Reminds me of the scene from Robocop 2 when they are introducing new robocop models and one literally rips his helmet off and his skull just screams. As a kid I found it both funny and scary at the same time.
This is something that can happen with an autonomous robot if it was trained via imitation learning, which is one of the common ways of doing things when using transformers, and transformers are in vogue right now.
But knowing how tech demonstrations usually work, it’s much more likely that this is actually just a robot being remote controlled by another human.
Yeah, Elon’s tech demos are usually remote controlled
Why it look like whoever was controlling it just took their headset off? User takes viser off robot falls from lack of user controls? Tell me this was not actually just another mechanical turk.
It was AI – Actually Indians
Anonymous Indians since we’ve never met them
i read that omg seems its always just a room full of exploited humans.
Pretty sure in that case, the Indians were the ones doing the exploiting.
Well at least it’s not Turk?
Look, the mechanical Turk was a good chess player at least… and his operator had to literally hide beneath the table…
This is why I know Elon won’t get that ridiculous bonus. He has to sell a million of these robots.
What is even the point of these robots? Like, will he provide the operators as well? Do I have to employ someone to work my robot? Am I supposed to use a robot as a proxy? I don’t get what’s being sold.
The idea is that a robot + guy in india to run it is cheaper than hiring and American
What’s being sold is the illusion that the robots are completely autonomous based on AI or something.
Buy it now, get the full AI self-walking update in
twothreesevenfifteensome years.It worked for Tesla cars, so why should the same strategy not work on the same idiots with robots?
The robots are supposed to be autonomous, not driven by an operator, but no one can make robots except Boston Dynamics and even they have struggled.
Even if they aren’t autonomous you could argue that the dexterity of these bots could be very useful for a human operator to work in a hazardous environment where the human being present directly is harmful.
Now selling them as fully autonomous while pulling off an Amazon checkout is total bullshit.
There’s a singular perfectly awesome use case for a robot you can drive as your proxy. Here it is:

Of course, the chances on Elmo being able to deliver anything even approaching this minimum level of usefulness are well into the negatives.
Given that his cars break from rain, and his rockets always explode, I wouldn’t put much stock in these, no.
Oh god, you’re right. One million robots in ten years. That’s impossible and hilarious.
On the other hand it’s good for the economy. They will have to employ one million people working full time to control them. Musk is such a good socialist.
Could they work for people with disabilites? That’s the only potential I could see for something like this but if you have to physically walk around to operate the robot they won’t even work for that.
It could be great for humanity except for a few details. I only know a few blind people, so my opinion sucks but: they usually don’t have a lot of money to spend, and I don’t think they would appreciate losing their autonomy to a commercial company, even more if the CEO is an unreliable pedophile drug addict that loves Nazism.
But I could be wrong.
I meant like a person who has limited mobility or something like that could operate the robot to do stuff around the house for themselves, instead of having to pay someone. But I guess it could work for a blind person that needs care to have someone else operate the robot for them but they could do so remotely instead of having to come to them. I am also not speaking of the ones musk is developing. Obviously no one should buy anything that dude has a hand in. As for paying for it, I would assume it’d be treated like other medical devices where your healthcare coverage helps with the costs, but yea, corporations will probably do everything in their power to make it as unobtainable as possible.
Thr question is can that thing jack me off and can it look much less creepy?
Did you see the water spray out the bottle at the end? You’re losing it if it grabs anything
Let me introduce you to the wonderful world of teledildonics. Because yes, you can actually wank with open source, not just about it! A good library for your toys https://buttplug.io/ and a prototype I did for it https://twitter-archive.benetou.fr/utopiah/status/1280079465847160832/ activating a toy from within a VR experience.

c/shirtsthatgohard
The Indian operator just finished his shift
He meant AI as in Affordable Indians
https://ia.acs.org.au/article/2025/the-company-whose--ai--was-actually-700-humans-in-india.html
I was gonna post a joke but nothing will top this lol
My optimist: surely Musk will not be able to get away with being caught in this obvious lie.
My realist: he’s gonna get away with it again.
The optimistic take is that no one is that stupid.
The realistic take is that musk fanboys will claim that the robot is so advanced, that it thinks it’s alive and mimics someone taking his headset off, because it has seen it.
