Huawei aims to mass-produce newest AI chip in early 2025, despite US curbs

In partnership with

Pivot 5: 5 stories. 5 minutes a day. 5 days a week.

The fastest way to build AI apps

We’re excited to introduce Writer AI Studio, the fastest way to build AI apps, products, and features. Writer’s unique full-stack design makes it easy to prototype, deploy, and test AI apps – allowing developers to build with APIs, a drag-and-drop open-source Python framework, or a no-code builder, so you have flexibility to build the way you want.

Writer comes with a suite of top-ranking LLMs and has built-in RAG for easy integration with your data. Check it out if you’re looking to streamline how you build and integrate AI apps.

Start building Writer AI Studio.

1. Huawei aims to mass-produce newest AI chip in early 2025, despite US curbs

Huawei plans to begin mass-producing its most advanced artificial intelligence chip in Q1 2025, despite facing US restrictions. The company has sent samples of its Ascend 910C chip to some technology firms and started taking orders.

The 910C is being made by top Chinese contract chipmaker Semiconductor Manufacturing International Corp (SMIC), but a lack of advanced lithography equipment has limited its yield to around 20%. Advanced chips need yields of more than 70% to be commercially viable. The US has imposed curbs on Chinese companies, arguing their technological progress poses a national security risk.

Read the full story here

2. Claude Pro subscribers can now interact with Google Docs documents directly within the app

3. New FLUX.1 Tools

FLUX.1 Tools is a suite of models designed to enhance control and steerability in the FLUX.1 text-to-image model. It includes four features: FLUX.1 Fill, which offers advanced inpainting and outpainting capabilities, FLUX.1 Depth, which provides structural guidance based on depth maps, FLUX.1 Canny, which uses canny edges, and FLUX.1 Redux, an adapter for mixing and recreating input images and text prompts.

The tools are available as open-access models in the BFL API.

Read the full story here

4. How Did You Do On The AI Art Turing Test?

A recent challenge to 11,000 people involved identifying fifty pictures as human art or AI-generated images. The test, which included paintings by Domenichino, Gauguin, Basquiat, and others, revealed that most people had a hard time identifying AI art. The median score was 60%, with a mean of 60.6%.

The test was designed to be fair, excluding submissions with garbled text, misshapen hands, or other deformities. The test also excluded AI art in the DALL-E "house style" and warned test-takers that it included both human and AI art.

Read the full story here

5. Snowflake beats Databricks to integrating Claude 3.5 directly

Snowflake has partnered with AI vendor Anthropic to advance AI projects for its customers. The data ecosystem giant has signed a multi-year strategic agreement to bring Anthropic's Claude 3.5 family of models to its core platform on AWS.

This will allow enterprises using Snowflake to leverage the cutting-edge Claude LLMs for building advanced AI applications, including conversational agents. The partnership will also power Snowflake's customer-facing agentic offerings and help employees accelerate their internal workflows. Snowflake's partnership is similar to Databricks, which offers a range of open-source and proprietary models.

Read the full story here