Sources: DeepSeek CEO Liang Wenfeng says training on Huawei chips is one of DeepSeek's biggest bets and Huawei is set to deliver training chips in Q4 or Q1 2027 (The Information)

DeepSeek CEO Liang Wenfeng told investors that a major priority for his company is to use more domestic chips to train its models …

theinformation.com ·2026-09-21

We haven't written up this one. Techmeme has the full story — the link below goes straight to it.

This story

This is one outlet's version. Read the fullest account.

Read the original at theinformation.com →

More in AI

GOV.UK founder warns Britain is ‘dangerously dependent’ on foreign AI

Britain risks becoming “dangerously dependent” on a handful of foreign tech companies as AI spreads through businesses and public services, the founder of GOV.UK has warned. Mike Bracken, who founded the Government Digital Service and led the creation of GOV.UK, said organisations were steadily giving up control of critical systems as…

2026-09-21 ·cityam.com

FPB plans AI-powered oversight of streaming, social platforms

The content watchdog approves an AI strategy targeting content monitoring and classification, to better scrutinise online content and video-on-demand streaming platforms.

2026-09-21 ·itweb.co.za

US and China Discuss Alerting Each Other to AI National Security Threats

Officials discussed setting up a mechanism for the two countries to notify each other of AI incidents which could threaten national security.

2026-09-21 ·wired.com

Work as you know it will be a relic of the past

AI won’t replace knowledge workers overnight, but it will fundamentally change how they work.

2026-09-21 ·techradar.com

Vocci’s $249 AI Ring Records Meetings, Connects Notes to AI Agents

Vocci’s wearable recorder combines eight-hour battery claims with MCP access to meeting context, but businesses still face limits around administration, consent, and subscriptions. The post Vocci’s $249 AI Ring Records Meetings, Connects Notes to AI Agents appeared first on TechRepublic .

2026-09-21 ·techrepublic.com

Artificial Structure Function Search: Preserving Artificial Functional Connectivity for Structured Pruning

Structured pruning is a model compression technique that is used to reduce the computational cost of deploying deep neural networks on resource-constrained devices. Popular methods of pruning rely on opaque heuristics or weight-based criteria that give no indication as to the structural dependencies in the network.

2026-09-21 ·arxiv.org