Humanoid robots. Again

[Goofyhoofy]

From todays WSJ:

https://www.wsj.com/tech/ai/humanoid-robot-hype-use-timeline-1aa89c66

Even the Companies Making Humanoid Robots Think They’re Overhyped

Billions of dollars are flowing into humanoid robot startups, as investors bet that the industry will soon put humanlike machines in warehouses, factories and our living rooms.

Many leaders of those companies would like to temper those expectations. For all the recent advances in the field, humanoid robots, they say, have been overhyped and face daunting technical challenges before they move from science experiments to a replacement for human workers.

“We’ve been trying to figure out how do we not just make a humanoid robot, but also make a humanoid robot that does useful work,” said Pras Velagapudi, chief technology officer at Agility Robotics.

And farther on in the story:

Company leaders say there is a narrow set of roles where humanlike robots make sense today, including performing simple, repetitive tasks such as moving boxes. Persona is building a welding robot for a shipbuilding company, a function Radford said is ripe for roboticization because the danger involved makes labor hard to find. For something like robot butlers, the market is farther off, he said.

The cautious, if not downright gloomy, outlook by leaders and engineers of humanoid robot companies stands in contrast to forecasts made by some of the biggest names in technology.

To be fair, there is one paragraph quoting Musk and Huang, whose businesses rely on the enthusiasm of investors to keep their stock prices elevated. And to be fair to Albaby & me, there’s also a section where the story (and industry) question the need for a robot to be humanoid at all:

Ultimately, there is a more fundamental question to answer: Do we even need a robot with arms and legs?

There are downsides to the human form: Robots that look like us are prone to tipping over and engineers struggle to create a mechanical version of the human hand. We rely on sensations from our skin to know how much pressure to apply, something robot builders struggle to replicate. Some engineers say the future isn’t in replicating the human shape but in improving on it, with four hands instead of two, or suction grippers instead of fingers.

“My point of view is that we are sticking to the humanoid form too much,” said Max Goncharov, the chief technology officer at RemBrain. “In the factories, it’s all about efficiency, and efficiency means more specialized robots.”

“I think humanoids will do a tiny layer of tasks in factories in the future,” Goncharov said.

[flyerboys]

I am with goofyhoofy (and the robotics companies he cites) on this one.

Human form is extremely evolved, but mostly evolved for very different uses within environments extremely different from those that we have constructed or created by destruction….

[Leap1]

[image]

Google AI

Elon Musk favors humanoid robots like Tesla’s Optimus because they’re built for our human-centric world, allowing them to navigate and use existing infrastructure (stairs, doors, tools) and perform tasks in diverse environments, essentially acting as general-purpose laborers for dangerous, repetitive, or boring jobs, leveraging Tesla’s AI from self-driving cars for a future where physical work is a choice, not a necessity.

My comment

We take that world for granted, but recreating buildings and infrastructure to exclude humans in effect is a bad idea.

[1poorguy]

@flyerboys

Human form is extremely evolved, but mostly evolved for very different uses within environments extremely different from those that we have constructed or created by destruction….

Yes. Many form factors can deal with our human world. Dogs are the most obvious. They can go through our doors, climb our stairs, and the bigger dogs can take food off the counter. The lack of opposable thumbs is about their only limitation.

Outside, a dog-form will be more stable in uneven terrain, and faster.

In factories, you would want specialized robots that do one task very quickly and efficiently. As the tech officer in the OP said.

[Goofyhoofy]

@1poorguy

In factories, you would want specialized robots that do one task very quickly and efficiently. As the tech officer in the OP said.

The other argument is that you want robots that can do everything, all at the same time. I’m not sure why, robots excel at doing on or two things extremely well, and can be repurposed by reprogramming, adding different end-of-arm tooling, or adding different hardware. There is a thriving market for ‘“used robotics” demonstrating that they are not confined to their single original purpose.

But there are surely jobs in which doing many things would be an advantage: cleaning a hotel room or carrying patients in a nursing home and tending to them. The people in the know about these things (some quoted in the WSJ article upthread) say we are nowhere close to having that happen, however.

