OpenAI CEO Sam Altman praised Aravind Srinivas, CEO of Perplexity AI, for developing a competing Deep Research project inspired by China's DeepSeek AI. This came shortly after OpenAI introduced Deep Research to ChatGPT, prompting Srinivas to launch a similar, albeit less accurate, but faster and more affordable version.
The exchange began with Altman announcing a new update to ChatGPT, powered by the GPT-4o model, which he described as "pretty good" and improving rapidly.
"We put out an update to ChatGPT (4o). It’s pretty good and will soon get even better; the team is working hard," Altman posted on X.
In response to this, Srinivas asked, "Sorry, what’s the update?"
Altman clarified, "Among other things, it’s the best search product on the web! Check it out and let me know your thoughts."
In a playful reply, Srinivas shared, "Lol, I just mogged you yesterday, check this out," linking to the announcement of Perplexity’s Deep Research launch.
Rather than engaging in any petty back-and-forth, Altman referred to a conversation, possibly from the recent Paris AI Summit, where Srinivas had apologized for his "mean tweets" and said he would let the matter go.
"Since you nicely apologized to me in person for all the mean tweets last week, I’m going to let this go. Keep working hard! Proud of you," Altman responded.
Deep Research on Perplexity vs. ChatGPT: A Comparison :
Perplexity's Deep Research tool claims to perform multiple searches, read hundreds of sources, and reason through next steps in a way that mirrors how a human researcher would tackle a topic. Once the research is complete, the AI generates a "clear and comprehensive report" that can be exported as either a PDF or a Perplexity page for sharing with others.
While Perplexity acknowledges that its Deep Research tool falls behind ChatGPT's version in terms of accuracy, it asserts that its AI outperforms competitors like Gemini Thinking, o3-Mini, Grok 2, and Claude 3.5 Sonnet on the same benchmark.
Although Perplexity's Deep Research may not match ChatGPT in precision, Srinivas argues that the tool is "an order of magnitude faster and cheaper" than OpenAI's offering. Notably, ChatGPT's Deep Research allows only 100 queries for Pro users, whereas Perplexity provides 500 queries per day for paid users, with limited queries available for free users.