I’m sure we’ll see, “the AI was trained on human operator recordings - which unfortunately included headset removal actions, Tesla is now working to filter that from the model”.
The operator typically goes straight to sleep after removing his helmet as it’s a very taxing job. Nothing unusual.
They’re actually shot in the head when their shift is over and they remove the headset - to prevent leaks of Optimus being a charade.
Sadly it’s not even limited to fanboys, just the completely oblivious. I was kinda shocked recently to explain to a work friend how idiotic musk is with receipts. He was perplexed, “… But he’s that smart tech billionaire, right?” Just taking the world at face value.
Just taking the world at face value.
I remember coming into my actualization in my late 20s and making the horrific realization that the vast majority of people never go through that.
If we wanted to to do that we’d say it wasn’t taking headphones off, it was throwing up its arms in panic that it was having a stroke. See the immediate collapse
Remember those Tesla solar roof tiles? Fake. Made by a prop company in LA.
Tesla solar still exists, but they just install typical panels, because PUTTING SOLAR TILES IN THE SHADY ROOF SIDE WAS STUPID.I’ve seen a local installation of the tiles. I asked the contractor all about them as they went in. They are real.
While I really don’t want to defend that flop, they didn’t. Their goal was to create identical tiles so the roof looked consistent but only some would be the more expensive solar tiles
Yeah but they delivered fuck all.
I know theyre banned in my town, fire marshal and electrical inspector said fuck that.
Theres actually a guy who has them on his roof. Hes not allowed to turn the system on, even islanded to make power, and he eats a decent yearly fine just for having them on his roof, which is deemed a hazard by the fire marshal, should he have a fire.
Are you sure? I remember seeing other legitimate roofing companies selling solar roof tiles too. I think I even remember watching an installation video. They weren’t rigid tiles, they were flexible and nailed down similarly to shingles.

What is this from?
It’s a scene from 1990’s RoboCop 2.
There was a RoboCop 2!?
Must’ve been so bad that I blanked it out of my memory.edit: Nevermind lol. I’ll check it out, thanks.
No, RoboCop 2 is alright. RoboCop is probably my favorite movie. RoboCop 2 is good, not as deep, not as gratuitous, nor perfectly campy as the first but good enough for a watch. 3 is not, don’t waste your time… Even Peter Weller didn’t come back.
The remake totally missed the mark; had some interesting ideas and callbacks to the originated but didn’t know how to capture the tone and thought of the original and took itself too seriously.The new game is supposed to be good. It’s been patiently waiting on my PS5.
You guys talking about RoboCop: Rogue City (the one that’s already on PC)?
If so, don’t get too excited. The game is mostly a mindless shoot-em-up. I was bored in under an hour.
Same here. Even with the upgrade tree it looked boring. Some here will argue it’s a great game because of the AA status and small dev team. I get that, but after playing so many other great titles, that game just doesn’t measure up.
I personally like 2 better, but you’re spot on. I would say the game on PS5 is the true RoboCop 3 lol.
I also love that the original is a perfect chiachism: a palindrome of a film, every scene at the start as a mirror at the end.
Robocop 2 is the superior movie IMO. The story, action, basically everything is improved. Definitely check it out if you’re a fan of the first one.
I don’t quite agree, RoboCop 2 lacked some of the depth of the original in my opinion and come on: can you really beat Ronny Cox and Kurtwood Smith as the villainous duo? 😉
Having said that, RoboCop 2 is definitely still worth watching!
I will agree the villains in Robocop were better. It’s been such a long time since I watched but I remember just enjoying the nuke addiction the whole city had lol. Plus the main villain was a nuked out junkie in a massive mech suit. My heart was full lmao. Both are definitely worth watching though! Great movies!
I always found it hilarious that the villain in Robocop 2 is actually called Robocop 2.
Oh sh— I gotta check that out.
deleted by creator
What I thought of immediately.
Excellent.
I came here expecting to post it myself, delighted to not have to put in the effort.
Kind of what I hoped happened, without the corpse. Could have been cool to see it rip its head off.
deleted by creator
Be funnier if the robot ripped its own head off.
I kinda clicked into here expecting that outcome
Robocop 2 is a hell of a movie.
