(Photo by Artur Widak/NurPhoto via Getty Images)
We don't know when OpenAI plans to launch GPT-5, but a company executive is already hyping up future iterations of ChatGPT as a major upgrade that put older versions of the chatbot to shame.
"I think we will look back in a year and realize how laughably bad they were," OpenAI COO Brad Lightcap said in a talk at the Milken Institute Global Conference on Monday.
Lightcap made the comment when asked about OpenAI’s business six to 12 months down the line. In response, he said the AI systems we use today are "laughably bad."
"The kind of turn-based, almost kind of oracle-like system that ChatGPT is today...we don’t see it as a long-term engagement model for what these systems can do,” he said. “We think that we’re going to move to a world where they are...going to be capable of taking on more complex work.”
(Credit: Milken Institute)The result promises to create a more "assistive relationship with users," with AI serving as a "great teammate,” Lightcap added. The change will be so drastic that talking to AI models will be like conversing with a friend or a co-worker.
OpenAI CEO Sam Altman has also been hinting that future upgrades to ChatGPT will be game-changers. “GPT-4 is the dumbest model any of you will ever have to use again by a lot,” he said during a seminar at Stanford University.
Lightcap also downplayed concerns that generative AI will replace human workers and cause mass layoffs. Instead, he expects future AI programs to spark demand for jobs yet to be envisioned. “We have a much more diversified economy, we have a much more resilient economy,” he said, predicting that the workforce will adapt with the technological changes.
“The thing that actually concerns me is that we don’t adopt [AI] fast enough,” he added. “We are an economy where real GDP growth is not as high as it needs to be. There are massive areas in the economy that are ripe for the type of benefits I think this technology could bring.”
One major challenge facing AI is the technology’s demand for electricity. As OpenAI tries to develop even smarter AI systems, it’ll require more cutting-edge, power-hungry GPUs.
"We think this is a big risk,” Lightcap said. “There’s a really big divergence there [and] one of the great challenges society is going to have to grapple with in the next few years.”