This Stanford computer science professor went to written exams 2 years ago because of AI. He says his students insisted on it

But two years ago, as the latest wave of artificial intelligence began to reshape education, Leskovik said luck He was shaking by the explosion of his field in the main current. He said that Stanford has the prominent computer science program he feels as if he “sees the future because it is born, or even before the birth of the future”, but the general version of GPT-3 was great.

“We had a major existential crisis among students a few years ago when it was not clear what our role in this world,” Leskovic said.

He said that it seems like breakthroughs in artificial intelligence will be to the point that “it will do just a research for us, so what do we do?” He said that he spent a lot of time talking to students at the level of Ph.D. on how to organize themselves, even about what their role in the world will advance. It was “existential” and “surprising”. After that, he received another surprise: a student led by the student to make a test change.

He said, “He left the group,” especially teaching assistants, the former generation of university students in computer science. Their idea was simple: “We do a paper examination.”

Amnesty International as a catalyst for change

Leskovec, a prominent researcher at Stanford whose experience is in the data regulating graphic and artificial intelligence applications in biology, is the axis of a mixture of surprise and thinking. Historically, its chapters relied on the open book and exams wandering at home, where students can benefit from textbooks and the Internet. They could not use the icon and solutions of others, but the rest was a fair game. When large language models such as Openai from GPT-3 and GPT-4 exploded on the scene, students and teaching assistants alike started to ask whether the assessments should be dealt with differently.

He now said much more than work for him and to get TAS, saying that these exams take “longer time” to the class. But they all agreed that it was the best way to test student knowledge. The era of artificial intelligence of Leskovec, who is a veterans of Amnesty International, surprised him by placing a higher work burden on himself and other human beings. Besides the presence of “less trees in the world” of all the cooking paper, he said that artificial intelligence has just created an “additional work”. Its 400 chapters feel as an audience at the “Rock concert”, but he insisted that he does not resort to artificial intelligence to help collect and analyze all exams.

“No, no, no, we are,” insist.

A solution moved by the student

Leskovec solution is directly in the midst of a pseudonym on how to change artificial intelligence in higher education, as outbreak reports have prompted many colleges to prohibit the use of artificial intelligence directly. Other professors return to the paper exam, as they revive the famous blue books of many memories of children in the nineties of high schools. Even one of New York University professor suggested obtaining the “Middle Ages”, which adopts old forms of test such as oral examination and book. In the case of Leskovec, the solution of the artificial intelligence professor of the era of artificial intelligence is the same as a distance from the artificial intelligence of the test.

When asked if he was concerned about students deceiving with artificial intelligence, Leskovec asked another question: “Are you worried about deceiving students with calculators? He said that Amnesty International is a surprisingly powerful tool “somewhat from just our appearance and surprised us all”, but it is also “very incomplete … We need to learn how to use this tool, and we need to be able to test people who can use the tool and the ability to think about themselves.”

What is the skill of artificial intelligence and what is the human skill?

Leskovec wrestles with a question that affects everyone in the workforce: What is the human skill, what is the skill of artificial intelligence, and where do you merge? Professor of the Massachusetts Institute of Technology, David Autor and Google SVP James Manica Atlantic Ocean Tools like a calculator or artificial intelligence are generally located in buckets: automation and cooperation. Think of the dishwasher, on the one hand, or the text processor, on the other hand. The tool of cooperation “requires human participation” and the issue with artificial intelligence is that it “does not go accurately in either of them [bucket]”

The job market sends a message about the application of artificial intelligence, which equals something like a response from magic 8: “Fogly response. Try again later.” The federal job report has revealed anemia growth since spring, the latest of which is disappointing expectations with only 22,000 jobs in August. Most economists attribute the lack of employment to uncertainty over president Donald Trump’s regime, which has ruled multiple illegal courts and appears to be heading to the Supreme Court. But the implementation of artificial intelligence is not running smoothly at the company level, with the failure of the Massachusetts Institute of Technology (not related to Autor) to find 95 % of the immunic artificial intelligence pilots, followed by a short period of the Stanford study that finds the beginning of the collapse of employment at the level of entry, especially in the jobs that were subjected to automatic operation by artificial intelligence.

For another perspective, the free Marketplace market has just launched the opening monthly employment report, revealing jobs that are not full of jobs that are rewarded by the market. The answer is “artificial intelligence skills” in the demanding demand, and even if companies are not employing full -time employees, they accumulate to independent workers with high wages and very skill.

Although the labor market is generally softer, UpWork finds that companies “benefit strategically from flexible talents to treat temporary gaps in the workforce”, as large companies lead a 31 % growth in what requires higher work (contracts that exceed $ 1,000) on the platform. Smaller and medium -sized companies accumulate to “artificial intelligence skills”, with the demand for artificial intelligence and 40 % machine learning. But UpWork also sees an increasing demand for the type of skills between: a person who is useful in cooperation with artificial intelligence.

Upwork says artificial intelligence “amplifying human talents” by creating demand for expertise in higher value, and more clearly through creativity, design, writing and translation categories. One of the best skills appointed in August is to verify facts, given “the need for human verification of artificial intelligence outputs.”

“Humans return directly to the episode” to work with artificial intelligence.

“We actually see the human skills that reach a premium,” she said, adding that she believes that people are aware that Amnesty International is very touched by time to completely replace human participation. “I think what people see, and now that they use the content created from artificial intelligence, is that they need to verify the facts.”

The extension of this line of thinking, Monahan said that the sophisticated scene of “artificial intelligence skills” shows what you call “field experience” growing increasingly. She said that the legal category has grown in August, as it highlights the legal expertise required for the legal writing that was created by AI. If you do not have advanced skills in a specific field, it is “easy to be deceived” through the content created from artificial intelligence, and companies are rented for protection from that.

Leskovec agreed when asked about the skills gap that appears to be facing workers for beginners who are trying to obtain employment, on the one hand, and companies that are struggling to implement artificial intelligence effectively.

“I think we almost need to restore the skill of the workforce. Human experience is much more important than what it was ever [before]He added that the issue of beginners is the “essence of the problem”, because it is assumed that young workers will get the experience of the field to cooperate effectively with artificial intelligence?

“I think it is due to teaching, recruitment, and rethinking our curriculum,” said Leskovic. He asked a rhetorical question: How can they have skilled senior workers if they do not take young workers and take time to train them?

When asked before luck In order to wipe the scene and evaluate where we are now using artificial intelligence, as students, professors and workers, Leskovic said, “Very early in this.” He said he believed that we are “next with solutions.” Solutions such as a manually exam and a professor finds news ways to verify his students’ knowledge.