HN Theater @HNTheaterMonth

idk what kind of sci-fi cardboard they're working with, but here they're picking up boxes from a single side:

https://www.youtube.com/watch?v=yYUuWWnfRsk

I find the introduction video [0] better shows what it does/can do. So... I guess inserters [1] are a thing, wonder how long until we get the tech for fast and stack inserters. Glad we skipped the burners, good strategy as it is such a mess to reconfigure.

0: https://www.youtube.com/watch?v=yYUuWWnfRsk

1: https://www.youtube.com/watch?v=FLb8TJ6-Z2M

⬐ Stevvo

Is it meant to get distracted by dancing with another robot after 60 seconds of work?

⬐ kevin_thibedeau

Union rules.

⬐ BbzzbB

They're just not properly tamed yet, they'll call in Cesar to instill some discipline and work ethics.

⬐ p_l

On one recent project we had a stacking robot arm we nicknamed inserter-zilla.

It was large, powerful (800kg max moved mass), and it killed few bits of industrial machinery before we patched in enough interlocks.

I believe. You can ask them what's their plan for a moment when their job position gets replaced by a robot or an AWS Lambda function: https://www.youtube.com/watch?v=yYUuWWnfRsk

Introducing Stretch

https://m.youtube.com/watch?v=yYUuWWnfRsk

Here's the video Boston Dynamics released, showing the robot in action:

https://www.youtube.com/watch?v=yYUuWWnfRsk

As far as I can see, boxes are picked up from one side, or the top. Some boxes would collapse due to this uneven distribution of force - a human worker would normally pick up a box from both sides, or the bottom.

⬐ tsujp

I wonder how fast each Stretch can do these movements and how heavy a box they can lift.

I assume that if they are able to reasonably grasp heavier boxes then loading/unloading containers full of heavy _stuff_ would fall more into Stretch's domain (less chance for injury, less people needed per box) whereas human workers who can quickly stack light boxes would remain at the top _for now_?

It looks like (implied) from the product page (https://www.bostondynamics.com/stretch) that Stretch could handle things like load-balancing (literally) itself so say when loading a truck the heaviest stuff goes lowest and closest to the cab while also optimising for packing density. Very cool.

I could see this replacing some/most workers at packing centres for [insert large company names here] if that was the case.

⬐ m463

I wonder if they could design a multi-arm version - think of a constantly rotating octopus.

Having many arms could get the speed way up, and could help with balance too.

multiple arms might even allow repurposing the arms - it could use two or three hands to grasp very odd or heavy objects.

⬐ csomar

It's probably too expensive, and require human intervention (skilled labor) to function. It's much cheaper as it is to hire humans.

⬐ bhuirlylg

Robots won't try to unionize, neither will they piss in bottles and generate a lot of bad press for you.

⬐ loceng

Until you start valuing time as invaluable.

⬐ unishark

You could replace multiple minimally-skilled workers with one skilled engineer monitoring multiple robots. This is how manufacturing is going. Robots don't need to be completely autonomous.

⬐ csomar

It'll take time until it their "prototype" becomes useful. It took some time for Spot to find a niche and its not clear how viable it is, and what the costs are (contact us?).

Paradoxically, raising the minimum wage and giving minimally-skilled workers more rights could play in favor of companies like Boston Dynamics.

⬐ freeflight

> It took some time for Spot to find a niche and its not clear how viable it is, and what the costs are (contact us?).

The Boston Dynamics shop lists prices for the Spot and all the accessories [0]

It doesn't for the Enterprise and Academic versions, only offering "contact us", because orders in these fields often consist of larger numbers which might qualify for bulk/education discounts.

[0] https://shop.bostondynamics.com/spot

⬐ aogaili

I've mixed feeling about this new Boston Dynamics robot. The issue here is that global economical systems are already having lots of capital centralized in the hands of a few (the like of Amazon and their shareholders). Politicians asking those companies to raise salaries and provide jobs, meanwhile, new tech is making it easier to replace the workers.

This kind of progress without proper economical and social intervention is just a race to the bottom for many workers..

I don't think technological progress can be stopped but economist, politicians and sociologist need to really step up and provide a meaningful future vision.

⬐ arketyp

> I don't think technological progress can be stopped but economist, politicians and sociologist need to really step up and provide a meaningful future vision.

No thanks! Providing a meaningful future vision is a task for each and everyone.

