Great talk by @emilymbender about what are ok use cases of synthetic text and what aren't. But even for those that are ok use cases, we need to ask the question of whether those use cases are worth the environmental impact. https://
youtube.com/watch?v=qpE40j
wMilU
…
RESEARCH
-
Synthetic text use cases and environmental impact assessment
By
–
-
Path to Human-Level AI: Starting with Fish-Level Intelligence
By
–
Before we reach Human-Level AI, we have to reach Fish-Level AI.
-
LLM Water Consumption Raises Environmental and Ethical Concerns
By
–
What I find remarkable is the continued research showing how much water is used (1/2 a liter for every 15 minute LLM interaction according to a recent paper from Cornell) for GenAI while so much of the world burns and is in drought. We don’t need hallucinations. We need water.
-
Token input limit discussion for model evaluation
By
–
There is a limit on 2048 token inputs for evaluation, is that what you mean?
-
AVIS: Autonomous Visual Information Seeking with Large Language Models
By
–
Today on the blog, read all about AVIS — Autonomous Visual Information Seeking with Large Language Models — a novel method that iteratively employs a planner and reasoner to achieve state-of-the-art results on visual information seeking tasks → https://
goo.gle/3P2y2mY -
Foundation Models Transforming Society Through Advanced AI
By
–
New advancements in AI are transforming many aspects of our society. At the forefront of this development are foundation models. Here, we explain what this technology is and its potential to society:
-
The Dark Forest of R&D and Capital Deployment in AI
By
–
The Dark Forest of R&D and Capital Deployment in AI – Michael Dempsey: Blog https://
bit.ly/47pkGIM #AI #MachineLearning #DeepLearning #LLMs #DataScience -
AI Language Becoming Primary Framework for Understanding World
By
–
Not necessarily saying this is good or bad, just that the language we use tends to reflect what we think is the most state-of-the-art way to describe the world. In my opinion, the language of AI is going to take that mantle from science sooner rather than later.
-
AI Training Paradigm Replaces Scientific Metaphor for Startups
By
–
Science (hypothesis, theory, experiment) has been the dominant metaphor for how to think about startups for the last decade. I think AI (training, prediction, loss) will become the dominant metaphor for how to think about startups over the next decade.
-
Baidu Research Advances Computer Vision Robotics with Seven Papers
By
–
Baidu Research RAL is leading the way in innovation! 🚀 Seven innovative papers in #ComputerVision and #Robotics showcase our advances in stereo matching, autonomous excavators, NeRF and more!
— Baidu Research (@BaiduResearch) 18 août 2023
Explore our work: https://t.co/agJCMDfp2q pic.twitter.com/EhFmABG9JOBaidu Research RAL is leading the way in innovation! Seven innovative papers in #ComputerVision and #Robotics showcase our advances in stereo matching, autonomous excavators, NeRF and more! Explore our work: http://
research.baidu.com/Blog/index-vie
w?id=186
…