Superalignment Fast Grants

OpenAI Blog2023-12-14 16:000 views

We’re launching $10M in grants to support technical research towards the alignment and safety of superhuman AI systems, including weak-to-strong generalization, interpretability, scalable oversight, and more.

Source：https://openai.com/index/superalignment-fast-grants

My favorites

PreviousIncreasing accuracy of pediatric visit notes NextPractices for Governing Agentic AI Systems

After 3 Weeks and More Than 451,000 Folds, the iPhone Duo’s Hinge Is Finally Starting to Fail

We built a machine to fold the iPhone Duo until it breaks. Twenty days of 24/7 folding later, Apple’s phone shows the fi

0CNET News

DeepSeek’s coding agent is too tidy for its own good — and benchmarks are punishing it

DeepSeek’s open-source coding agent is losing points on a major benchmark for a habit developers usually treat as good p

0The New Stack

The context problem: It’s what’s really breaking AI agents at scale

AI models are getting smarter , tokens are getting cheaper, and agents are improving. But the context gap sits between t

0The New Stack

Water-based batteries store more energy in iodine with new electrolyte design

New chemistry developed by a UCL-led team could increase the energy capacity of aqueous (water-based) batteries—a step f

0Phys.org Tech

Russia turns to AI filters to hunt down VPNs in 2.2-billion-ruble crackdown

Russia spent 2.27B to integrate AI into its national web filtering system It flags VPN connections by analyzing packet t

0TechRadar

45+ best early Black Friday deals in the UK — Nintendo Switch 2, LG OLED TVs, AirPods 5, Kindle, Garmin Forerunner, and more of my top picks this weekend

Is it already Black Friday? Technically, it isn't, as the official date is November 27. However, that hasn't stopped ret