Who is Aravind Srinivas?

Aravind Srinivas is an Indian-American computer scientist, researcher, and programmer who is one of the founders of Perplexity AI. Before this, he interned at some of the finest AI institutions in North America.

Education 

Srinivas earned his Master’s in Engineering from the Indian Institute of Technology, Madras, in 2017. Over this time, he researched transfer learning, reinforcement learning (RL), and dynamic action repeats for hierarchical RL.

Four years later, in August 2021, Srinivas earned a Ph.D. in Computer Science from the University of California, Berkeley. In addition to RL, he concentrated on contrastive learning for computer vision, image recognition, video generation, and transformers for image generation. 

Research internships

Between 2019 and 2022, Srinivas worked as a research intern at various companies:

  • OpenAI – in mid-2018 for a period of four months, Srinivas worked on solving RL problems with policy gradient algorithms. 
  • DeepMind – Srinivas then traveled to London in mid-2019 where he interned at DeepMind for around five months. There, he worked on large-scale contrastive learning.
  • Google – he then returned to the United States and worked at Google for a twelve-month stint between May 2020 and April 2021. He was involved with the self-attention-based model HaloNet and SOTA vision models such as ResNet-RS. 

OpenAI

Srinivas returned to OpenAI in September 2021 as a research scientist where he worked on language and diffusion generative models.

While the specifics of his role at OpenAI are unclear, Srinivas did post several images to Twitter from the company’s DALL-E 2 text-to-image generator. In one post on April 13, 2022, the prompt “a robot attempting to learn a new language” results in the near hand-drawn image of a droid writing random characters on a whiteboard. 

Other images released by colleague Aditya Ramesh show time-lapse videos of DALL-E 2 as it is guided by language and transforms images into artistic renditions. The reimagining of the image of a cat into a samurai master is particularly impressive.

Perplexity AI

According to his LinkedIn profile, Srinivas left OpenAI in August 2022 and started Perplexity AI that same month. The company’s core product is an AI chat-based search engine that uses advanced artificial intelligence such as GPT-3 to answer user queries.

Srinivas co-founded Perplexity AI with Denis Yarats, a former Facebook AI research scientist and ML engineer at Quora, and Andy Konwinski, the co-founder of DataBricks.

In an episode of the No Priors podcast with Elad Gil about how search engines will become answer engines, Srinivas explained the rather circuitous route the three took to starting the company. 

Based on the belief that the only way to beat Google was search based on camera pixels, Srinivas initially wanted Perplexity AI to be a visual search engine. But he quickly realized that it would not be possible because search was as much a distribution game as it was technological.

When Konwinski left Facebook and joined the team, they incorporated the company, settled on generative models as the foundation, and bounced around several ideas. One of these was Text-To-SQL, an NLP task that generates SQL queries from natural language text.

Today, Perplexity AI uses reinforcement learning to train its model based on real human feedback from users about its summaries, completions, and the like. 

When asked about chat functionality, Srinivas noted that while it had its place in search, Perplexity would perhaps not rely on it as heavily as some other platforms: “We think chat UI is the future. People are using it pretty heavily. At the same time, if you can try to get the answer right in the first attempt, you should, right? Like you have a responsibility to save people’s time.

Key takeaways:

  • Aravind Srinivas is an Indian-American computer scientist, researcher, and programmer who is one of the founders of Perplexity AI. Before this, he interned at some of the finest AI institutions in North America such as OpenAI and DeepMind.
  • Srinivas returned to OpenAI in September 2021 as a research scientist where he worked on language and diffusion generative models. One of his projects involved the text-to-image generator DALL-E 2. 
  • According to his LinkedIn profile, Srinivas left OpenAI in August 2022 and started Perplexity AI soon after. The company’s core product is an AI chat tool-based search engine that uses advanced artificial intelligence such as GPT-3 to answer user queries.

