Reid Hoffman has said he uses a specific tool daily to gain insight into how AI products could be “workers in the future.”
The LinkedIn cofounder and investor said he did “at least” one prompt daily with OpenAI’s Deep Research tool, an agentic tool for automating complex multi-step internet research. He also said there was many companies building “strong” offerings in the race to make AI agents.
Hoffman, who stood down as an OpenAI director in 2023, citing potential conflicts of interest with his other AI investments, was asked about the startup during an interview on Bloomberg Television on Friday.
He said he was using Deep Research once a day, and that it “gives you the lens to the amplification we’re going to get with these products as workers in the future.”
The rise of agentic AI, which can independently act on a person’s behalf and make decisions without human intervention, has fuelled speculation about how and when AI might replace human workers.
A group of Carnegie Mellon researchers ran a virtual simulation designed to test how AI agents fare in real-world professional scenarios. They found that the top-performing model finished less than one-quarter of all tasks.
“While agents may be used to accelerate some portion of the tasks that human workers are doing, they are likely not a replacement for all tasks at the moment,” Graham Neubig, a computer science professor at CMU and one of the researchers, previously told BI.
Hoffman, who cofounded Manas AI, said he saw no clear leader in the race to develop agentic AI, saying there was “a bunch of folks who are doing very strong things,” and “not just OpenAI, Anthropic, Microsoft, Google.”
Bloomberg interviewer Ed Ludlow told Hoffman he was increasingly talking to AI in voice mode, which he called “a psychological thing that, as a consumer, you kind of have to get over.”
