Future ChatGPT versions could replace much of the work people do today, says Ben Goertzel

According to cognitive scientists Ben Goertzel, the world is experiencing a historic moment in artificial intelligence technology.best known for his work as co-developer of sophia the robotAnthropomorphic AI, Goertzel admits there’s a lot of hype around the tech, but there’s a new large scale language model that power Generative AI It has the potential to change the world.

Goertzel believes that generative AI models like those powering ChatGPT could replace many tasks currently performed by human workers.

“At the end of the day, you don’t have to be incredibly creative, innovative, or take big leaps to get most people’s work done,” says Goertzel.

again: ChatGPT looks more like ‘alien intelligence’ than human brain, says futurist

He also discussed how AI tools that automate a significant portion of people’s work could lead to industry restructuring and job reassignment. For example, drive-thru fast food restaurant workers and news copy editors are affected by AI.

“Tools like Grammarly reduce the need for human copy editors,” says Goertzel. “They don’t completely eliminate [the job] But they reduce that need.auto tools [can be used for] Write journalism articles. They have long written summaries of sports results and weather forecasts. “

(Note: There is no substitute for a good editor, but I use Grammarly Pro as my copy checking tool for most of my articles.)

Nonetheless, he identifies two main areas that AI will not replace. It’s a job that relies on human interaction and requires breakthrough creativity.

“One class that never goes out of style is work where human contact is essential.” Preschool teachers, political strategists, artists, and more.

again: AI can automate 25% of all jobs.The highest (and lowest) risks are:

Generative AI chatbots are so powerful that it’s inevitable to wonder if the system is working. Conscious.

Goertzel says it doesn’t matter.

Instead, he hypothesizes that humans come to accept AGI self-awareness based on an intuitive, gut-level understanding, just as we accept the consciousness of other humans.

“I don’t think we need to solve open problems in the philosophy of consciousness to build human-level or superhuman thinking machines,” he said. “But I think you can make these questions feel irrelevant.”

Goertzel also commented on the differences between AI models considered “narrow” and those of AGI, emphasizing that AGI is capable of human-like thinking and creativity. He said developers are closer than ever to realizing his AGI. He predicts that within the next three to ten years he could see a breakthrough in AGI. The current state of AI technology is not AGI to him, but Goertzel is optimistic that progress towards AGI will accelerate through a combination of large-scale language models, machine reasoning and evolutionary learning.

Today’s generative AI “can pretend to be a generic AI just by having a wide variety of training data,” he said. “You don’t have to go far beyond that training data to do amazing things. It’s a testament to the power of computer networks and multi-GPU server farms.”

again: How ChatGPT works

Goertzel highlights philosophical issues surrounding AI and consciousness. He draws parallels with the concept of time, pointing out that while the philosophy of time remains unresolved, humans have learned to exploit relativistic time delays. He suggests that a similar approach could be taken with AI and AGI, allowing significant progress without the need for philosophical answers.

Goertzel’s insight is a glimpse into the future of AI, and his optimism about AGI’s potential to revolutionize the world is a reminder of the impact AI can have on society. Whether you watch with excitement or with caution, advances in AI and AGI will provoke deep thought and debate for years to come.

“Not all philosophical problems are necessarily solved, right? And I think that’s okay,” he said.