Precisely the scene I thought of too.
He doesn’t shoot them off, he tears them off.
The most recent Defunctland video made me believe we might actually see full autonomous robots performing pretty rudimentary tasks while looking humanoid in the relatively near future. This uh… Isn’t that.
Best we can do send a live video feed from a camera on the robot and have someone in India control the robot.
It’s like having a robot, you just have to accept that someone is seeing everything you do in your home!
2025 version of the mechanical turk
Indeed, a case for Amazon’s Mechanical Turk, a service that is shamelessly and deliberately named after its historical model. The principle has not changed after all this time: poorly paid people do the work to make it look as if machines could perform the task - as if it were an unprecedented technological breakthrough, as if it were some kind of magic.
It is a very popular thing among all the companies that claim that “artificial intelligence” was the future.
Edit: However, this does not appear to be a demonstration of autonomous technology.
Edit edit: Apparently, this was actually intended as a demonstration of autonomous technology - any source other than Reddit would be more credible.
They’ve already demonstrated time and again that these robots only exist so the rich can have slaves without actually having to see or interact with the slaves. They even had their robot strength nerfed so that there could be no uprising.
I thought Mturk was primarily used for labeling data for ML models, i.e., “here’s data. Look at it and give it a label according to our specifications”. Do they have a component of Mturk for piloting devices?
It existed before that use case was prominent. Basically it was for whatever trivial for people but hard for machines task you could have people do over the Internet.
I can’t answer that competently, but I can well imagine it, because there is demand for it.
Mecha Türk intensifies
Remember cyberguy? I like the energy.

With his delicate all these robots move it’s nice to see one throwing some weight around. And it’s holding up really well, you’d assume a connection would come loose after the first swing.
Trial and
ERROR!!!amirite
Elon Musk once again proving that he’s more of a Justin Hammer than a Tony Stark (who was already a problematic figure, but at least had talent and SOME sense of right and wrong)

Elon Musk was in Iron Man 2 because he was the closest thing to Tony Stark. And then Lex Luthor looked and sounded like Elon thinks he does, because he’s the closest thing to Lex Luthor.
Problematic? He was a Genius, Billionaire, Playboy, Philanthropist and the Hero that saved the world!
Everything about Elon Musk is a scam.
I was so confused, I thought it had accidentally done a Nazi salute while removing the headset and then it shut down as some sort of rule.
I had no idea that instead of a person standing there doing whatever a person controls it remotely. What a great idea for nuclear waste clean up, fucking terrible for handing out water.
Sorry to disappoint you, but there’s no way this thing is usable for nuclear waste cleanup. in comparison, the cleanup crews in Chernobyl wanted to use robots to clear the graphite rubble off the roof of the power plant after the accident because of the high radiation levels there, but the radiation was crashing them pretty much instantly, forcing them to use human liquidators.
Components these days are surely even less resistant to radiation, because of much higher density parts which ensures that the memory and cache in this thing would look like after a blender treatment.
Components these days are surely even less resistant to radiation
I’d disagree, we’ve had 4 decades to learn how to better harden electronics in high radiation environments. off the shelf stuff? sure that would be fucked. purposely designed:
https://www.jalopnik.com/these-robots-go-into-fukushima-daiichi-so-people-don-t-1850032340/
Yeah I used to work in a chip company many years ago and we had rad hardened chips that had special outer packages instead of the normal consumer ones at a minimum. They’re typically used for space stuff. Still not sure how well they’d hold up in Chernobyl in fairness.
for specially designed purposes, I’d say from the links above someone’s figured it out.
Yep, we do this for particle accelerators and in other extreme radiation environments all the time.
We are talking about something designed by a company owned by Musk that is already faking autonomy, The tech you write about is specialized for this line of work, and none of it is wireless, humanoid and have the processing power to work autonomously. But i’m happy that a part of this work can finally be done remotely.
wasn’t opining on musks shitware. just the idea that there’s no way to operate in high rad environments.
Have NASA design them. They already do radiation hardening for outer space.
I believe the levels of radiation are several orders of magnitude different. I don’t think you can even use a digital camera for a robot near these open reactors as the signal is completely swamped by the radiation, while in space you would just have a couple of inaccurate pixels at any point in time.
