… they want to put ChatGPT in charge of metal bodies with arms and legs
Project GR00T – apparently named after, though not explicitly linked to, Marvel’s arboriform alien – is a new foundation model from Nvidia developed for controlling humanoid robots. A foundation model, such as GPT-4 for text or StableDiffusion for image generation, is the underlying AI model on which specific use cases can be built. They are the most expensive part of the whole sector to create, but are the engines of all further innovation, since they can be “fine-tuned” to specific use cases down the line.
Nvidia’s foundation model for robots will help them “understand natural language and emulate movements by observing human actions – quickly learning coordination, dexterity, and other skills in order to navigate, adapt, and interact with the real world”.
Yeah, it seems to be that the superchips are more efficient, faster, and smaller, so they’ll be more effective for powering AI software? And those chips will go into the robots and be able to learn more effectively making them eventually autonomous… it’s just around the corner, I’m sure!
But really, this seems more about making data centers more energy efficient (or rather, making it so you can power more off the same energy grid size)…
“ZeroEyes software was installed on 300 of SEPTA’s 31,000 security cameras — slightly under 10% of the system […] The ZeroEyes-connected cameras were spread out across the sprawling transit system, covering most of the Market-Frankford and Broad Street lines’ stations.
It is unclear how many arrests were made because of ZeroEyes. What ZeroEyes’ representatives did see was that the software was no match for SEPTA’s outdated technology.”
We don’t need proper planning, especially for integrating the new stuff into the existing stuff. We also don’t need some sort of pilot project at a small scale, maybe limited to one or two stations to start with, to see what works, what doesn’t and what might have to be modernized in the old stuff or tweaked in the new stuff, and then scale up step by step.
No, we’ll sprinkle magic pixie dust on everything this is AI, and everything will just work perfectly by itself.
I also love the techbro spin: this new thing doesn’t work but it’s totally the fault of the old stuff.
Do it again with moar new stuff that we sell you and this time it will be different.
Hippocratic directly promotes how it can undercut the living wages of real nurses as a feature, not a bug. One page of the company’s website compares a human nurse’s $90 per hour salary to an AI agent’s $9 an-hour running costs. Hippocratic claims its AI nurses outperform human nurses regarding bedside manner, education, and narrowly miss on satisfaction, according to a survey.
