AI & Robotics News and Discussions

[firestar464]

DeepMind develops SAFE, an AI-based app that can fact-check LLMs

https://techxplore.com/news/2024-03-dee ... d-app.html

[firestar464]

[weatheriscool]

Hyundai's DAL-e delivery bot takes on the morning coffee run
By Paul Ridden
April 03, 2024
https://newatlas.com/robotics/hyundai-k ... ery-robot/

Hyundai and Kia are gunning for office mailroom jobs with the introduction of the DAL-e Delivery robot, that can not only distribute packages around a building autonomously, but can also do the coffee run too.

The new delivery bot from the Hyundai/Kia Robotics Lab is not based on the DAL-e "humanoid" launched in 2021, but a model introduced at the end of 2022 that was built on Hyundai's Plug & Drive platform featuring "intelligent steering, braking, in-wheel electric drive and suspension hardware" in one modular unit.

The DAL-e Delivery also features four Plug & Drive modules, with enhanced suspension to ensure that your cup of Joe within its belly arrives stirred but not shaken. The Lab reports that though the new bot doesn't stand as tall as its ancestor, the cargo hold has grown to accommodate up to 16 cups of coffee or packages weighing up to 10 kg (22 lb).

[firestar464]

Computer scientists show the way: AI models need not be so power hungry

https://techxplore.com/news/2024-04-sci ... ungry.html

[weatheriscool]

[Powers]

[weatheriscool]

A framework to improve air-ground robot navigation in complex occlusion-prone environments
https://techxplore.com/news/2024-04-fra ... mplex.html
by Ingrid Fadelli , Tech Xplore

Robotic systems have so far been primarily deployed in warehouses, airports, malls, offices, and other indoor environments, where they assist humans with basic manual tasks or answer simple queries. In the future, however, they could also be deployed in unknown and unmapped environments, where obstacles can easily occlude their sensors, increasing the risk of collisions.

Air-ground robots could be particularly effective for navigating outdoor environments and tackling complex tasks. By moving both on the ground and in the air, these robots could help humans search for survivors after natural disasters, deliver packages to remote locations, monitor natural environments, and complete other missions in complex outdoor settings.

[firestar464]

I feel like OpenAI's never going to declare AGI. Daddy Microsoft won't allow it.

[wjfox]

[weatheriscool]

Tiny AI-trained robots demonstrate remarkable soccer skills

by Bob Yirka , Tech Xplore
https://techxplore.com/news/2024-04-tin ... occer.html

A team of AI specialists at Google's DeepMind has used machine learning to teach tiny robots to play soccer. They describe the process for developing the robots in Science Robotics.

As machine-learning-based LLMs make their way into the public domain, computer engineers continue to look for other applications for AI tools. One purpose that has long captured the imagination of scientists and the public at large is the implementation of robots that can carry out traditionally human tasks that are difficult or arduous.

The basic design for most such robots has typically involved using a direct programming or mimicking approach. In this new effort, the research team in the U.K. has applied machine learning to the process and has created tiny robots (approximately 510 mm tall) that are remarkably good at playing soccer.