[1poorguy]

@Goofyhoofy

The other argument is that you want robots that can do everything, all at the same time

Which doesn’t make sense. When the assembly line came about, you have humanoid (humans) standing in one place (mostly) as pieces went by them. They mounted the widget to the thingy, and sent it to the next person in line. They were specialized for one task. Today, we have robots in manufacturing lines that do the same thing…they work on one bit, and the item moves down the line to a different robot that does a different bit of the process.

[Goofyhoofy]

@1poorguy

When the assembly line came about, you have humanoid (humans) standing in one place (mostly) as pieces went by them. They mounted the widget to the thingy, and sent it to the next person in line. They were specialized for one task. Today, we have robots in manufacturing lines that do the same thing…they work on one bit, and the item moves down the line to a different robot that does a different bit of the process.

Yes, but the argument is that there are lots of jobs which that kind of automation can’t do because it’s limited to doing one thing over and over. Hence: the humans still employed doing things.

Think: hotel maids. They have to make the bed, empty the trash can, clean the sink, check the shampoo container, put the TV remote back in place, etc. That’s a lot of different things and you wouldn’t have “a robot” to do each one separately, so the holy grail is to have one robot do it all.

Now that’s not a particularly high value sort of work - nor is it likely to see falling costs with an army of robots doing it, but perhaps if there is just one, and it’s cheap enough, and doesn’t break down too often, and doesn’t use too much electricity, and can be amortized over 10 years use - it might make some sense.

Perhaps that’s a terrible example, I don’t know, but whenever, wherever I see humans standing around I think: automation? The kids at a fast food restaurant? Do I care if it’s a humanoid robot in the back making the burger or ice cream cone? I do not.

Salesperson in a department store? Yes, I probably care because I have been locked in too many conversations with automation (think: phone answering decision tree robot) to want to deal with that. Retail clerks? Not for me.

Aides in nursing homes? Maybe. Depends on how well they could operate, which we are a long way away from. But an aide in my own home so I can “age in place”? A more likely use, but still decades away.

So I see some, I just don’t see the (investment) enthusiasm for something so far away and so unpredictable except in a “this might be good for humankind, someday” sort of way.

It would be like watching the first connection of the ARPANET in 1969 and thinking “Oh, I must invest in a store selling books by email, because that may be big one day.”

[1poorguy]

@Goofyhoofy

Do I care if it’s a humanoid robot in the back making the burger or ice cream cone? I do not.

Sure. But as you allude in your next sentence, you probably would care about the kid taking the order so you can complain “I asked for this without mayo”.

I hate the phone systems, and almost always starting hitting “0” or speaking “representative” over and over until I get one. Works about 90% of the time.

The maid you mention doesn’t need to be humanoid. Sure, we can automate lots of stuff. The central question in a few threads now is whether we have to make them bipedal, or some other form factor is an option. I contend it’s not only an option, it is a superior option, and simpler to design.

As an aside, we stayed in a hotel in Asia (Seoul). The receptionists were robots (mostly for show…you checked in on a screen). But room service was robots. You wanted extra towels? It would be delivered by robot. If you had a problem with the room, it was hard to get an actual person (which, to me, is a failing of such a system). I’m assuming the maid was human, but never saw him/her.

[DrBob2]

@Goofyhoofy

Salesperson in a department store?
Aides in nursing homes?

You left out an area sure to be a success – sex robots.

All the lonely people
Where do they all come from?
All the lonely people
Where do they all belong?

DB2

[JimA759s]

@DrBob2

You left out an area sure to be a success – sex robots.

This may be the first posting of yours that I 100% agree with!

You had to be right eventually.

JimA

[1poorguy]

@DrBob2

You left out an area sure to be a success – sex robots.

I read a while back that they are already available. I think they called it “Harmony”, or maybe that was the name of the manufacturer. Not sure, too lazy to look it up again. I think it was part of an article about AI robots, which implies this has an AI.

I’m sure it’s not as functional as replicant Daryl Hannah, but I’m also sure one of the first applications for replicants would be sex. And combat. Those are the first two things any new tech gets used for.