Svarat Chadhur was obsessed with puzzles as a child, he says.
Growing up in Kolkut, he spent afternoon, deciding every pyramid and the word he could find in local publications. He then graduated from Bengal translations of Martin Gardner’s American mathematical books on mathematics, logic and puzzles.
Chadhur dreamed of the world in which he could decide the puzzles for life.
That is, more, what he now does, as a researcher and professor of computer science at the University of Texas, Austin.
One of the puzzles he works is especially important. The fact is whether artificial intelligence can actually expand the scale of human knowledge.
Here’s how the 46-year-old Chadhur goes: in his Trishul laboratory at the university, he developed a Copra AI program (short-based content), which works with big language models (in this case GPT-4) to prove mathematical theorems.
It may not seem that fun, but here’s what the line is waiting. As both systems work together, the possible goal of Chadhur is to make Copra new mathematical problems and then work on their solution.
This would be a decisive step towards determining whether II can imitate the research nature of the human mind. This would somehow answer such questions as: Can II as a result of the scientific newspaper co -author?
In other words: Can AI program go beyond what knows not just connect the points (they already do it), but also reach and collect more points to add to the matrix? (The points in which we can not consider at all.)
“This will mean a big leap: from the engines of AI, which we see around us, which are engaged in affordable data and perform several repetitive tasks, to a system that uses much more logic and can perform” inhuman “tasks,” he says.
***
This “inhuman” bit is what it interests him because it will mean new and faster solutions to some of our most pressing problems. Such programs can potentially change as we see our world and move it.
New mathematics problems will only be the beginning. Down, he believes these programs can be staff working with researchers.
“They can be like interesting children to go out to find things on their own and find out what works and what is not,” he says. With such a program, small startups and single masters can take over giant puzzles such as cleaner energy and urban planning.
New answers can arise on questions such as: How can we move a large number of people from point B and back every day? How can we resolve the issue of solar elements that have such a short life? Or, how can we better control the temperature control against the background of the climatic crisis?
***
Chadhuri was a while. In 2017, he and his students created Bayou, an early tool led by AI, which could only create a code based on text tips. This early victory created the basis for its current Copra AI agents.
Currently, Chadhur was awarded the prestigious Guggenheim Scholarship for his work and for the projects he is leading at an “open mathematical discovery”.
He knew early that he was saying that the path to the biggest and heavy puzzles in the world runs through the world of computing.
After school, he studied computer science and entered the Indian Institute of Technology (IIT) -kharagpur. After college, he began to study neural networks-which are the types of programs that seek to imitate the possibilities that connect to the human brain rather than rely on linear (therefore, therefore) that control traditional software.
At this time, 20 years ago, the idea that the computer could once sift the options and choose the right, not spit out the ready answer that was filed with it. Today, of course, all LLM do it. It’s like a chatgpt talking; How Midrus and Shara create their horribly realistic images and videos.
***
Five years (if not before) Chadhur believes that for AI programs will “think” in such a way that more closely imitates the human brain and has inhuman abilities in many areas of human activity.
The transition to renewable resources will be vital to reduce the impact on the environment of server farms and data centers that support such systems, it adds.
What does he think that even this advanced AI version is fighting?
Create deep art, says Chadhur. For the simple reason that art is moving, perhaps more than any other human endeavors, the live experience of the artist.
“It is unlikely that AI will start writing as Rabindranate Tagore, or pulling innovative film scripts, because to create something like literature, constantly moving from the world and the artist’s interaction with the world are vital,” Chadhur says. “This entry level is a long way for AI.”