Camera issues are a whole nother issue, but other then tech debt, NASA likes to use older computers because the traces on the chips are bigger and less likely to have their bits flipped. A friend of mine programs for satellites and his systems use some sort of PowerPC chip.
Why would consumer electronics be radiation hardened? But I didn’t to say that we can’t do radiation hardened robots it’s just that these ones won’t be it.
Well you can make radiation resistant electronics or shield them with lead.
But I would design a robot that has at least 4 wheels or legs.
Falling over is sth you don’t want to happen.
You aren’t going to get a wireless connection in a radiation zone.
Yesnt. Just use a different frequency.
Or cables
4 legs, cables, … it’s already tricky to navigate a complex space on 2 legs without cables …
I feel like this is close to https://rodneybrooks.com/why-todays-humanoids-wont-learn-dexterity/ recent piece, namely that, yes, all that is easy to imagine but in practice, to reach the seemingly basic level of movement an average human can do for a similar weight, not 1 ton, is actually ridiculously hard. Biological organisms aren’t magical by any stretch of the imagination but somehow to manufacture an equivalent is not something we are able to do. Each extra wire adds a bit more weight, which in turns needs more powerful servos, themselves making the one below requiring more power too and keeping the whole thing mobile needs a very powerful battery… so yes making design suggestion is easy but fully comprehending the consequences of those choices often means going back to the drawing board.
Fun fact : VR players know quite well what moving while tethered means. I can tell first hand, it’s damn annoying BUT if you don’t manage it, you will both fall and break your system.
Who says it needs to be wireless?
We literally have drones right now tethered in active war zones because of jamming, and they work fine until someone purposely snips the wire.
Hopefully we can ask the mole people in Chernobyl not to snip the wires.
The rate that fukushima robots fails, you are going to.have a load of cables trailing into the radiation zone.
That’s very interesting. Robots are less resistant to radiation than humans? So when robots take the jobs of people, production is made more vulnerable to nuclear weapons?
in a way. Cell damage can be repaired when it occurs in low amounts and even broken DNA strands can be fixed by the machinery in our cells. Most importantly, our systems are very much redundant on a cellular level, losing a few cells is not so much of an issue, since we lose cells every day anyways. Computers have nearly no redundancy; in some cases, a single bit flipped by a gamma ray can cause a system crash in any computer. There is stuff like ECC for memory which helps, but even that isn’t foolproof. Computers for space missions outside of earths magnetosphere are designed to make sure the density of components isn’t too high, with lots of error correction code, backups and a lot of lead shielding, which equals lower performance.
I think you are both overestimating the ability of biological systems and underestimating the ability of mechanical systems to be repaired.
Biological systems have incredible self-repair capabilities, but are otherwise largely unrepairable. To fix issues with biological systems you mostly have to work within the bounds of those self-repair mechanisms which are slow, poorly understood and rather limited.
Loosing a few skin cells is perfectly normal. Corrupting a few skin cells can cancer cancers or autoimmune disorders. Loosing a few Purkinje cells can lead to significant motor impairment and death.
Computers, and mechanical systems in general, can have a shit ton of redundancy. You mention ECC, but neglected to mention the layers of error connection, BIST, and redundancy that even the cheap, broken, cost-optimized, planned obsolescence consumer crap that most people are mostly familiar with make heavy use of.
A single bit flipped by a gamma ray will not cause any sort of issue in any modern computer. I cannot overstate how often this and other memory errors happen. A double bit flip can cause issues in a poorly designed system and, again, are not just caused by cosmic rays. However, it’s not usually that hard to have multiple redundancies if that is a concern, such as with high altitude, extreme environment, highly miniaturized, etc. objects. It does increase cost and complexity though so____
The huge benefit of mechanical systems is they are fully explainable and replaceable. CPU get a bunch of radiation and seems to be acting a bit weird? Replace it! Motor burnt out? Replace it! The new system will be good as new or better.
You can’t do that in a biological system. Even with autografts (using the person’s own tissues for “replacements”) the risk of scarring, rejection and malignancy remains fairly high and doesn’t result in “good as new” outcome, but is somewhere between ‘death’ and ‘minor permanent injury’. Allografts (doner tissues) often need lifelong medications and maintenance to not fail, and even “minor” transplants carry the risk of infection, necrosis and death.
