ChatGPT vs Perplexity: Sam Altman praises Aravind Srinivas’ Deep Research AI; ‘Proud of you’ | Mint -DellyRanks

OpenAI CEO Sam Altman praised Perplexity AI CEO Aravind Srinivas for building a rival Deep Research project based on China’s DeepSeek AI. Notably, OpenAI had recently rolled out Deep Research to ChatGPT which prompted Srinivas to bring a similar offering, albeit less accurate but faster and cheaper. 

The conversation began with Altman announcing a new update to ChatGPT (based on GPT-4o model), which he claimed is ‘pretty good’ and would soon get much better. 

“we put out an update to chatgpt (4o). it is pretty good. it is soon going to get much better, team is cooking.” Altman wrote on X

The vague post got a reply from Srinivas who asked, “sorry what’s the update?”

In reply, the OpenAI CEO explained, “among many other things, it’s the best search product on the web! check it out and lmk what you think.” 

Seeing the response, Srinivas could contain his excitement and wrote, ‘lol, i just mogged you yday, check this out:” while refrecing a post about the launch of Deep Research for Perplexity. 

Instead of being petty about the matter, Altman references a conversation, perhaps from the recently concluded Paris AI Summit, and stated that he would let this go since Srinivas had apologized to him in person for all the ‘mean tweets’. 

“since you nicely apologized to me in person for all the mean tweets last week, im going to let this go 🙂 keep cooking out there! proud of you.” Altman wrote back.

Deep Research on Perplexity vs ChatGPT: 

Perplexity says that its Deep Research tool performs dozens of searches, reads hundreds of sources and reasons what to do next, similar to how a human researcher might approach a topic. Once the source material is complete, the AI agent summarises all the information in a “clear and comprehensive report” that can be exported either as a PDF or as a Perplexity page to share with friends and family.

By its own admission, Perplexity lags behind ChatGPT’s Deep Research in the Humanity’s Last Exam accuracy rankings. However, Perplexity claims that its latest AI offering beats the likes of Gemini Thinking, o3-Mini, Grok 2 and Claude 3.5 Sonnet on the same benchmark.

While Perplexity’s Deep Research may fall short of ChatGPT in terms of accuracy, Srinivas claims in a post that the new AI tool is “an order of magnitude faster and cheaper” than the OpenAI offering. Notably, ChatGPT Deep Research only offers 100 queries to its Pro users, compared to Perplexity’s 500 queries/day for paid users and limited queries for free users.