AI in practice Archive

Jan 17, 2026

Chinese AI startup DeepSeek ran into trouble developing its new flagship model and had to switch to Nvidia chips. According to insiders, Deepseek initially tried using chips from Huawei and other Chinese manufacturers last year, reports the Wall Street Journal. But the results weren't good enough. The company ended up switching to allegedly smuggled Nvidia chips for some training tasks, which finally got things moving. The new model is expected to ship in the coming weeks.

At a recent conference in Beijing, leading Chinese AI researchers admitted that Chinese AI models won't be able to keep pace with US companies without access to better hardware. Justin Lin from Alibaba's Qwen team put the odds of overtaking OpenAI or Anthropic within three to five years at 20 percent at best. Meanwhile, the Chinese government is pushing to cut US chip imports to boost domestic production.

Comment Source: WSJ

Jonathan Kemper

Jan 17, 2026

AI in practice

Flux 2 small brings AI image generation and editing to consumer graphics cards

Collage of photorealistic AI images: Portraits, landscapes, animals and city scenes with a high level of detail

Matthias Bastian

Jan 16, 2026

Short News

OpenAI is pushing "Open Responses," an open interface that works with language models from different providers. The project builds on OpenAI's Responses API and lets developers write code once and run it with any AI model.

Currently, Google, Anthropic, and Meta all handle their APIs differently, which means developers have to rewrite code when switching between models. Open Responses tries to fix that with a shared format for requests, responses, streaming, and tool calls. Vercel, Hugging Face, LM Studio, Ollama, and vLLM have already signed on.

Of course, if successful, this move works in OpenAI's favor. If its API becomes the default, competitors would need to adapt to OpenAI's approach, while existing OpenAI customers wouldn't have to change a thing. The "open" label also lets the company signal a spirit of collaboration, even though it's not sharing any technology beyond what's already public.

Comment Source: OpenAI via X | Open Responses

Matthias Bastian

Jan 16, 2026

AI in practice

OpenAI will soon test ads in ChatGPT despite CEO Sam Altman once calling the idea dystopian

OpenAI will start testing ads in ChatGPT, despite CEO Sam Altman’s earlier objections. With a valuation of up to $750 billion to justify and only around five percent of users paying for the service, the company is under enormous pressure to find new revenue streams.

Read full article

Comment

Matthias Bastian

Jan 16, 2026

Short News

Anthropic has expanded access to its new Claude Cowork feature. When Cowork debuted on Monday, it was limited to Max subscribers, a $200 per month tier that put it out of reach for most users. Pro subscribers can now access the feature for $20 per month, though Anthropic warns they may hit usage limits faster since Cowork consumes more tokens than regular chat. Max subscribers still get higher usage limits.

Cowork brings the agent-based capabilities of Claude Code to the desktop app for everyday tasks that don't require programming knowledge. With computer access enabled, Claude can handle more complex tasks on its own: sorting files, gathering context from multiple documents, and similar workflows. For now, the feature remains exclusive to the macOS desktop app.

Anthropic has already shipped several updates since Monday's launch: users can now rename sessions, connections to external services are more reliable, file previews work better, and the app prompts for confirmation before deleting files.

Comment

Maximilian Schreiner

Jan 16, 2026

AI in practice

Cloudflare is acquiring British startup Human Native to create a new payment system for AI training data. The company runs a marketplace for AI training data and converts multimedia content into structured, licensable datasets.

The move addresses a growing problem: AI crawlers scrape the web on a massive scale without paying website operators. Instead of letting AI companies grab content for free, publishers could make their data available through an index and get paid for it.

Cloudflare has already been developing tools like "AI Crawl Control" and "Pay Per Crawl" that let website operators control who can use their content for AI training. The company also co-founded the x402 Foundation with Coinbase to enable automatic machine-to-machine payments. Cloudflare runs its own AI platform and recently expanded it by acquiring Replicate.

Comment Source: Cloudflare | Human Native

Matthias Bastian

Jan 15, 2026

Short News

Wikipedia has landed major AI companies as partners: Amazon, Meta, Microsoft, Mistral AI, and Perplexity have joined the Wikimedia Enterprise partner program, with Google, Ecosia, and others already on board. These companies use the APIs to integrate Wikipedia content into their products.

Wikipedia is considered one of the highest-quality datasets for training large language models, and its content powers chatbots, search engines, and voice assistants. The Wikimedia Foundation argues that human-curated knowledge is more valuable than ever in the AI era, but without financial contributions from companies profiting from this data, the open knowledge model could be at risk.

In late October, Wikipedia raised concerns about declining traffic from AI systems that display its content without sending users to the website and later called for fair licensing through its API.

This tension is likely to grow. Chatbots are extracting value from the web at scale, and the legal landscape remains murky. Not every site can follow Wikipedia's path, or that of Tailwind, by offsetting lost revenue through partnerships and paid APIs.

Comment Source: Wikimedia Enterprise

Jonathan Kemper

Jan 15, 2026

AI in practice

Google's new open TranslateGemma models bring translation for 55 languages to laptops and phones

TranslateGemma shows how targeted training helps Google squeeze more performance out of smaller models: the 12B version translates better than a base model twice its size and runs on a regular laptop. With the growing Gemma family, Google is staking its claim in the race for open AI models.

Read full article

Comment

Maximilian Schreiner

Jan 15, 2026

Update

Short News

OpenAI reportedly walked away from Apple to focus on building its own AI hardware instead

Apple is betting billions on Google to power Siri’s AI future, and OpenAI says it walked away from the deal on purpose.

Read full article

Comment

Matthias Bastian

Jan 15, 2026

Short News

What China can do, the US can do too: OpenAI has published a call for proposals to boost domestic AI hardware production. AI relies on a broad ecosystem of physical components beyond chips, OpenAI says. The company is seeking manufacturers and suppliers of data center components like cooling systems, power supplies, and networking equipment, plus consumer electronics and robotics. Applications are open through June 2026.

The move comes as China reportedly restricts Nvidia H200 imports and pushes domestic manufacturers to source hardware locally. If China succeeds in decoupling its supply chain, the US can expect countermeasures—and if both countries want true independence, they'll each have to build their own.

The initiative fits squarely with Trump's "America First" agenda. OpenAI frames it as "reindustrialization of the country." Notably, OpenAI President Greg Brockman donated $25 million to Trump's campaign.

Comment Source: OpenAI