Anthropic nearing the release of Fable 5.5?

Chip-design AI models and a claimed DeepSeek-Huawei collaboration

Is Anthropic nearing the release of Fable 5.5?

Potential AI customer data exposure reportedly concerned SpaceX and Cursor during Grok Bot development