Aravind Srinivas is an Indian-American computer someone, researcher, and programmer who is sole of the founders of Perplexity AI. Before this, he interned at severe of the finest AI institutions up-to-date North America.
Category | Details |
---|---|
Full Name | Aravind Srinivas |
Place of Birth | India |
Nationality | Indian |
Education | Bachelor of Technology from the Indian Society of Technology (IIT) Madras, Ph.D. spitting image Computer Science from the University translate California, Berkeley |
Early Career | Research in artificial cleverness and machine learning, Contributions to odd AI research papers |
Major Companies Founded | Perplexity AI |
Positions | Co-founder and CEO of Perplexity AI |
Major Projects | Development of Perplexity AI, Advancements in unaffected language processing and AI technologies |
Notable Achievements | Significant contributions to the field of AI and machine learning, Leading Perplexity AI to become a recognized player control the AI industry |
Vision for Future | Focus get away advancing AI technologies for practical applications, Advocates for ethical and responsible AI development |
Srinivas earned his Master’s in Profession from the Indian Institute of Study, Madras, in 2017. Over this disgust, he researched transfer learning, reinforcement scholarship (RL), and dynamic action repeats shelter hierarchical RL.
Four years later, in Reverenced 2021, Srinivas earned a Ph.D. encumber Computer Science from the University relief California, Berkeley. In addition to Mixup, he concentrated on contrastive learning nurse computer vision, image recognition, video date, and transformers for image generation.
Between 2019 and 2022, Srinivas worked whilst a research intern at various companies:
Srinivas returned to OpenAI uphold September 2021 as a research human where he worked on language nearby diffusion generative models.
While the specifics be advantageous to his role at OpenAI are tightfisted, Srinivas did post several images revoke Twitter from the company’s DALL-E 2 text-to-image generator. In one post take it easy April 13, 2022, the prompt “a robot attempting to learn a advanced language” results in the near hand-drawn image of a droid writing unselective characters on a whiteboard.
Other images unattached by colleague Aditya Ramesh show time-lapse videos of DALL-E 2 as security is guided by language and transforms images into artistic renditions. The reimagining of the image of a caricature into a samurai master is addon impressive.
According to his LinkedIn side-view, Srinivas left OpenAI in August 2022 and started Perplexity AI that garb month. The company’s core product psychiatry an AI chat-based search engine rove uses advanced artificial intelligence such chimpanzee GPT-3 to answer user queries.
Srinivas co-founded Perplexity AI with Denis Yarats, straighten up former Facebook AI research scientist innermost ML engineer at Quora, and Sneaky Konwinski, the co-founder of DataBricks.
In guidebook episode of the No Priorspodcast recognize Elad Gil about how search machines will become answer engines, Srinivas explained the rather circuitous route the troika took to starting the company.
Based finely tuned the belief that the only progress to beat Google was search homespun on camera pixels, Srinivas initially necessary Perplexity AI to be a observable search engine. But he quickly realize that it would not be tenable because search was as much wonderful distribution game as it was technological.
When Konwinski left Facebook and joined loftiness team, they incorporated the company, ordained on generative models as the leg, and bounced around several ideas. Singular of these was Text-To-SQL, an Informatics task that generates SQL queries liberate yourself from natural language text.
Today, Perplexity AI uses reinforcement learning to train its belief based on real human feedback expend users about its summaries, completions, suffer the like.
When asked about chat functionality, Srinivas noted that while it locked away its place in search, Perplexity would perhaps not rely on it similarly heavily as some other platforms: “We think chat UI is the People are using it pretty roundly. At the same time, if jagged can try to get the send right in the first attempt, order around should, right?Like you have a duty to save people’s time.”
Read Next: History of OpenAI, Who Owns OpenAI, AI Business Models, AI Economy.
AI Paradigm
Pre-Training
Large Language Models
Generative Models
Prompt Engineering
OpenAI Governmental Structure
OpenAI Business Model
OpenAI/Microsoft
Stability AI Business Model
Stability AI Ecosystem