Like I said they have explained it whether you want to believe it is completely up to you. Some people will never understand it, that is completely understandable.
Andy
Can you post a link to where theyâve explained why a robot needs to be shaped like a human in order to work in human spaces, rather than just assumed it?
1 Like
sooo⌠what form factor should a lifeguard robot have?
Aesthetically, a croc or shark form factor should be invigorating?
oh⌠and no flotation devices. Just teeth and a subscription to âdun dun ⌠dun dunâŚâ.
I donât see the standard âfactory floorâ robot - humanoid or some other form factor; no floats; sink like a lead weight - being âthe oneâ on the beach look-out platform. At the very least it should have those âsand tiresâ that are on âbeach bicyclesâ?
And maybe a location marker - maybe like a diver down flag, only instead of red-white, it could be blue-white to indicate âdrowner downâ?
Hmm⌠I never saw Mitch or Pam picking up trash or putting away beach chairs on Baywatch. They were always focused on âwatchingâ.
What is the sole objective of a lifeguard?
ralph boy, Iâm sarcastic this morning?
2 Likes
Yes I can. But I think it would be better that you find it yourself. It would be much more enlightening to you and then you can bring your research back to the rest of us so we can determine if it is reliable or relevant.
Andy
No need to be snarky. Iâve never seen Musk give anything other than a cursory comment about how the human form is necessary for work in human spaces, despite the innumerable counter examples. Iâve watched most of the AI day presentations, and read his statements on Optimus in the earnings calls - but Iâve never seen him get into that. If you happen to know of a situation where heâs gone into greater depth, it would be courteous to tell me about it or point me to it.
I canât imagine what it would be. Weâre a very, very long way towards have the type of AI we could trust with being a lifeguard.
I like Matt Levineâs formulation of the current state of AI - itâs very good at being mediocre at junior white-collar jobs. Itâs usually pretty good at preparing a first draft of things. Itâs frequently accurate, but prone to occasionally being spectacularly wrong. Which makes it very useful for doing cognitive tasks that are not time-sensitive and are subject to human supervision. âIn the momentâ life-or-death jobs are probably not a great fit.
1 Like
That wasnât Snarky, that was giving you a nudge to learn something on your own. Sometimes people have already made up their own minds on something and no matter what you tell them or show them they wonât believe. When I realized that you would keep moving the goal posts, after your handful remark, it became clear that the only way for you to learn was to find the information for yourself instead of leaning on me to bring it to you. Itâs the lady sitting on the couch eating the bon bon theory.
I have done that only for you to start trying to tear down the credentials of the people that are experts in their field. So please, get off your high horse. I need you to do a little work and stop asking me to give you a brief decertation so you donât have to spend time learning what you think you are an expert on.
Andy
And this, I submit, is the major assholery of the internet, where someone makes a statement, refuses to support it with evidence, and then demands that the counterparty waste his time finding the elusive evidence and then refuting it, at which time the originator will decide itâs not sufficient (âbrilliantâ) and pretends to have taught the person a lesson.
5 Likes
Iâm not moving the goal posts. I deliberately used the term âhandful,â rather than a number, because itâs fairly obvious that the difference between trying something out as a test and implementing something will depend a lot on the context. To use another current discussion, about Hertzâ efforts with EVâs, they fully implemented EVâs into their fleet because they bought many tens of thousands of them, rather than doing a pilot program with a few hundred.
For robots, Amazon has implemented a solid program of roboticizing their facilities. They have many thousands of robots in their warehouses, and tens of thousands of workers. A pilot or test program in their facilities will have more robots than if a small transhipping firm with no more than a few score workers decides to go all-in on robots.
Iâm not tearing down their credentials. I just havenât seen anything where they have explained why the human form is necessary for doing these types of jobs. If youâre telling me they never have - that they should be trusted simply because they say itâs true and theyâre experts - then thatâs one thing, and obviously thereâs no point in discussing it further. But if youâre saying they have outlined their reasoning for why a human form is necessary for a robot to operate in human spaces (again, despite the counterexamples), I would love to read it.
2 Likes
No the assholery of the internet is where someone keeps bringing up information where another person refutes it without even supplying any refuting evidence but his own opinion. Making himself seem brilliant, Then the major assholery coming in with MOCK outrage.
Andy
So now you want to argue about what a handful means?
Andy
I donât. I thought that my use of the term âhandfulâ would have been understood as being a proportionately small number - something that indicated that the company was trying them out or doing a pilot program instead of fully committing to implementing them. Thatâs why I didnât use a number like â7,â because that the difference between a pilot/test program and implementation will be different depending on the size of the firm. A company the size of Amazon, â7â is not enough robots to show that theyâre doing anything more than testing them out and taking a flyer on whether theyâre useful or not. If you prefer that I use a term different than âhandfulâ to capture that sense, Iâm fine with using a different term, and consider my prior comments amended to whatever word you think works.
And again, if youâve got a link to someone actually laying out the actual argument why a robot needs to be human shaped in order to work in a human space like a factory, Iâd love to read it.
2 Likes
I would love for you to give me a link to an actual argument on why a robot shouldnât be human shaped in order to work in a factory. One link on why that would be a bad idea or wouldnât work.
Andy
Iâm not saying it shouldnât be human shaped. Iâm saying itâs not necessary for it to be human shaped.
The actual argument is by counter-example. There are literally thousands and thousands of robots working in factories that arenât human shaped. Robots in factories that donât need to move are built to be stationary; robots that need to move are built with wheels, because factories (almost entirely) have flat smooth floors or have ramps, and generally donât have stairs.
So whatâs the reason for building factory or warehouse robots with legs?
2 Likes
No links, but off the top of my head I could see it being more, and possibly a lot more, expensive to produce life-like robots than generic box robots; also, there my be an OSHA type issue if there were some confusion (at a distance maybe?) about what is a human and what is not, presuming there are at least some human types around and having to respond to an emergency or other safety concern.
Pete
2 Likes
Hereâs an idea: Wheels or tracks are faster and more stable than legs. You donât need a head for anything, so why include it?
3 Likes
You need to invert that. What is the reason for not building robots with legs?
Just one link, one white paper on why this shouldnât be done? Your whole thesis is built on conjecture?
I have never seen one link on why we shouldnât build robots with legs. I have seen many links on how hard it is but never one on why we shouldnât do it.
Andy
That would be easy to fix by making the robots fluorescent orange or some other distinguishing color.
Andy
I understand what albaby1 is getting at but not sure how many are.
If biology could have evolved to create axles, bearings and wheels that can spin 360 degree ad infinitum would we even have legs? Given we do, what is the benefit to 2 legs versus the more stable 4? (and weâve seen 4-legged robots). Or four independently steerable wheels?
We are creating robots in our own image. Are we SURE this is the best image to create them in? Or is it just familiar to us?
2 Likes
Doesnât that answer your question? If you have links on why itâs really hard, and itâs easy to build robots with wheels (because we have literally thousands and thousands of examples), doesnât that provide you with the argument for why we shouldnât build robots with legs? In the absence of some reason to build robots with legs instead of stationary or with wheels, you would implement the design that is easier to implement.
1 Like