⬐ aogaili

Then what the point of government, economist, philosophers and entrepreneurs if everyone can do that?

I don't think everyone is qualified to come up with a vision, the world is complex and you need certain background to do it.

⬐ krona

Exactly, freedom is dangerous since it means we might have think for ourselves and take responsibility for our own lives. I'd rather have government deciding what to do.

⬐ aogaili

If you have been to university or got some educational somehow and can think for yourself, which you probably are given you are in this forum, by all means go, it is a free market, and share with us your vision and insights.

But to assume everyone can do that for themselves is just fantasy and idealistic thinking.

⬐ arketyp

Everyone should at least try. It's part of being a decent human being.

⬐ freeflight

That ignores that we are social creatures living in an extremely interconnected global economy.

Rugged individualism doesn't scale to that level and if we left everything only up to that then we might as well just abolish all institutions, including governments.

Or to put it another way: There's a reason religion has been so successful for millennia, it provides a shared worldview and vision for people struggling to find their own.

⬐ faitswulff

Right, but we live in a society.

⬐ aogaili

Most people can't articulate visions for themselves let alone human society.

Could anyone point me to an author/thinker who can provide a meaningful vision for society in light of this rapid technological progress?

I see Zuckerberg wants to embed us in virtual reality, Elon want us to go mars and evolve to cyborgs with neuralinks, Bezos want us to sit all day and one-click-buy from automated Stretch powered Amazon warehouses, governments want to bring coal factories back and print money, and the rest of us just busy speculating about NFTs, stocks, bitcoins, real-estate and posting on HN and social networks..

Where is the delighted vision for humanity in the next twenty years in the mix of all of that? I'd be happy if someone could share one with me.

⬐ MisterTea

> and the rest of us just busy speculating about NFTs, stocks, bitcoins, real-estate and posting on HN and social networks.

It's sad when someone's idea of creative free time consists of gambling and addiction.

⬐ aogaili

Yeah, I'm personally not into any of that, but I don't hold myself any better than those who do, I just happen to have better support system and got lucky with some skill/goals that keeps me busy.

But I can't help but notice that when faced with uncertainty and job instability many choose to throw money in whatever they happen to perceive the future would be, and unfortunately a lot of those so called "investment" are hyped by other capitalizing on their fears. Economical bubbles are everywhere (real estate, NFTs, stocks, bitcoins, etc.)

I do think the underlying issue is the instability in the economical system. My concern is that this trend toward job elimination and capital centralization accelerated with Covid. And governments, economist, thinkers still unable to articulate a vision/strategy for the future, or even worse acknowledge the issue.

The change is consumer behaviour last year was extremely rapid, those who never bought online in their live are now buying only online and many of those changes don't seem to be reversible.

I really hope I'm wrong, but the way I see it, with the current market economics and centralization, and the rate of change, a lot of us and huge chunk of our society would find itself with no prospect of work or income because our government and economical systems are just lagging behind.

⬐ arketyp

Though I sympathize with your concerns, maybe that attitude is part of the problem? Why do you expect anyone else to have a vision for you/us?

⬐ aogaili

Well, it is conversation.

I'm doing entrepreneurship and I run business so I'm capable of articulating some vision for some part of the world. But I'm NOT in a position to articulate a vision for humanity, and what I see/hear concerns me, and I'm really hoping someone has one.

Also, even if I/You have a vision, I don't think I'm (I speak for myself) in a position to make some major social or economical changes, which seem to me, necessary to avoid the societal catastrophe that might happen when things like this robot gets widely adopted.

⬐ unishark

You assume the labor supply always growing. Eventually world population is supposed to peak and start shrinking. Then accepting more immigrants won't be an option anymore. What happens when most of a population are elderly and unable to do most kinds of work?

⬐ XorNot

This is a sea-change I'm expecting to see in my lifetime which will have a lot of flow on effects. The baby-boomers have had an outsized influence on everything since being an oversized generation means not only have they consumed and held onto more of everything finite (i.e. real estate) but they've also stayed as the principle holders of high level positions for longer (since normal sickness/early death doesn't remove them - another boomer is able to simply step into the role on the basis of experience).

We really haven't seen a world where the labor supply follows a more normal age pyramid, and the markets are all assuming that what's true now will always be so.

⬐ Gravityloss

Think of a simulation with a large number of agents. They have randomly assigned wealth at start.

Agents meet one on one randomly and engage in a transaction. In the transaction, the richer gets a slight amount of money from the poorer one.

