Skip to content
Nikhil KamathNikhil Kamath

Nikhil Kamath ft. Perplexity CEO, Aravind Srinivas | WTF Online Ep 1.

In this episode, we sat with Perplexity AI co-founder & CEO, Aravind Srinivas, to explore the evolution of artificial intelligence, what the big AI giants are up to & if we can even predict the future. We also speak about the biggest AI advancements, the role of India in this fast-moving sector - where the real opportunities lie, what’s being overlooked, and finally, the questions we didn't even know we should be asking. #NikhilKamath Co-founder of Zerodha, True Beacon and Gruhas Twitter: https://x.com/nikhilkamathcio LinkedIn: https://www.linkedin.com/in/nikhilkamathcio Instagram:https://www.instagram.com/nikhilkamathcio Facebook: https://www.facebook.com/nikhilkamathcio #AravindSrinivas Co-founder and CEO of Perplexity AI Twitter: https://x.com/AravSrinivas LinkedIn: https://www.linkedin.com/in/aravind-srinivas-16051987 Instagram: https://www.instagram.com/aravindsrinivas Timestamps - 00:00 - Intro 00:45 - Meeting Aravind Srinivas | His Journey & Career Path 12:14 - AI’s Evolution | From Basics to Super Intelligence 29:06 - Understanding the Process Behind AI 35:54 - WTF is a Neural Network? 45:25 - Large Language Models (LLMs) & it’s Evolution 53:59 - The Latest AI Shifts 57:03 - Aravind’s Hustle | Work, Education & Family 01:05:13 - What are Big Players of AI Doing? | Perplexity, Google, Meta, Open AI, Anthropic, and more 01:22:00 - Where the Real AI Opportunities Are 01:34:44 - AI Features & Tools | Text-Videos, Chatbots, Translations 01:39:15 - Why Data Centers Are the Next Big Thing 01:49:42 - India’s Role & Scope in this Industry 02:02:47 - Aravind’s AI Platform Recommendations 02:05:26 - Where AI is Headed Next 02:08:43 - AI Regulations | Tackling Complications 02:16:17 - Outro #WTFiswithnikhilkamath #WTFOnline #nikhilkamath #perplexityai #ai #google #meta #neuralnetworks #perplexity #chatgpt #openai #gemini #manus #deepseek #technology #tech #data #datacenter

Nikhil KamathhostAravind Srinivasguest
Mar 22, 20252h 16mWatch on YouTube ↗

At a glance

WHAT IT’S REALLY ABOUT

Perplexity CEO explains AI basics, industry shifts, and India’s opportunities

  1. Aravind Srinivas recounts his path from Chennai and IIT to Berkeley, OpenAI, and founding Perplexity, emphasizing learning through humility, fundamentals, and sustained effort.
  2. He explains AI from first principles—narrow vs general intelligence, neural networks, machine learning, and how large language models are trained via next-token prediction plus post-training (e.g., RLHF).
  3. The discussion argues the 2020s AI leap came from scaling compute with higher-quality data and training methods, and that differentiation is shifting from “chat” to agentic systems that take actions and complete transactions.
  4. They examine competitive dynamics (Google’s distribution moats, Meta’s network effects), data centers and chips (NVIDIA’s CUDA/software moat), India’s role (model-building and voice), and a light-touch approach to regulation focused on applications rather than models.

IDEAS WORTH REMEMBERING

5 ideas

Modern AI progress was driven by scaling simple ideas with compute and data.

Srinivas describes a key lesson from OpenAI/Ilya Sutskever: sophisticated academic ideas often lose to simpler approaches once you “throw a lot of compute” at them—provided data quality is high and training is done correctly.

General-purpose capability—not single-task performance—is what feels disruptive now.

Earlier “AI” like chess engines or calculators excelled at narrow tasks; today’s LLMs are one system that can handle thousands of economically valuable tasks (coding, writing, summarizing), creating broad labor and business impact.

An LLM is a giant neural network trained mostly to predict the next word.

Pretraining consumes massive text corpora (internet-scale tokens) using transformers; post-training then reshapes the model into a useful chatbot via fine-tuning and learning from human feedback (RLHF).

Neural networks learn patterns only when the task and data contain real signal.

Using the stock-market example, he notes models can’t reliably extract predictive power from irreducible noise; performance depends on whether the dataset and objective expose true structure that generalizes.

Chatbots are converging; the next differentiation is “agentic” action and workflow.

He predicts question-answering becomes a commodity, while winners will integrate personal context (email/calendar), tools/APIs, voice UX, and execution (booking, purchasing, emailing) with reliable reasoning.

WORDS WORTH SAVING

5 quotes

AI is just two circles... The big circle is generative AI, and the smaller circle is reinforcement learning... and the only thing that remains is to throw a lot of compute at it.

Aravind Srinivas

Even though other people in academia... respect you for the more complicated ideas, what matters in reality is making things work, and it's often the simplest ideas... thrown a lot of compute at them.

Aravind Srinivas

A large language model... is essentially a giant neural network that's trained on... predicting the next word... training on the whole internet.

Aravind Srinivas

I feel like the real magic is gonna come from AIs doing things.

Aravind Srinivas

Regulating models is not necessarily a great idea... The best way is to regulate applications.

Aravind Srinivas

Aravind’s journey: IIT → ML → Berkeley → OpenAI → PerplexityDefining AI, AGI, superintelligence; narrow vs general systemsNeural networks, loss functions, backprop; ML vs neural netsLLMs: pretraining, transformers, tokens; post-training and RLHFWhy AI took off: compute scale + high-quality data + product interfaceAI product differentiation: search, sources, latency, multimodal UXAgents and integrations: action-taking assistants, transactions, voiceCompetitive moats: Google default search, Android/Play Store, Meta networksData centers and vertical integration; inference vs training economicsNVIDIA and CUDA; Google’s full-stack TPU/JAX/XLA approachIndia opportunities: domestic models, Indian voice and dialectsRegulation: focus on risky applications (kids/companionship) vs model bans

High quality AI-generated summary created from speaker-labeled transcript.

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.

Add to Chrome