Few things prompt as much anxiety in science and the wider world as the growing use of artificial intelligence (AI) and the rising influence of China. This spring, these two factors created a rift at the European Geosciences Union (EGU), one of the world’s largest geoscience societies, that led to the firing of its president.
The whole episode has been “a packaging up of fear of AI and fear of China,” says Michael Stephenson, former chief geologist of the United Kingdom and one of the founders of Deep-time Digital Earth (DDE), a $70 million effort to connect digital geoscience databases. In 2019, another geoscience society, the International Union of Geological Sciences (IUGS), kicked off DDE, which has been funded almost entirely by the government of China’s Jiangsu province.
The dispute pivots on GeoGPT, an AI-powered chatbot that is one of DDE’s main efforts. It is being developed by Jian Wang, chief technology officer of e-commerce giant Alibaba. Built on Qwen, Alibaba’s own chatbot, and fine-tuned on billions of words from open-source geology studies and data sets, GeoGPT is meant to provide expert answers to questions, summarize documents, and create visualizations.
Stephenson tested an early version, asking it about the challenges of using the fossilized teeth of conodonts, an ancient relative of fish, to define the start of the Permian period 299 million years ago. “It was very good at that,” he says.
As awareness of GeoGPT spread, so did concern. Paul Cleverly, a visiting professor at Robert Gordon University, gained access to an early version and said in a recent editorial in Geoscientist there were “serious issues around a lack of transparency, state censorship, and potential copyright infringement.” A group of publishers, led by Phoebe McMellon, CEO of GeoScienceWorld, wrote to IUGS in February, arguing that GeoGPT had been illegally built off unlicensed literature and was not transparent because it did not cite its sources. “Just because all these commercial companies feel OK to scrape stuff and ask for permission later, that doesn’t mean we in geosciences have to do the same,” McMellon says.
But the letter did not cite specific copyright violations, and so DDE President Chengshan Wang, a geologist at the China University of Geosciences, did not feel compelled to end the project. To address transparency concerns, DDE said it would not make GeoGPT widely available until it could cite sources for its answers. Meanwhile, DDE had arranged a session at EGU’s April meeting, including a talk that would introduce GeoGPT. But weeks before the event, EGU received a complaint about the abstract that made arguments similar to those in the publishers’ letter to IUGS.
It arrived at an EGU whose leadership was already under strain. A year earlier, Irina Artemieva, who is Russian born but left the country decades ago, had taken over as president. An expert in continental interiors, Artemieva is affiliated with Germany’s GEOMAR Helmholtz Centre for Ocean Research Kiel but is also paid by the Chinese Academy of Geological Sciences to advise it on its geophysical research. Although she was elected by EGU’s members, she says EGU’s leadership didn’t want her—especially after Russia’s invasion of Ukraine. “They were looking for opportunities to undermine my authority,” she says.
Artemieva says the GeoGPT complaint was “complete arm waving.” “Nothing was documented,” she says. But other executives sought to withdraw the conference abstract right away. “To them, the only thing that appeared to matter is that it came from China,” she says. Artemieva forwarded an email containing the complaint to Chengshan Wang, to get his view, but forgot to delete the name attached to it—a privacy violation.
Artemieva wanted an investigation before taking any action. But the week of the meeting, a group of EGU leaders, declaring themselves the “acting EGU ethics committee,” withdrew the abstract without her knowledge. Artemieva, who ordinarily would have led the ethics committee, was not included because of her confidentiality breach and a “possible conflict of interest,” says Hazel Gibson, EGU’s head of communications.
Things came to a head at the DDE session, which included a contingent of Chinese scientists and plans to ceremonially sign a memorandum of understanding with Springer Nature, the publishing giant. EGU’s vice president, Peter van der Beek, a geochronologist at Potsdam University, became agitated, several attendees later said, reprimanding attendees, ripping up a paper program that shuffled several virtual talks, and escorting two DDE members out of the room. Gibson says van der Beek was just enforcing the rules, because the attendees were unauthorized, and the program should not have been revised.
Several scientists filed complaints about the session and van der Beek’s behavior, claiming his actions might have constituted “harassment and discrimination.” The next day, he apologized in person to Chengshan Wang. But Wang says all session attendees deserve a similar apology from him. Any suggestion that he acted aggressively is a “gross mischaracterization, that anyone who knows Peter van der Beek will attest to,” says Gibson, who responded on van der Beek’s behalf after Science sent questions to him.
In May, EGU leaders met and voted to fire Artemieva, elevating van der Beek to president. Among other allegations, they said Artemieva was difficult to work with and had breached confidentiality when she forgot to delete the complainant’s name. They also said she threatened EGU’s executive secretary when she asked him for details of his employment contract after he called a meeting of the executive board without her approval.
Artemieva made a mistake in sharing the name, says Hans Thybo, a close friend of Artemieva’s and a former EGU president who works at Istanbul Technical University and two Chinese universities. But he says the leadership staged a “coup d’état in council to cover their own mistakes. … In reality, they should be dismissed.” To John Ludden, IUGS’s president and another former EGU president, the EGU leadership’s “behavior was a little too accusatory. … They weren’t willing to listen.”
Seeking to broker a peace deal around GeoGPT, Ludden has invited all parties to a workshop this month in London. He hopes DDE will agree to an international governance body for GeoGPT. Meanwhile, the GeoGPT team has already said users will be able to choose from Alibaba’s Qwen or Meta’s open-source Llama as the underlying AI. Chengshan Wang says it is also developing a licensing deal with Springer Nature.
China’s support of DDE is filling a void left by other countries, adds Robert Hazen, a mineralogist at the Carnegie Institution for Science who is a DDE grantee. And models like GeoGPT aren’t going away: Just last week, NASA announced its own space and earth science AI language model, called INDUS, which uses studies from the American Geophysical Union, among others. Although Hazen understands the wariness about China’s involvement, “I keep my eyes open all the time,” he says. “I sense no agenda whatsoever.”
More: https://www.science.org/content/article/chinese-ai-stirs-panic-european-geoscience-society
