Imagine the future and work backwards
RESEARCH
-

AlphaFold breakthrough drives hard science discoveries
By
–
Building on the incredible breakthrough of #AlphaFold. For anyone who is an AI-skeptic, consider how it is driving breakthroughs in the hard sciences. (biology, chemistry, physics.) It’s no fluke.
-

xAI’s Grok model selector update
By
–
xAI is working on a new model selector for a standalone Grok For now, only grok-latest (2) is available, but eventually, we should see there an addition which everyone is waiting for
-
The Dopamine Cycle of Daily AI Research Experiments
By
–
Somewhat meta but there is a dopamine cycle in doing AI research that is pretty interesting Every day you wake up and you think about what experiment to run. You think thing X matters so you decide to improve it or ablate it. Then you write the code and pay some compute to find
-

Sonar Pro Outperforms Search Engines LLMs SimpleQA Benchmark
By
–
Purpose-built for factuality, Sonar Pro outperformed leading search engines and LLMs in terms of answer quality in recent SimpleQA benchmark findings.
-

LLMs Show Bias Toward Socially Valued Personality Traits
By
–
When @Stanford researchers surveyed LLMs on the “big five” personality traits, the models started to bend their answers toward what society values. @JEichstaedt @AadeshSalecha https://
stanford.io/3E1DyTL -
Open Source Importance: Models, Training Processes, and Data
By
–
This is why Open Source is important; not just the model, but also the training process and data.
-

Scaling Laws in AI: A Decade of Research Insights
By
–
"I've been in this field for 10 years. I've worked at all the major companies, including Google and OpenAI. "Me and some of my colleagues were the first ones to document what are called scaling laws. This is the observation that when you pour more computing into AI systems,
-
Apple’s DepthPro Revolutionizes Monocular Depth Estimation
By
–
Apple's DepthPro ⚡️is quite impressive, producing pixel-perfect, high-resolution metric depth maps with sharp boundaries through monocular depth estimation.
— Satya Mallick (@LearnOpenCV) 21 janvier 2025
It outperforms all of its contenders like Metric3D v2 and DepthAnything in "in-the-wild" dynamic scenes.
In this article,… pic.twitter.com/FjE0nHyJrhApple's DepthPro is quite impressive, producing pixel-perfect, high-resolution metric depth maps with sharp boundaries through monocular depth estimation.
It outperforms all of its contenders like Metric3D v2 and DepthAnything in "in-the-wild" dynamic scenes. In this article, -
LLM Architecture Becomes Demystified and More Accessible
By
–
Also I feel like LLM architecture is demystified now and we have enough open datasets too. It’s not as high skill bracket as it was a couple of years ago.