AI experts say we are on the wrong way to get a person similar to a person

According to one panel of hundreds of artificial intelligence, the area is currently conducting artificial intelligence in the wild manner.

This concept was found in the Union for the progress of the 1925 presidential panel on the future of the AI ​​research. This Long Report Examination, the State of the AI ​​infrastructure was placed together by 24 AI researchers from the state of artificial intelligence to the social aspects of artificial intelligence.

The report includes a community opinion section, as well as the community opinion for each section, as well as the section of respondents.

MIT computer scientist Rodney Brooks is a “AI Perception and Reality” section, Gartner hype turnover, a five-stage period for technological hype. In November 2024, Gartner estimated that the hype was only in the summit for the Gartner, “the report said that 79% of respondents did not comply with the EU’s research and development of EU opportunities, the EU preventing the EU research.” AI He was managed by the instructions of the study of the study, “he said.

Artificial General Intelligence (AGI) Hypothetic intelligence of a machine that interprets the information and informs the information as a person. AGI is a sacred gray of the field for automation and efficiency for countless areas and subjects. Review any gift assignments you do not want to spend more time planning a trip to present your taxes. AGI can be placed to reduce the load of Rote tasks, as well as in other areas, can catalyze progress without transporting education and technology.

The surprising majority, 76% of 475 respondents will not be enough to give AGI to expand current approaches to AI.

“In general, the answers show a careful but forwarding approach: AI researchers receive security, ethical governance, benefit, gradual innovation, and gradual innovation and gradual growth,” he said.

Researchers in the optimal way to the EU do not allow the most optimistic way against AGI-technology, despite the hype and current approaches that distort researchers.

“We could have this conversation five years ago – EU, Henry Kautz, a computer scientist in a high percentage of mistakes, product recommendations, and a computer scientist, the General AI began working and began to work in the public.”

The AI ​​function is “far from the solution”, read the report and the best LLMS answered approximately half of a number of questions in a benchmark test only 2024. However, new training methods can improve the strength of these models and new ways to organize AI can further improve their performance.

“I believe that in the next stage of improving reliability, the agents of the agents operated by individual AI agents with cooperation groups, and each other, who are trying to be honest,” Kautz added. “Most common public assesses the quality of the best EU systems today, as well as the AI ​​researcher society; AI’s perceptions remain a year or two behind the technology.”

AI does not go anywhere; After all, Gartner does not end in the hype of the hype “obliviona”, but in exchange for “productivity plateau”. Different arenas of AI use, have different levels, but all the noise from the private sector, Heck, Heck is a refreshing reminder of our own families, AI researchers. AI systems have a place for innovation and development from the road in the way around the world. A mother without AI is the only direction as we are not returning.