“A resource that's packed with tips, news, and discussion around generative AI.” 〰️ wired.com
✨ About Gradient Ascent
Gradient Ascent is the go-to AI newsletter trusted by thousands of researchers, engineers, and managers from top tech firms like Google, Meta, Microsoft, and Amazon, as well as top universities and research labs from over 125 countries worldwide.
As AI weaves itself into the fabric of our daily lives, there is a clear need for resources that demystify complex research into actionable insights and how this work impacts technology and our world.
We also need an easy and enjoyable way to become AI-literate. Online courses are a great starting point. But given how rapidly state-of-the-art research becomes fossilized, we need something that’s always current.
That’s where Gradient Ascent comes in.
🙌 What Readers Get:
🤿 Epochs: Longform illustrated articles that connect cutting-edge research with real-world applications and takeaways that get you a deep understanding of a topic.
☕ S.N.O.Re (Short Notes On Research): Don’t have time to read lengthy research papers? No problem. I write crisp breakdowns of the most interesting ones to save you time and trouble finding them.
🧑💻 The Lab: Hands-on experimental coding projects that teach the inner workings of ML while having lots of fun.
🖌️ Tensor Toons: Having a solid foundation as an ML practitioner is extremely important. I cover the fundamental concepts that underpin current research, usually with hand-drawn illustrations. When AI is explained so simply, even doodles get it.
📚 The Tokenizer: Hundreds of papers are posted to ArXiv daily. I carefully curate and share research papers and articles that will keep you updated with the most relevant pieces of AI research. I also share videos, code repositories, and interesting demos that you can use as inspiration for your next ML project.
📜 Letters from Latent Space: Periodically, I’ll write freeform essays on the intersection of AI, art, creativity, and meaning drawn from personal observations enriched by storytelling.
💸 Discounts: Exclusive discounts on workshops, courses, and coaching sessions.
🤗 The joy of supporting a one-man band on a mission to make the most accessible and enjoyable AI education.
Your support allows me to pay for cloud computing credits, API bills, and services like ChatGPT+ and try new experiments to make the content even more valuable. It also allows this content to be made available to those who can’t afford it.
💵 Expense Gradient Ascent
Many subscribers expense Gradient ascent using their companies’ learning and development budget.
If this newsletter helps you even with just one project, it’s already paid for itself.
Here is an email template you can adapt and send to your manager:
🥳 Invite friends and get Gradient Ascent for free
You can get a free membership by inviting your friends to read this newsletter!
When you use the referral link below or the “Share” button on any post, you'll get credit for any new subscribers. Even the free ones!
Simply send the link in a text or email, or share it on social media with friends.
❤️ Get a Discount
If you are a student or live in a country with a low purchasing power, reach out to me and tell me more about you. I’m happy to give you a full subscription at a steep discount :
🙏 What others are saying
— Jéssica Leão, Partner at Decibel VC
“Sai has a really unique mix of cutting-edge knowledge in AI and incredible/creative writing ability. His newsletter finds a way to make complex topics fun and easy to read about. I highly recommend it for ML practitioners at any level!”
— Cameron R. Wolfe, Ph.D., Senior Scientist at Netflix and author of Deep (Learning) Focus
“I love Gradient Ascent for it's simple explanations and range of topics. Great resource to stay in touch with the crazy field of Machine Learning and AI”
— Devansh, from Artificial Intelligence Made Simple
"Sairam's ability to seamlessly blend captivating stories with technical details makes reading and learning about AI fun in ways few technical writers can."
— Alvin Tsui, Senior Software Developer and author of Below the Surface
"Sairam has an encyclopedic knowledge and enthusiasm for the latest research in Machine Learning and a knack for explaining difficult concepts in a clear and engaging manner."
— Bijan Forutanpour, Computer Graphics R&D Engineer
"As a software engineer, I’ve tried for years to wrap my head around AI/ML. The jargon and equations in those Coursera courses went over my head.But finally, through Sairam’s newsletter, Gradient Ascent, I have been able to learn these concepts with his concise analogies that connect new knowledge with things I already know. There’s nobody I’d trust more than Sairam to take a complicated concept and make it digestible for folks from all walks of life."
— Janahan Sivaraman, Distinguished Engineer at Walmart and author of Often Wrong Club
"I cannot imagine anyone else being able to teach AI as simply and as succinctly as Sairam.Through his newsletter, Twitter threads, and illustrations, I have grasped more AI concepts than from any other source."
— Louie Bacaj, Entrepreneur and Former Engineering Director
— Zander Mackie, Engineer and author of From Very Little Gravitas Indeed
👋 Hi, but who are you?
Hi! I'm Sairam, a Machine Learning Scientist with over 14 years of experience in the tech industry. Currently, I spearhead advanced perception research on autonomous vehicles at Valeo. Before this, I spent almost six years at Intel Labs centered on deep learning algorithms, specifically algorithm-hardware co-design and multimodal models.
After getting my Master's degree from the University of Michigan, I worked for eight years at Qualcomm Technologies Inc. on groundbreaking Computer Vision and Machine Learning projects for multimedia applications. My work on real-time 3D reconstruction was featured in Forbes. I also developed cutting-edge object tracking algorithms and Qualcomm's FastCV library, which earned significant recognition. You may have used this technology on your phone without even realizing it.
Collectively, my work has earned me dozens of patents and publications.
But beyond these, I'm an educator at heart. I’ve served as a Machine Learning Lead and Mentor at the Frontier Development Lab for three years. During this time, I led an interdisciplinary team on projects to understand our Sun, focusing on Starspots, Solar Wind, and reconstructing its surface. My teams achieved significant recognition, with their work mentioned by Google CEO Sundar Pichai.
These days, my time is divided between research, writing, and two tiny humans who’ve got me wrapped firmly around their thumbs.
Outside of Substack, you can find me on Twitter/X and LinkedIn.
Disclaimer: All opinions and views expressed here in the newsletter are mine alone and do not represent those of my employer.
