Talking to Zerodha cofounder and investor Nikhil Kamath on his podcast, Srinivas mentioned, “All of them are doing related issues. There may be not a real differentiation between ChatGPT, Anthropic, Grok or Gemini proper now. For Perplexity, our differentiation was that we had been the one ones to ensure from the start that we had extremely correct sources for every part, which everyone seems to be now realising.”
At present, excluding the search half, most fashions give the identical solutions, he mentioned. “It is because they’re making an attempt to climb on the identical benchmarks, similar leaderboards. There is not a whole lot of qualitative distinction,” he defined.
The place is the nuance then for a buyer to make use of a selected mannequin? Srinivas mentioned it is determined by the aim. “For those who use AI for reality checks and analysis, monetary analysis and so forth, you’ll choose whichever mannequin does that finest.”
The Perplexity CEO added that within the coming 12 months, differentiation will come from the agentic behaviour of AI. “Answering questions will probably be seen as a commodity—simply answering in textual content won’t minimize it. Parts like graphics and charts will probably be essential,” he mentioned, including that the subsequent magic will probably be about AI “doing issues”.
Requested why this was not working in AI techniques at this time, Srinivas mentioned agentic behaviour can’t function with out reasoning, and has simply begun to begin working efficiently.
Uncover the tales of your curiosity
Discover more from News Journals
Subscribe to get the latest posts sent to your email.
