(Credit: Bloomberg / Contributor / Bloomberg via Getty Images)
Would you hire an AI for $20,000 or work alongside one?
OpenAI is mulling over charging that much for "PhD-level" AI systems, according to The Information. While it sounds like a lot, keep in mind that human PhD students are only paid between $20,000 - $30,000 per year on average.
"Just hire a PhD researcher for that money [and] give them access to open-access AI and I bet they will outperform OpenAI," says Jay Van Bavel, PhD, a Professor of Psychology at NYU. "Plus you'll have the advantage of a human with real expertise to oversee the work!"
OpenAI would also offer lower-priced AI agents, such as $2,000 per month for a "high-income knowledge worker" agent and a $10,000-per-month software developer agent.
"I'm personally skeptical that they're close to having reliable agents that are truly that level and can 1:1 replace a PhD level researcher," says one Redditor. "Obviously it won't be able to do the actual physical work required to work in most research labs. Even generating good and reliable documents seems unlikely to me." Another adds that while the $2,000 tier could work, the $10,000 and $20,000 tiers may be unrealistic.
OpenAI CEO Sam Altman has been playing with the company's product offerings. This week, he tweeted a proposal to give subscribers "credits" they can cash in for services rather than fixed subscription pricing.
"An idea for paid plans: Your $20 plus subscription converts to credits you can use across features like Deep Research, [GPT-]o1, GPT-4.5, Sora, etc.," Altman says. "No fixed limits per feature and you choose what you want; if you run out of credits you can buy more. What do you think? Good/bad?"
OpenAI debuted its first AI agent, Operator, in January. It can automate simple web-based tasks like ordering groceries and is available with a $200-per-month ChatGPT Pro subscription, the company's most expensive plan. OpenAI burned through an estimated $5 billion last year, The New York Times reports, and is in talks to raise another $40 billion, CNBC reports.
Anthropic launched its own autonomous AI agent that can do web-based tasks for its Claude chatbot in October 2024. This week, CEO Dario Amodei warned the government of the serious economic challenges these AI agents present. He predicts that by 2026 the tech industry will have developed AIs that can mimic what "highly capable" humans can do today, including navigating digital interfaces and "interfacing with the physical world" by controlling lab equipment and manufacturing tools.