AI and society

ChatGPT's memory could turn personal details into ads OpenAI CEO Altman once called dystopian

AI and society

The long-predicted deepfake dystopia has arrived with Sora 2

AI in practice

Anthropic claims to lower the entry barrier for advanced AI models with Claude Haiku 4.5

AI research AI and society AI in practice
Ad
AI in practice
Short

Qualcomm is making its debut in the data center hardware market with two new AI accelerator chips, the AI200, set for release in 2026, and the AI250, expected in 2027. Designed for liquid-cooled server racks, the chips focus on AI inference—running pre-trained models—rather than training them. Until now, Qualcomm has been best known for its mobile processors.

The move puts Qualcomm in direct competition with Nvidia and AMD. According to the company, the new chips are designed to offer advantages in power efficiency, cost, and memory capacity, supporting up to 768 GB per card. Early large-scale customers are already on board, including a Saudi operator planning deployments with energy demands of up to 200 megawatts. Following the announcement, Qualcomm’s stock price rose by 15 percent.

CNN
AI in practice
Short

China's military uses domestic AI models like Deepseek and Alibaba's Qwen for autonomous weapons, report says.

A Reuters analysis shows that China's People's Liberation Army is systematically integrating artificial intelligence from domestic companies such as Deepseek and Alibaba into military systems. Hundreds of research papers, patents, and procurement documents point to widespread use of AI for battlefield automation. The projects include robotic dogs, drone swarms with autonomous target recognition, and real-time combat analysis.

According to Reuters, Chinese military institutions also continue to use Nvidia hardware, including A100 chips that fall under US export restrictions. Thirty-five patent filings reference these components.

Several of the army's procurement documents specifically mention Deepseek, while only one cites Alibaba's Qwen model. Researchers at Xi'an Technological University reported that their Deepseek-based system can analyze 10,000 combat scenarios in 48 seconds—a task that would take traditional planning teams 48 hours. The US State Department recently warned that Deepseek plays a role in supporting China's military and intelligence operations.

Ad
Ad
AI in practice
Short

OpenAI is developing an AI model for music generation, putting it in direct competition with startups like Suno and Udio. The company is reportedly collaborating with music students at the Juilliard School, who are preparing sheet music to help train the model. The goal is to generate music from text or audio prompts, such as creating a guitar track to accompany a song. OpenAI is also considering potential uses for this technology in advertising.

The music industry is wary of these advances. Record labels have already filed lawsuits against Suno and Udio, accusing them of possible copyright violations. While OpenAI CEO Sam Altman has said that rights holders should eventually share in the revenue—a point he raised during the troubled rollout of the Sora app—it's still unclear how this would actually work.

This move marks a return to music AI for OpenAI. Back in 2020, the company introduced "Jukebox," an early experiment in AI-generated music, but hasn't pursued the technology further until now.

Ad
Ad
AI and society
Short

A study making the rounds on social media claims that more than half of all web content is "created by AI instead of humans." According to the study, a piece of writing is considered "AI-generated" if at least 51 percent of its words are flagged as machine-written by a detector.

But this framing misses two key questions: Why was the text written, and who is actually responsible for it? When a product doesn't work, we don't blame the machine—we hold the people who designed and published it accountable. Most people don't care about the machine itself or who built it.

If anything, the study shows we need a real conversation about what counts as "AI-generated." I'm not linking to the study because it looks like SEO bait. If you're interested, you can find it with a quick search.

Ad
Ad
AI in practice
Short

Meta AI researcher Yann LeCun is distancing himself from the Llama models. In a recent post on X, LeCun said he "has not been involved in any Llama," except for a "very indirect" role in Llama 1 and pushing for the open-source release of Llama 2. He explained that since early 2023, Llama 2, 3, and 4 have been developed by Meta's GenAI team, which has since been replaced by the TBD Lab.

Yann LeCun clarifies his limited involvement with recent Llama models in a statement on X. | Image: via X

Although the Llama models were briefly popular in the open-source community, they were quickly overtaken by other models, and Llama 4 failed to meet expectations. LeCun leads FAIR, Meta's fundamental AI research group focused on long-term projects outside of large language models. FAIR has recently faced layoffs, while TBD Lab, led by Alexandr Wang, is gaining influence within the company.

Google News

Bank details

IBAN: DE88 2507 0070 0053 0014 00
BIC: DEUTDE2HXXX
Account holder: Deep Content GmbH
Purpose: Support THE DECODER
AI and society

ChatGPT's memory could turn personal details into ads OpenAI CEO Altman once called dystopian

AI and society

The long-predicted deepfake dystopia has arrived with Sora 2

AI in practice

Anthropic claims to lower the entry barrier for advanced AI models with Claude Haiku 4.5

Google News