That study doesn’t seem to support the point you’re trying to use it to support. First it’s talking about machines with error correcting RAM, which most consumer devices don’t have. The whole point of error correcting RAM is that it tolerates a single bit flip in a memory cell and can detect a second one and, e.g. trigger a shutdown rather than the computer just doing what the now-incorrect value tells it to (which might be crashing, might be emitting an incorrect result, or might be something benign). Consumer devices don’t have this protection (until DDR5, which can fix a single bit flip, but won’t detect a second, so it can still trigger misbehaviour). Also, the data in the tables gives figures around 10% for the chance of an individual device experiencing an unrecoverable error per year, which isn’t really that often, especially given that most software is buggy enough that you’d be lucky to use it for a year with only a 10% chance of it doing something wrong.
it’s talking about machines with error correcting RAM, which most consumer devices don’t have.
It’s a paper from 2009 talking about “commodity servers” with ECC protection. Even back then it was fairly common and relatively cheap to implement though it was more often integrated into the CPU and/or memory controller. Since 2020 with DDR5 it’s mandatory to be integrated into the memory as well.
gives figures around 10% for the chance of an individual device experiencing an unrecoverable error per year, which isn’t really that often
Yes, that’s my point. Your claim of “computers have nearly no redundancy” is complete bullshit.
It wasn’t originally my claim - I replied to your comment as I was scrolling past because it had a pair of sentences that seemed dodgy, so I clicked the link it cited as a source, and replied when the link didn’t support the claim.
Specifically, I’m referring to
A single bit flipped by a gamma ray will not cause any sort of issue in any modern computer. I cannot overstate how often this and other memory errors happen.
This just isn’t correct:
- loads of modern computers don’t use DDR5 or ECC variants of older generations at all, so don’t have any error-correcting memory. If the wrong bit flips, they just crash.
- loads of modern computers don’t exclusively use DDR5, e.g. graphics memory (which didn’t have error correction until GDDR7 but can still cause serious problems, e.g. if a bit flips in a command buffer and makes the GPU write back to the wrong address in main memory, overwriting something important), and various caches (SRAM is vulnerable to bit flips from various kinds of radiation, too). If the wrong bit flips, they just crash.
- Compared to other computer problems that can put the wrong data into memory, like experiencing a bug because a programmer made a mistake, or even just a part wearing out from age, memory errors are really rare, so anything implying normal people need to care is thoroughly overstating their prevalence.
It also means humans will be progressively pushed into the most dangerous jobs because the robot circuitry can’t cope with harsh environments. The easy cushy jobs will go to the robots.
Could be some exceptions.
Off the top of my head: Anything with poisonous gases. Anything where there’s a RISK of an explosion or something (so the robots would work before the explosion; this is kinda already a thing with bomb disposal robots, isn’t it?). Etc.
So for sure anything nuclear will have to be human, but there could be other environments where robots survive, but humans won’t.
Humans are, in general, absurdly robust. You can absolutely mess them up, and they will keep chugging along for a while before breaking down. Not to mention their almost frightening ability to make a full recovery from horrendous injuries.
Most robots/machines will be more or less completely disabled by a faulty connection, clogged valve, or torn hydraulic line. Sure, you can shield them more, but for stuff like radiation, dust, and harsh environments that cause gradual degradation, you’re going to have a very hard time beating the resilience of humans.
Bleep Bloop… it is clearly advantageous that we use humans to operate in harsh environments rather that robots… Bleep Ding.
And don’t forget cheaper!
Which is why its imperative that little Timmy is sent to the mines despite all the risks and occupational health hazard that will eventually kill them.
it is clearly advantageous that we use humans to operate in harsh environments rather that robots
Robots were sent into the Chernobyl reactor and they stopped working immadiately. Gamma radiation fries circuits.
in the end, they sacrificed soldiers above dumping sacks of cement, and miners below laying a foundation to stop the core melting into the earth.
You can know that isn’t the case because a Nazi salute would be encouraged by Musk, not shut it down.






















