Ai Japanese start Block Says that AI is created in one of the scientific publications. But while the claim does not need to be real, there is careats to record.
The debate that is swimming around AI and the role in scientific process grows in the day. Many researchers do not think that AI is satisfied to be “scientist together,” while others think it is the beginning.
Sakana fell into the last camp.
The company says it applies ai system called AI Science-V2 to create documents at Saklana, famous and famous. Sakana claims the agenda of the meeting, as well as the leadership of ICLR, settled to work with the company to verify the production application.
Sakana said it cooperated with resembling a British Columbia and University Oxford to send a 3-time meeting for peer reviews. Scientific Science-V2 created the “End-End File File,” including a scientific balance, and analysis, messages and terms.
“We make an opinion on research and scientific research and Techcrunch research and the appropriate submissions.”
One of the three units were recognized at ICLR conference – paper throws an important lenses in the training technique for AI. Sakana said it removed the paper immediately before it was published in the interest of transparency and respect for ICLR convention.
“The acceptable documents offer new ways, how to train neural network training and shows the remaining challenges,” LANGE. “It provides interesting data points to investigate additional science investigation.”
But success is not an impressive as it seems at first.
In the blog article, Sakana admitted that its occasionally led to “the” embarrassment “reference to the Invalidation of 20197 was incorrect.
Sakana’s paper has not been checked as much as some reviewed prints together. Because the company removes it after reviewing the initial peer, the paper did not receive further reviews, during the theory that deny it.
Then there is the fact that acceptance rates for general conventions are increasingly higher than its internal tab for the follow-up bar of ICLR.
Matthew guzdial, AI researchers and assistants at the University of Alberta University, which is called Sakana Results. ”
“The Sakana Folks chosen the document from the manufactured number, means they are using the AI.” Thit
Mike Cook, Research at King London University of Experts in Ai, asked the toughness of the inspector and conference.
He tells the technology “for a new session is reviewed by the more young research.
COOD added that he was not surprised that AI can test the friend, consider Ai excellence in writing human prose. Partially Ai-giveated Document The passenger review is not yet new, cooking points, and the ethical environment is interested in science.
Ai’s technical defect – such as the trend it will make many scientists who are mailing for adoption for serious work. Moreover, experts are afraid AI can Sim In scientific literature, do not raise progress.
“We need to ask yourself [Sakana’s] The Result Is ABOut How Good Ai Is at Designing Ideas – Which It’s Ai ISHOVE AI, “Cook Sleady. “It’s a difference between passing peer reviews and knowledge contribution to the field.”
Sakana, to give credit, no claim that AI it can produce space – or even a particular novel – scientific. Of course, the trial goal is to “study quality of research, and highlight the urgent requirements for the scientist.”
“[T]This is a hard question about that [AI-generated] Scientists should be judged in its own goodness before avoiding prejudices against it. “” Go forward, we will continue to share opinion with research conditions in the review of the review process. “