OpenAI, developer of ChatGPT, plans to invest significant resources and establish a new research team aiming to keep its artificial intelligence safe for humans, eventually using AI to self-promote itself. It announced Wednesday that it will monitor.
OpenAI, developer of ChatGPT, plans to invest significant resources and establish a new research team aiming to keep its artificial intelligence safe for humans, eventually using AI to self-promote itself. It announced Wednesday that it will monitor.
“The sheer power of superintelligence … could lead to the disempowerment and even the annihilation of humanity,” OpenAI co-founder Ilya Sutskever and coordinator Jan Reich wrote in a blog post. rice field. “At the moment, there are no solutions for piloting or controlling potentially super-intelligent AI and preventing AI misbehavior.”
“The sheer power of superintelligence … could lead to the disempowerment and even the annihilation of humanity,” OpenAI co-founder Ilya Sutskever and coordinator Jan Reich wrote in a blog post. rice field. “At the moment, there are no solutions for piloting or controlling potentially super-intelligent AI and preventing AI misbehavior.”
Super-intelligent AI (systems smarter than humans) could emerge within a decade, the authors of the blog post predicted. According to the authors, humans will need better technology than is currently available to control superintelligent AI, so they focused on ensuring that AI remains beneficial to humans. A breakthrough in so-called “alignment research” is needed.
Super-intelligent AI (systems smarter than humans) could emerge within a decade, the authors of the blog post predicted. According to the authors, humans will need better technology than is currently available to control superintelligent AI, so they focused on ensuring that AI remains beneficial to humans. A breakthrough in so-called “alignment research” is needed.
OpenAI, backed by Microsoft, will dedicate 20% of the computing power it secures over the next four years to solving this problem, they wrote. Additionally, the company is forming a new team to organize around this effort, called the Super Alignment Team.
OpenAI, backed by Microsoft, will dedicate 20% of the computing power it secures over the next four years to solving this problem, they wrote. Additionally, the company is forming a new team to organize around this effort, called the Super Alignment Team.
The team’s goal is to create a “human-level” AI alignment researcher and scale it through vast computing power. According to OpenAI, this involves training an AI system using human feedback, training the AI system to assist in human assessments, and finally training the AI system to actually do alignment surveys. said to mean
The team’s goal is to create a “human-level” AI alignment researcher and scale it through vast computing power. According to OpenAI, this involves training an AI system using human feedback, training the AI system to assist in human assessments, and finally training the AI system to actually do alignment surveys. said to mean
Connor Leahy, an AI safety advocate, warns that early human-level AI could run amok and wreak havoc before it is forced to solve AI safety issues. said to be fundamentally flawed.
Connor Leahy, an AI safety advocate, warns that early human-level AI could run amok and wreak havoc before it is forced to solve AI safety issues. said to be fundamentally flawed.
“We need to work out the coordination before we build human-level intelligence, otherwise we won’t be able to control it by default,” he said in an interview. “Personally, I don’t think this is a particularly good or safe plan.”
“We need to work out the coordination before we build human-level intelligence, otherwise we won’t be able to control it by default,” he said in an interview. “Personally, I don’t think this is a particularly good or safe plan.”
The potential dangers of AI are of paramount concern to both AI researchers and the general public. In April, a group of AI industry leaders and experts signed an open letter calling for a six-month moratorium on OpenAI’s development of systems stronger than GPT-4, citing potential risks to society. A May Reuters/Ipsos poll found that more than two-thirds of Americans are concerned about the potential negative impacts of AI, with 61% believing it could threaten civilization.
