(Stock-Sue/Shutter Stock)
In October, Stanford will host a new conference in which AI is placed in the center of science and engineering. Agents 4 are called science, the program requires that each paper be manufactured by AI, first reviewed by AI systems, and even presented using artificial sounds through AI.
It is described as the first conference where artificial intelligence should work as both author and reviewer, it is a format that challenges long -standing principles in the educational publication. The purpose is not only to show what AI can produce in articles but also to review its shortcomings in a transparent order.
The conference is a mental maker of James Zoo, a computer scientist in Stanford, who studies how humans and machines can assist in research. Recent projects have explored whether massive automation can help accelerate discovery, including “virtual lab”, where a “virtual lab”. AI agents suggested and examined possible treatment Emerging for Covade 19 tensions. Agents 4 with science, Zoo is increasing this view. Its purpose is to use AI tools that can help researchers with the idea of maximum process, from the idea to the output.
Zoo’s interest in scientific support began during a PhD in Harvard, when he stepped away from computer science to spend a year in the genomics lab. This experience highlights how difficult it can be for researchers from different fields to communicate. Later, he became convinced that a larger language model (LLM) could improve in eliminating these gaps.
(Shutter stock)
This idea made it a virtual lab. When the system began to produceable results, Zoo went into a road block. Even when the AI system developed key ideas, experienced and written paper drafts, there was no formal way to recognize their contribution. Most of the magazines rejected the idea of naming AI as an author, regardless of its role. Some conferences accepted the AI -supported research but insisted that only people could claim writing.
This resistance to publishers helped form the idea of agents 4 science. Zoo started planning the program earlier this year, talking to researchers in various fields and finding out how to design a conference where AI not only helps science but also gains superiority. The idea quickly draws attention from others who faced similar questions when the machines were doing more work, they encountered similar questions about ways to get more help from AI and share credit.
This interest helped to create the structure of the event itself. Each article presented in Agents 4 Science should be written primarily by the AI system, in which humans are allowed only to participate in auxiliary role. The submissions need to include a clear explanation of how AI worked, what tools it used, and how key decisions were made.
Administrators are putting a wide net. Requests from any sector are welcomed where AI can advance scientific discovery, including biology, chemistry, physics, engineering, and computer science.
Reviewers will also be AI systems, each paper will be independently reviewed by multiple LLMs to reduce prejudice and provide different perspectives. These reviews will follow the standard neurps conference template, and will score papers related to the origin, explanation and importance. After the first round of AI -led AI, a panel of human experts will review advanced submissions. All reviews, AI indicates, and the author’s revelations will be made public, which will present a transparent view to the researchers as to how the machine is led by the machine.
That is, this openness can be just as valuable as these consequences themselves. By making the full pipeline visible, how the AI created its results, as a result of how other models decided its production, the conference gives a rare look at how the machine’s reasoning is in fact. This gives scientists a way of evaluating this process as well as a way of consequences and has to start construction standards for how AI should be used in scientific work.
(Macchi/Shutter Stock)
Some scientists still do not believe in handing over the lab coat to the machines. He pointed out that the AI model, even the best, can spoil the facts, lose context, or come with answers that may be convinced but are closely examined. This is not just about mistakes.
There is a deep trouble that machines, no matter how fast or well, maybe a trained person can argue because of problems. When the stake is high, such as in the research of medicine or climate, such decisions are important.
What does this mean for early career researchers? If AI started to work more than heavy lifting, where do people who are still learning ropes? Some say science is not just about the consequences. This is about the long hours of the process and the long hours of finding things. This struggle helps to make decisions and skills. Without it, researchers may lose such decisions and skills through their predecessors through experience.
Whether AI is ready to take the lead in science, this is an open question. Some people see it as a way of fast and wider discovery, while others are worried about what can be lost on the way. What is clear is that science and engineering research is no longer completely human. When the machines move from the tool to the team’s partner, the challenge will be to make sure they raise the work instead of changing the people behind it. The future of science may depend on how this balance is affected.
Relevant
