Deezer Sees Surge in AI-Generated Music, Now Receiving 20,000 Tracks Daily

Nearly 20% of all new music uploaded to streaming platform Deezer is now AI-generated, with over 20,000 such tracks submitted each day, the company announced.

This marks a sharp increase from January, when AI-generated songs made up just 10% of daily uploads. Deezer said the volume has nearly doubled in three months.

Chief Innovation Officer Aurelien Herault said AI-generated content continues to flood platforms like Deezer, with no signs of slowing.

To manage the influx, Deezer has implemented automated detection tools capable of identifying fully AI-generated songs created using the most common generative tools.

However, the company noted that many tracks are a blend of human and AI input, making detection more complex.

A major concern is that some AI-generated music may be trained on copyrighted material without permission.

Deezer is the only streaming platform to sign a global statement on AI training, which opposes the unlicensed use of creative works for training generative AI.

While AI-generated music remains a minority among the more than 100,000 songs uploaded daily, Deezer warned that it could soon become the majority if current trends continue.

In January, Deezer said it had made significant progress in developing what it calls the best AI detection tool on the market, aimed at protecting the estimated USD 16 billion in annual music creator revenue.

The company estimates that up to 25% of that revenue could be at risk by 2028 due to AI-generated content.

There are now more than 100 generative AI tools for music creation, including text-to-music platforms like YouTube’s Music Assistant, Google’s Dream Track, and Meta’s MusicGen.

Other tools include OpenAI’s MuseNet and Jukebox, composition software like AIVA and MusicLM, and production plugins such as LANDR, TAIP, and Synthesizer V.

Voice-cloning and AI vocal tools like Suno, Udio, and Supertone are also contributing to the surge.

Market.us projects the global AI in music market will grow from nearly USD 4 billion in 2023 to USD 39 billion by 2033, a 26% compound annual growth rate.

The report also states that 60% of musicians are now using AI in their creative and production processes.

This suggests that many of the same artists Deezer aims to protect may also be contributing to the rise of AI-generated music.

Deezer draws a clear line at fully AI-generated content. Herault said the company supports responsible development of generative AI to protect artists’ rights and maintain transparency for listeners.

He added that Deezer is already removing fully AI-generated songs from its algorithmic recommendations.

---

• Deezer now receives over 20,000 AI-generated songs daily

• AI music makes up nearly 20% of all new uploads

• Deezer uses detection tools to identify and manage AI content

Source: FORBES

AI Companions Disrupt China’s Adult Entertainment and Dating Industries

Artificial intelligence is rapidly transforming China’s adult entertainment sector, with AI-powered emotional robots reshaping traditional concepts of intimacy, companionship, and even marriage.

Credit: EX ROBOT

Nicknamed “China Dolls,” these AI companions have triggered a surge in the A-share market, boosting the stock prices of upstream material suppliers, midstream algorithm developers, and downstream assembly plants.

The dolls, produced primarily in China, now account for 70% of global output. Unlike traditional silicone models, these AI-enhanced companions feature realistic textures, body temperature simulation, and emotional interactivity.

Their “souls” are powered by large AI models such as Deepseek, enabling them to remember conversations for up to three months and switch between eight personality types, including cute, shy, and enthusiastic.

These AI partners can engage in conversations about work, science, and philosophy, offering a different experience each day. Their rapid evolution mirrors the pace of internet product updates, with new emotional features added weekly.

The trend is also impacting Japan’s long-established adult industry, which still relies on traditional craftsmanship. In contrast, Chinese manufacturers are integrating silicone, algorithms, and internet-driven design to dominate the market.

The rise of AI companions is influencing broader social behaviours. Many young Chinese are opting out of traditional relationships, marriage, and parenthood, drawn instead to AI partners that require no emotional labour, financial burden, or long-term commitment.

These companions do not age, gain weight, or demand attention. They remember personal preferences, such as disliked coffee flavours, and adjust home settings like lighting based on user routines.

The phenomenon reflects the convergence of the loneliness economy, the silver economy, and the otaku economy, making silicone and algorithms essential emotional commodities.

While technology is not solely responsible for declining marriage rates, the high cost of relationships in urban China—often reaching millions—contrasts sharply with the affordability and reliability of AI partners.

An AI companion can be acquired for the equivalent of one or two months’ salary, offering instant emotional gratification without the risks of betrayal or incompatibility.

As AI-generated personas become more sophisticated, traditional dating rituals—such as betrothal gifts and years of courtship—are being replaced by code-driven emotional fulfilment.

---

• AI-powered dolls dominate 70% of global production, led by Chinese manufacturers

• Emotional robots feature realistic bodies, memory retention, and switchable personalities

• Surge in A-share market driven by AI companion supply chain

Huawei Unveils AI Supernode to Rival Nvidia’s NVL72 in Computing Power Race

Huawei Technologies has launched a new artificial intelligence infrastructure architecture that reportedly rivals Nvidia’s top-tier offerings, marking a significant step in China’s push for tech self-sufficiency.

Credit: Jason Alden/Bloomberg

The CloudMatrix 384 Supernode, introduced last week, is said to match Nvidia’s NVL72 system in addressing computing bottlenecks in AI data centres, according to STAR Market Daily.

Huawei’s supernode, currently deployed in its data centres in Wuhu, Anhui province, reportedly delivers 300 petaflops of computing power, surpassing the 180 petaflops offered by Nvidia’s NVL72.

Nvidia’s NVL72, launched in March 2023, features a 72-GPU NVLink domain that functions as a single GPU, enabling real-time inference for trillion-parameter large language models at speeds 30 times faster than previous generations.

Huawei’s CloudMatrix 384 Supernode is part of its broader CloudMatrix infrastructure, launched in September to meet growing demand for computing capacity amid the global AI boom.

Supernodes are AI infrastructure systems equipped with enhanced resources, including central and neural processing units, network bandwidth, storage and memory, allowing them to boost cluster performance and accelerate foundational model training.

Huawei is reportedly working with Chinese AI infrastructure start-up SiliconFlow to support DeepSeek-R1, a reasoning model developed by the Hangzhou-based firm and launched in January.

The collaboration suggests Huawei’s system, powered by Chinese-made chips, has achieved performance on par with Nvidia’s H100 chips, reaching a throughput of 1,920 tokens per second with high accuracy.

The development underscores Huawei’s progress in overcoming US sanctions and advancing domestic computing capabilities amid ongoing US-China tech tensions.

Other Chinese firms are also ramping up AI infrastructure investments. In Feb., Alibaba Group Holding announced a capital expenditure of 380 billion yuan (USD 52.4 billion) over the next three years for computing and AI infrastructure.

This marks the largest-ever investment by a private Chinese company in a computing project.

---

• Huawei launched the CloudMatrix 384 Supernode to rival Nvidia’s NVL72

• The system delivers 300 petaflops, surpassing Nvidia’s 180 petaflops

• Huawei is collaborating with SiliconFlow to support DeepSeek-R1

Source: SCMP