Looking ahead, @WadhwaniAI envisions a shift to language-oriented interfaces and scaling solutions. Read our CEO, Shekar Sivasubramanian's full interview to know how AI can catalyze solutions for India's persistent developmental challenges.
RESEARCH
-
OpenAI Adds Seed Parameter for Deterministic Output in API
By
–
Since this tweet, OAI added a seed parameter that you can now get deterministic output — or, at least, a "best effort" at deterministic output with a clear error when not possible. See here: https://
cookbook.openai.com/examples/deter
ministic_outputs_with_the_seed_parameter
… -
AI Automates Surgical Feedback Analysis for Training
By
–
Our paper "Deep Multimodal Fusion for Surgical Feedback Classification" won the best proceedings paper at @SymposiumML4H
! http://
tinyurl.com/deep-surgical-
feedback
…
We're revolutionizing #SurgicalTraining by automating feedback analysis using #MachineLearning. We've developed the first #AI -
Description Is Not a Plan: Common Academic Mistake
By
–
Description is not a plan. A common mistake that academics make.
-
Gartner’s Hype Cycle: 2000+ Technologies and GenAI Insights
By
–
This hype cycle distills insights from more than 2,000 technologies and applied frameworks that @Gartner_inc https://
gtnr.it/3QLYUam #EmergingTech #Technology #GenAI #AI #ML #ChatGPT #chatbots #chatgpt4 #chatgpt3 #GenerativeAI @Khulood_Almani @ipfconline1 @PawlowskiMario -
XGBoost Could Have Improved Results Six Years Ago
By
–
They could have been way better six years ago if they had used XGBoost.
-
2023 AI Events Recap: Year’s Biggest Developments
By
–
2023 has been a heck of a year! I made a full recap of all of the biggest events (IMO) that happened throughout the year. Of all the videos I've made so far, this one probably took me the longest to research and produce… Hope you enjoy it. 🙂 https://
youtu.be/I_m54jvnmgE -
Medical Training Cannot Rely on Text Alone
By
–
If the practice of medicine could be learned purely from text, training physicians would not require years of residency.
-
AI Safety: A Technical Responsibility for Specialists
By
–
1. I'm not surprised, but a lot of people reading this might be.
2. I do care about AI safety just as much as turbojet engineers care about the reliability of their engines (which is a lot). But both are deeply technical questions for specialists to solve through careful -
Key figures from top AI labs and companies
By
–
People from top AI labs and companies (OpenAI, Deep Mind, FAIR, NVIDIA, Mistral, etc.) would be a good starting point. Too many to list right now.