Timeline

  • Aravind Srinivas’s Background: Aravind Srinivas is an Indian-American computer scientist, researcher, and programmer. He is one of the founders of Perplexity AI, an AI chat-based search engine.
  • Education and Early Career: Srinivas earned his Master’s in Engineering from the Indian Institute of Technology, Madras, where he researched topics related to transfer learning and reinforcement learning. He later obtained his Ph.D. in Computer Science from the University of California, Berkeley, with a focus on contrastive learning for computer vision and transformers for image generation.
  • Research Internships: Srinivas worked as a research intern at prominent AI institutions, including OpenAI, DeepMind, and Google. His internships covered topics like policy gradient algorithms for reinforcement learning, large-scale contrastive learning, and self-attention-based models.
  • Work at OpenAI: Srinivas returned to OpenAI as a research scientist in 2021. He worked on language and diffusion generative models and contributed to projects like DALL-E 2, a text-to-image generator.
  • Co-Founding Perplexity AI: After leaving OpenAI in August 2022, Srinivas co-founded Perplexity AI with Denis Yarats and Andy Konwinski. Perplexity AI developed an AI chat-based search engine using advanced artificial intelligence, particularly GPT-3, to provide answers to user queries.
  • Perplexity AI’s Core Product: Perplexity AI’s chat tool-based search engine employs reinforcement learning to improve its model based on user feedback. The company aims to provide accurate answers in the first attempt to save users’ time.
  • Vision for the Future: While initially considering a visual search engine, Perplexity AI settled on generative models as the foundation for its search engine. They believe chat UI is the future, but the emphasis is on delivering accurate answers efficiently to users.

Read Next: History of OpenAI, Who Owns OpenAI, AI Business Models, AI Economy.

Connected Business Model Analyses

AI Paradigm

current-AI-paradigm

Pre-Training

pre-training

Large Language Models

large-language-models-llms
Large language models (LLMs) are AI tools that can read, summarize, and translate text. This enables them to predict words and craft sentences that reflect how humans write and speak.

Generative Models

generative-models

Prompt Engineering

prompt-engineering
Prompt engineering is a natural language processing (NLP) concept that involves discovering inputs that yield desirable or useful results. Like most processes, the quality of the inputs determines the quality of the outputs in prompt engineering. Designing effective prompts increases the likelihood that the model will return a response that is both favorable and contextual. Developed by OpenAI, the CLIP (Contrastive Language-Image Pre-training) model is an example of a model that utilizes prompts to classify images and captions from over 400 million image-caption pairs.

OpenAI Organizational Structure

openai-organizational-structure
OpenAI is an artificial intelligence research laboratory that transitioned into a for-profit organization in 2019. The corporate structure is organized around two entities: OpenAI, Inc., which is a single-member Delaware LLC controlled by OpenAI non-profit, And OpenAI LP, which is a capped, for-profit organization. The OpenAI LP is governed by the board of OpenAI, Inc (the foundation), which acts as a General Partner. At the same time, Limited Partners comprise employees of the LP, some of the board members, and other investors like Reid Hoffman’s charitable foundation, Khosla Ventures, and Microsoft, the leading investor in the LP.

OpenAI Business Model

how-does-openai-make-money
OpenAI has built the foundational layer of the AI industry. With large generative models like GPT-3 and DALL-E, OpenAI offers API access to businesses that want to develop applications on top of its foundational models while being able to plug these models into their products and customize these models with proprietary data and additional AI features. On the other hand, OpenAI also released ChatGPT, developing around a freemium model. Microsoft also commercializes opener products through its commercial partnership.

OpenAI/Microsoft

openai-microsoft
OpenAI and Microsoft partnered up from a commercial standpoint. The history of the partnership started in 2016 and consolidated in 2019, with Microsoft investing a billion dollars into the partnership. It’s now taking a leap forward, with Microsoft in talks to put $10 billion into this partnership. Microsoft, through OpenAI, is developing its Azure AI Supercomputer while enhancing its Azure Enterprise Platform and integrating OpenAI’s models into its business and consumer products (GitHub, Office, Bing).

Stability AI Business Model

how-does-stability-ai-make-money
Stability AI is the entity behind Stable Diffusion. Stability makes money from our AI products and from providing AI consulting services to businesses. Stability AI monetizes Stable Diffusion via DreamStudio’s APIs. While it also releases it open-source for anyone to download and use. Stability AI also makes money via enterprise services, where its core development team offers the chance to enterprise customers to service, scale, and customize Stable Diffusion or other large generative models to their needs.

Stability AI Ecosystem

stability-ai-ecosystem

About The Author

Scroll to Top
FourWeekMBA