Well, to start with, If you drive your car too fast, you might kill people. If you drive your computer too fast, it might get data corruption and/or reboot unexpectedly.
I could've sworn Peter Sripol built one while on Flite Test too, but I can't find it atm.
Looks like the rocket pendulum fallacy is about expecting meaningful passive stability from the location of the center of thrust vs center of mass, but even 2-rotors need to be able to tilt independently (or deflect thrust) for active control.
Theoretically this could still work even if the center of mass was above center of thrust, but the tilting/vectoring responsiveness would need to be very high. These RC models at least move so slowly that the air resistance of swinging back and fourth really does help dampen oscillations passively, but they all still have active flight controllers that are trying to keep angular velocity at zero without control input.
I dunno if you've seen the subreddit, "Sub Simulator GPT2", but I found it around 2020-2021. It seemed to contain GPT2-style models trained/finetuned on several popular subreddits, talking to each other as stereotypical regulars of each sub would. Most of the replies were fairly coherent and somewhat related to the "thread topic", but of course even GPT3.5 would make all of them look beyond drunk only a few years later. I already had a vague understanding of neural networks and the advances in image processing at the time, but couldn't have predicted where we are now. I wonder what it'll look like in a few more years as we continue how to learn how to make this capability useful and reliable, and hopefully sometimes keep finding additional conscionable entertainment and educational applications.
This feels rather overstated. Your firmware check isn't going to tell you if a repo package installed/updated by your package manager got compromised, which is apparently increasingly common.
The tone is written as abrasive to anyone who doesn't already agree, which shows this is more of an emotional opinion piece than open minded objective research.
Hype cycles never last forever, but that doesn't mean all the value has been tapped by any means. The fact that modern GPUs can solve ridiculously complex high dimensional functions is a superpower in every possible field of research.
Privacy and offline operation are valuable or non-negotiable in some cases, but the difference is pretty categorical between what can run on a single card and what can run on a DGX GB200 NVL72 cabinet. Doesn't mean it's not worth seeing how far local models can be pushed. Not every problem needs a senior engineer.
Mercury is orbiting partially inside the Sun, and Jupiter is nearly as wide as the Sun when it should be 1/10 as much, so the planet nodes should be scaled down 10x relative to the Sun.
Also, I did a top-down pixel measurement, where I could see the distance to Tau Ceti as well as the orbit of Neptune. The radius of Neptune's orbit was 32px, while the distance to Tau Ceti was 1152px, for a ratio of 36, when in reality, Tau Ceti is 11.9 ly away, while Neptune has an orbit radius of 30 AU, which means Tau Ceti is around 25,000 Neptune orbits away, so the planet orbit scale is too big (or distance to other stars too small) by a factor of ~694 (25000/36)
Edit: Since this was top-down, the vertical displacement didn't factor into the distance, which also contributed to Tau Ceti appearing too close on screen, so the error is slightly better than that, maybe a factor of 600.
Edit 2:
Tau Ceti is rendered at 3.652 pc × 3 world units/pc = 10.956 world units
Neptune’s orbit radius is rendered as 30.05 AU × 0.0065 world units/AU = 0.195325 world units
The rendered ratio is 10.956 / 0.195325 = 56.09 Neptune-orbit radii
The real ratio should be 25,067.5 Neptune-orbit radii
More like people try doing anything other than use the base OS, and realize the bottom-tier x86 mini-PCs are 3-4x faster for the same price, and can encode a basic video stream without bogging down.
If the RPI came with any recent mid-tier Snapdragon SOC, it might be interesting. Or if someone made a Linux distro that supports all devices on one of the Snapdragon X Elite laptops, that would be interesting.
Instead, it's more like the equivalent of a cheap desktop with integrated GPU from 20 years ago, on a single board, with decent linux support, and GPIO. So it's either a linux learning toy, or an integrated component within another product, and not much in between.
Double-sided USB-C connections require a handshake before sending voltage. USB-A ports can have the 5v line active at all times. Cheap USB C gadgets often don't make the handshake, they just use it as a 5V input, necessitating an A to C cable.