Robots to Retrieve Radioactive Sandbags at Fukushima Plant 

The tsunami-crippled Fukushima Daiichi nuclear power plant is seen from Namie Town, Fukushima prefecture, Japan August 24, 2023, in this photo taken by Kyodo. (Kyodo/via Reuters)
The tsunami-crippled Fukushima Daiichi nuclear power plant is seen from Namie Town, Fukushima prefecture, Japan August 24, 2023, in this photo taken by Kyodo. (Kyodo/via Reuters)
TT
20

Robots to Retrieve Radioactive Sandbags at Fukushima Plant 

The tsunami-crippled Fukushima Daiichi nuclear power plant is seen from Namie Town, Fukushima prefecture, Japan August 24, 2023, in this photo taken by Kyodo. (Kyodo/via Reuters)
The tsunami-crippled Fukushima Daiichi nuclear power plant is seen from Namie Town, Fukushima prefecture, Japan August 24, 2023, in this photo taken by Kyodo. (Kyodo/via Reuters)

Robots will begin moving sandbags that were used to absorb radiation-contaminated water after the 2011 Fukushima nuclear disaster as soon as next week, a spokesman for the plant operator said Friday.

TEPCO, the operator of the stricken Japanese power plant, says the bags on underground floors of two buildings have been left untouched following the worst nuclear accident since Chernobyl.

Radiation levels on the sandbags' surface are as high as 4.4 sieverts per hour, which means "humans can die if they approach" them, TEPCO spokesman Tatsuya Matoba told AFP.

Japanese media reports said there were 2,850 bags to be collected, a number which has not been confirmed by TEPCO, which says that they weigh 41.5 tons (91,500 pounds) in total.

Two robots developed to collect the bags, one with a moving claw, were on Wednesday placed on the underground floors, Matoba said.

Workers will use them to "carefully" bring the sandbags out in an operation that TEPCO aims to finish by the end of the 2027 fiscal year.

The bags will then be placed inside containers for radioactive material and kept at a temporary storage site outside the buildings, the spokesman said.

Three of Fukushima's six reactors went into meltdown 14 years ago after a huge tsunami swamped the facility.

The tsunami, triggered by a 9.0-magnitude earthquake, left 18,500 people dead or missing.

No one was recorded as having been directly killed by the nuclear accident, which forced evacuations and left parts of the surrounding area uninhabitable.

In addition to contaminated sandbags, around 880 tons of radioactive debris remain in the plant.

Removing this is seen as the most daunting challenge in the decades-long decommissioning project because of the dangerously high radiation levels involved.

A trial removal of nuclear debris from the plant began last year.



Huawei's AI Lab Denies that One of its Pangu Models Copied Alibaba's Qwen

FILE PHOTO: Huawei logo is seen during Munich Auto Show, IAA Mobility 2021 in Munich, Germany, September 8, 2021. REUTERS/Wolfgang Rattay/File Photo
FILE PHOTO: Huawei logo is seen during Munich Auto Show, IAA Mobility 2021 in Munich, Germany, September 8, 2021. REUTERS/Wolfgang Rattay/File Photo
TT
20

Huawei's AI Lab Denies that One of its Pangu Models Copied Alibaba's Qwen

FILE PHOTO: Huawei logo is seen during Munich Auto Show, IAA Mobility 2021 in Munich, Germany, September 8, 2021. REUTERS/Wolfgang Rattay/File Photo
FILE PHOTO: Huawei logo is seen during Munich Auto Show, IAA Mobility 2021 in Munich, Germany, September 8, 2021. REUTERS/Wolfgang Rattay/File Photo

Huawei's artificial intelligence research division has rejected claims that a version of its Pangu Pro large language model has copied elements from an Alibaba model, saying that it was independently developed and trained.

The division, called Noah Ark Lab, issued the statement on Saturday, a day after an entity called HonestAGI posted an English-language paper on code-sharing platform Github, saying Huawei's Pangu Pro Moe (Mixture of Experts) model showed "extraordinary correlation" with Alibaba's Qwen 2.5 14B.

This suggests that Huawei's model was derived through "upcycling" and was not trained from scratch, the paper said, prompting widespread discussion in AI circles online and in Chinese tech-focused media.

According to Reuters, the paper added that its findings indicated potential copyright violation, the fabrication of information in technical reports and false claims about Huawei's investment in training the model.

Noah Ark Lab said in its statement that the model was "not based on incremental training of other manufacturers' models" and that it had "made key innovations in architecture design and technical features." It is the first large-scale model built entirely on Huawei's Ascend chips, it added.

It also said that its development team had strictly adhered to open-source license requirements for any third-party code used, without elaborating which open-source models it took reference from.

Alibaba did not immediately respond to a Reuters request for comment.

Reuters was unable to contact HonestAGI or learn who is behind the entity.

The release of Chinese startup DeepSeek's open-source model R1 in January this year shocked Silicon Valley with its low cost and sparked intense competition between China's tech giants to offer competitive products.

Qwen 2.5-14B was released in May 2024 and is one of Alibaba's small-sized Qwen 2.5 model family which can be deployed on PC and smartphones.

While Huawei entered the large language model arena early with its original Pangu release in 2021, it has since been perceived as lagging behind rivals. It open-sourced its Pangu Pro Moe models on Chinese developer platform GitCode in late June, seeking to boost the adoption of its AI tech by providing free access to developers.

While Qwen is more consumer-facing and has chatbot services like ChatGPT, Huawei's Pangu models tend to be more used in government as well as the finance and manufacturing sectors.