Cool... their biggest failure pushed them to find what they are actually good at.
It's robot from 1990 and no, there is no video of the robot actually walking.
I opened the article expecting it was going to be about clone robotics https://youtu.be/5mSE6Tkhy4g?si=tDp0DUI9OOXAwsX2
Nightmare fuel
Makes perfect sense, in nature you have a lot of both practical and odd functionality out of filling "bags" with air or liquid.
This is a pretty cool approach. If they can improve the visual presentation it can also look pretty awesome. Gives me some inspiration for drawing scifi designs too.
Liquid seems like a better approach from an engineering standpoint because it is non compressible. But then I imagine dealing with liquid is more of a pain than air.
This is what non-commercial tech looked like back before the gold rush and vulture capital. Geeks and nerds in basements doing weird stuff that would be laughed at by most people on the street. Most STEM professions were middle class, not lottery tickets.
I don't think the pay is really what changed all that much. Median pay for SWEs according to ZipRecruiter is $118k, which is $40k in 1987 USD (the year the person in the article started). BLS data from 1987 puts that at mid-way through the 4th income quintile, which is middle clash-ish.
Levels.fyi slants towards the higher end of pay and they say $192k, which is $65.4k in 1987 or right at the bottom of the top quintile.
Part of what changed is that software abstracted enough for hardware and software to become separate fields, so a much smaller portion of software folks are able to wire together batteries and motors and what not to go with their software.
What do you mean? Things like those still happen, and probably will for a long time. They did end up being slightly nicer looking though, cheap 3D printers and CNC machines really increased a baseline of what hobbyist in the garage can do.
Go to website like https://hackaday.com and there will be plenty of projects like those. (Although this one is more complex than usual, so you might have to lookover a few months' worth of history to find something on that level)
A guy named Walker developing legged-robot software is even more on the nose than a guy named Karpathy developing autonomous-vehicle software.
Oh my god, how have I never noticed Karpathy and Car-Path-y? Amazing!
A noble principle, albeit not without its lamentable failures.
Car-(em)pathy... Now I can't unsee it!
Thanks.
I would love to see this with nitinol wire muscles.
Power use would be immense and it would be insanely slow.
Power use might be high depending on configuration, but speed shouldn't be that slow using capacitors. Sufficiently strong pneumatics tend to require quite a bit of power too.
Considering 90%+ of the input energy goes to heat with NiTi actuators, Your walking robot would also double as a great space heater.