Within a span of 12 months, ChatGPT, the flagship query robot of OpenAI which took the world by storm with its launch in 2022 and continues to amaze us, will be "laughably bad," company COO Brad Lightcap has said.
Lightcap boldly declared that if individuals find ChatGPT impressive now, they are in for an even greater marvel in the near future.
According to Lightcap, who was addressing the 27th annual Milken Institute Global Conference, the current generation of generative AI tools, including ChatGPT, will soon pale in comparison to the forthcoming advancements.
"In the next couple of 12 months, I think the systems that we use today will be laughably bad," Lightcap said during a panel discussion, according to Business Insider.
"We think we're going to move toward a world where they're much more capable," he added.
Lightcap said he envisions a future where large language models, currently utilised to aid in professional and personal tasks, will transcend their current capabilities to tackle "more complex work."
He said that AI will evolve to establish a deeper "system relationship" with users and that it will serve as a "great teammate" that can assist users on "any given problem."
"That's going to be a different way of using software," he said.
The executive said he recognises the challenge individuals face in truly grasping and embracing the concept of a world populated with robotic assistants, but there will be a shift over the next decade.
According to the COO, engaging in conversations with AI in a manner akin to conversing with a friend, teammate, or project collaborator will become the prevailing norm.
"I think that's a profound shift that we haven't quite grasped," he said, referring to his 10-year forecast.
His remarks hint at what's ahead with GPT-5, OpenAI's upcoming model slated for release as early as this summer.
According to a CEO who received a demo of the new model from OpenAI, GPT-5 is significantly better than its predecessor, GPT-4. He called it "really good" and "materially better" than GPT-4.
Although the CEO didn't provide specifics, they mentioned OpenAI suggesting other unreleased features, such as the ability for AI agents to autonomously perform tasks.
Despite advancements, there's still progress needed before AI can be fully trusted as a personal assistant due to issues biases in current systems. However, Lightcap suggests that at the very least, more advanced AI could serve as a "stand-up comedian," a prospect he deems plausible.
"We're just scratching the surface on the full kind of set of capabilities that these systems have," Lightcap said. He added, "That's going to surprise us."
Image Source: Unsplash
Comments