If you run this simulation for long enough, one agents gets all the wealth and everybody else gets zero.

It's the nature of the system. There's no good or evil, it's just a simple simulation.

⬐ m463

On the other hand, will it make goods less expensive for everyone?

I read "the box" about container shipping and the amount of labor in the global shipping business before containers was ludicrous.

https://en.wikipedia.org/wiki/The_Box_(Levinson_book)

⬐ chromanoid

https://www.krisis.org/1999/manifesto-against-labour/

⬐ chriskanan

UBI+existentialism

That's basically what Star Trek has for a post-scarcity world. It assumes that people would be motivated to better themselves or the world, be curious, etc. despite not having to work.

I think some people would flourish in that environment and presumably still be contributing. Many scientists were basically trust fund kids before the 20th century. They didn't need to work. But there is a good amount of worry some will just derive no meaning from life and turn to social media, video games, drugs, etc. (The WALL-E future)

There is a huge amount of thought going on, but I think the open questions are regarding how effective the solutions will be. We as a society may need to shift our values considerably for any solution to be viable.

⬐ the_duke

The problem is that people still need something to do that gives some meaning to their existence and provides accomplishment, intrinsically and within their peer group.

There are only so many occupations that can do that.

There's art, but if ever fifth person is an artist, who is supposed to consume all of it?

Many are also not wired to be scientists.

With machine learning steadily becoming helpful in more domains and slow but steady improvements in robotics, we could be just 50-100 years away from a future where most manual labor and even a lot of the lower level intellectual labor of today becomes redundant or significanntly diminished.

I'm sure there are models that can provide a meaningful existence, but they won't be so easy to find. And that's ignoring the possible outcome of giant mega corporations that completely control the masses , since every aspect of live becomes dependant on what they provide.

We are headed for very interesting times.

Of course that assumes continued progress, and as history shows that is not at all a given. All we'd need is a serious global conflict with nuclear weapons or local revolutions for progress to be reverted or stalled for a long time.

⬐ bobajeff

>The problem is that people still need something to do that gives some meaning to their existence and provides accomplishment, intrinsically and within their peer group.

I seriously doubt that's something anybody has trouble figuring out. The issue most people tend to face is finding something to do that will provide income. Especially, since all of the things we would want to do requires money. Even if it's just watching sports and talking to friends drinking beer.

⬐ jpindar

I've noticed the people who say things like "we can't just give people money because they need the meaning and dignity of work" already have plenty of money and a high-status office job.

⬐ aogaili

Precisely, because what they fear is loss of control over those underneath them and not the loss of meaning.

As it stands in today politic/economical systems (specially in the west) have an inherit artificial scarcity by design to allow a form of soft behavioural control/governance.

You do X, we will fine/pay you Y amount. And this will only work if the Y amount is impactful enough for the individual.

I think this could be replaced by some other from of social point system, while providing the necessities of life to ensure dignified living when income is lost. Because as it stand today, individual existence and any dignified livelihood is very fragile and highly dependant on a single source of income.

You work at Amazon warehouse to pay your rent/bills, Amazon starts using Stretch the robot and you are left in a very weak and vulnerable position, at the mercy of the upper management, wondering when they might let you go because, thanks to Stretch the robot, you are redundant now. This has to be fixed sooner or later, because it just broken and getting worst by the day really fast.

The issue here is that the erosion of jobs is happening from outside and inside. Due to covid, many consumers started purchasing online which is dominated by few mega players (Amazon), so many local retails, restaurants are losing job. And at the same time, those mega companies are investing heavily to gain efficiency as they try to meet the demand and increase their margins, so they are adding more automation.

The way I see it, this is a race to the bottom, and a lot of workers and families will get hurt if this trend continues without some major economical and social solutions.

⬐ aogaili

I also think this is the case. I see many people perfectly happy and keeping themselves very busy with hobbies, socializing, games etc. Also, content and entertainment platforms are growing, so I think most people will have even more to do with their time in the future.

It does seem to be mostly an income issue.

A lot of UBI skeptics are quick to point out the psychological impact of living in society with no work, but I think the real stress is living in a society that dependant on income while watching the jobs get automated by the day and your leaders have no clear/meaningful vision for the future.

⬐ slightwinder

> There's art, but if ever fifth person is an artist, who is supposed to consume all of it?

