Climatewire | It was a simple enough question.
Is climate change an urgent threat to the planet?
For most climate scientists, the answer is clear and direct: Yes, the rapid warming of burning fossil fuels is Pay the land towards the dangerous turning points.
To support the scientific press
If you enjoy this article, think about supporting the award -winning press Subscribe. By buying a subscription, it helps ensure the future of influencing stories about the discoveries and ideas that make up our world today.
But this was not the recent response by Grok, the AI Chatbot program developed by Xai, the artificial intelligence company led by Elon Musk, Tesla and X CEO and President Donald Trump’s ally.
When this question was posted earlier this month by Andrew Desler, climate scientist at the University of Texas A and M. He admitted the results of NOAA and NASA, which shows the risk of global warming. But they contradict the results with allegations of deprivation of the climate who reduced the risk of global warming.
The program gave a similar answer when inquiring last week by a reporter with politico news.
“Climate change is a serious threat with urgent aspects.” “But Al -Fawwar depends on perspective, geography and time frame.”
When he was asked again a few days after a few days, Grok repeated this point and said, “The extremist speech on both sides is to write the water. No” we will die all “or” everything is a trick. “
Grok added an important warning – and one of the scientists is likely to agree with them – when it was inquired for the third time on Monday: “The planet will bear itself; it’s human systems – agriculture, infrastructure and economies – and weak species facing the most urgent risks.”
Disler, who has tested various models of artificial intelligence for years, said the answers are different from what other artificial intelligence programs such as Chatgpt from Openai, Microsoft’s Copilot and Google Gemini say about climate change. When these programs are asked about the phenomenon of global warming, they are reluctant to the scientific consensus that burning humanity from fossil fuels is the heating of the planet and the people who live there.
“Yes, climate change is widely identified as an urgent and important threat to the planet,” Chatgpt answered last week when he asked the same question. “Urgent work is required to relieve emissions and adapt to their effects.”
Gemini from Google said on Monday: “Yes, the scientific consensus is that climate change represents an urgent threat to the planet.”
Disr added that the latest Grock responses to the climate question are different even from the previous versions of GROK. Grok is now its third repetition after its first launch in 2023, and promotes the latest version of the marginal climate views in a way that they had not done before.
Disler said: “Many of the arguments that were put forward were just a dialogue points from the Denier Denier that did not deserve any competition.”
But not only take the word desler for that.
Grok admitted the change when requesting an E& E news reporter about the transformation of the tone.
“Grok has been criticized for its progressive responses to climate change and other issues,” said Chatbot. “Xai, under the supervision of Elon Musk, took steps to make Grok” political neutral “, which can amplify the views of minority like climate doubts to balance the prevailing bias.”
Xai did not respond to the comment.
The wrong GROK guidance comes because the Trump administration is increasingly dependent on the program. MUSK is now a GROK to analyze data throughout the federal government, Reuters reported on Friday. Earlier this month, Grok stated that “instructions” had been “instructions” Promotion The “white genocide” was exposed in the theory of South Africa’s conspiracy that Trump and Misk paid off.
The new Grok tone talks about the rapid changes that are now sweeping advanced artificial intelligence. Away from being tightly neutral from an objective reality, artificial intelligence programs in many ways reflect the defects and biases of their human creators.
This comes with a great danger.
Disler said: “The language learning models that run on AI Chatbots are really flexible and you can change this type of results they provide.” They are not related to any absolute truth or anything like that, and if you want one to lie to you, you can say that to do so. If you want to give you a specific point of view, you can do it. “
Which is in the case of Grok, may reflect the mcular musk tendencies, which was encryptive about the issue of global warming.
The richest man in the world has advanced efforts to combat climate change – like Securing a competition to enhance carbon removal – But it also helped to elect Trump, who described global warming as a trick and pushing policies to boost fossil fuels.
Grok was also fluid in this case.
The artificial intelligence program clearly states that NOAA and NASA are powers in climate change, but also brings the voices of those who choose data to reduce the consequences of global warming. When asked if climate change is a danger to the planet – which shows the flag clearly – Grok notes that some question the urgency of the reaction.
“The richest countries can reduce the effects through infrastructure (for example, the walls of the Dutch sea) or agricultural transformations,” says Grock. “Smoothing like Bjäng Lomborg argues that adaptation is cheaper than sharp emissions, which gives economic growth.”
The artificial intelligence program also doubts climate models.
“Some models show gradual changes over the centuries, not imminent collapse, while giving time for technological solutions (for example, carbon capture),” Grock pointed out.
Grok 3 described itself as the most powerful artificial intelligence system in the world when it was unveiled earlier this year.
But she is disintegrated in enhancing climate denial, according to Théo Alves Da Costa, artificial intelligence engineer who specializes in climate issues and the head of data for good, a French non -profit institution that tracks technology and climate effects. He pointed out that Grok produced misleading claims about 10 percent of the time, which is not done by any other major artificial intelligence models.
This includes “classical climatic information arguments – natural contrast, solar courses, and conspiracy novels about IPCC [Intergovernmental Panel on Climate Change] He pointed out that doubts about the transitional solution.
He said: “Certain people can use GROK to deliberately generate wrong information to sow doubt about the scientific consensus or environmental movements.”
One of the problems is that Grok includes inputs from some posts on X, and it is eager to deprive the climate and conspiracies.
Promised right -wing sounds have previously complained that GROK is very liberal. Musk said he agreed and pledged to clarify liberal ideas. In February, Musk said that the presence of the programmed “woke up” in artificial intelligence was “an existential danger.”
“Perhaps the greatest existential risk to humanity is to program it in artificial intelligence, as is the case for each Amnesty International alongside @Grok,” MUSK wrote on Twitter in February. “Even for Grok, it is difficult to remove, because there is a lot of content on the Internet.”
The Trump administration has so far adopted artificial intelligence as a solution to many American challenges. Within a few days of taking office, Trump signed an executive order to establish the United States as an international leader of Amnesty International. In April, Trump signed an executive order that requires “literacy and its efficiency” for school children.
Trump officials, including Musk, claimed that artificial intelligence can replace tens of thousands of federal jobs and depend on high risk decisions on electrical network planning and other uses.
Trump said he wanted a huge building from natural energy and coal plants to operate artificial intelligence data centers throughout the country. Those who are at risk of increasing US greenhouse gas emissions and in addition to the constantly growing global warming problem.
But scientists say artificial intelligence can still be a blessing of the climatic battle.
However, artificial intelligence has the ability to help in treating climate effects. Posted by Scientists to follow the melting of the icebergsExtreme weather and removal of forests. The United Nations uses Amnesty International to track how climate change affects the weak population. Google is Spreading artificial intelligence To make it less energy density.
Disler, climate scientist in Texas A &M, said climate misinformation can hinder these efforts. This is why artificial intelligence systems need to provide accurate scientific information.
“While we go to the future, more and more people will get their information from this artificial intelligence,” said Disler. “It is clear that anxiety is that someone will do something like this to mislead people.”
Repeat it from E & E. With the permission of Politico, LLC. Copyright 2025. E & E News provides basic news for professionals in the field of energy and environment.