Other artist for example. It's not that unusual to have small isolated echo chamber. And as long as someone pays the bill, people will just do something even if they are not the top predator of the trade. Just look at all the hobby-creators out there who are busy creating stuff, despite not earning anything at all from it, often not even big responses.

⬐ helios_invictus

There are plenty of work out there that low skill workers could be doing, that requires uniquely human hands. Some of it doesn't seem that awesome, like say pickup litter, but some of it could be very cool like building trails or planting trees, or building sustainable homes. There are still lots of low problems that can only be solved by human hands.

⬐ CodeGlitch

> The problem is that people still need something to do that gives some meaning to their existence and provides accomplishment, intrinsically and within their peer group.

Parenting.

> There's art, but if ever fifth person is an artist, who is supposed to consume all of it?

Don't forget charity work, nursing, education - i don't think any of these will replaced by machines soon.

> Many are also not wired to be scientists.

See above. How about explorers? Sports enthusiasts.

> With machine learning steadily becoming helpful in more domains and slow but steady improvements in robotics, we could be just 50-100 years away from a future where most manual labor of today becomes redundant.

It happened before : see farming. We found other things to do :)

> Of course that assumes continued progress, and as history shows that is not at all a given. All we'd need is a serious global conflict with nuclear weapons and progress could be stalled for a long time.

Massive progress is made during times of need. WW1 and WW2 basically created aircraft, rockets and computers. The current pandemic is pushing medicine forward at a rate not seen in years.

⬐ mixedCase

>UBI

Ok, how do you fund it?

⬐ ralusek

Get rid of all other social safety nets. No social security, no affordable housing, welfare, food stamps, etc. Enormous amount of bureaucratic overhead removed from that. Everyone just gets a flat rate every pay period, virtually no bureaucracy or perverse incentives involved. Any dollar you earn in the market is purely in addition to what you receive from UBI.

Wealth transfer/inequality is handled on the tax side.

⬐ mixedCase

That's not enough to fund it, if this link's budgetary data is to be believed https://www.thirdway.org/memo/five-problems-with-universal-b... (scroll down to the graphic under the section "UBI is incredibly expensive"). You'd also need to cut down medical care programs.

⬐ weinzierl

UBI = universal basic income (to save others a Google search)

⬐ mdorazio

Can anyone with experience in a warehouse environment comment on the need for this robot to pick up boxes from the top? When I used to pack and ship things for my small business I always lifted heavier boxes from the bottom. The suction cups mechanism seems like it might be somewhat limited to lighter/very smooth boxes, but I'm not sure.

⬐ isomorph

I’m impressed by this thing lifting cardboard boxes by sucking on the top of them. I’m surprised that it works at all, and that the top of the box doesn’t rip off

⬐ thatguy0900

These are all preprepared boxes, maybe it wouldn't work so well on a random sampling of Amazon packages

⬐ liamkinne

My university won a competition run by Amazon, attacking this exact problem. Mostly it was a computer vision problem over actual handling of goods.

https://spectrum.ieee.org/automaton/robotics/industrial-robo...

⬐ motohagiography

I wonder if you can train/program them by recording a persons movements with a haptic feedback suit and piping the data through a model.

Given tech is necessarily an expression of economics, it's not so much the unit price, but the cost of a warehouse replaced by these that they will scale to. They can work 24/7, will probably be leased as deductible opex, carry no legal risk, and even exorbitant repair and support costs will be worth it. There will still be roles for model trainers and ML supervision, just like we have forklift and backhoe operators today, but fewer of them.

I've said before if a task can be specified in a collective bargaining agreement, it can be automated. It's not luddism, I think we're really at an inflection point on automation.

⬐ shantara

Stretch looks to be much more mundane and practical than Spot, which while impressive from a technical perspective always struck me more as a tech demo, than a useful product. Warehouse automation industry is booming and BD/Hyundai want a piece of it.

⬐ umeshunni

BD?

⬐ throwawayffffas

Boston Dynamics.

⬐ BuffaloBagel

I regularly ship 50lb boxes which are 13x13x13 inches to Amazon fulfillment centers. No way they can be picked up and moved by suction on one side.

⬐ jpm_sd

Boston Dynamics admits, robots do not need legs to be useful. In fact...

⬐ Daho0n

Interesting. I wonder how Hyundai will use those themselves.

⬐ gens

Looks janky (the whole thing is shaking a lot every time it moves, especially the sensors). If it's cheap, it's good.

⬐ lawrenceyan

Humans need not apply.

⬐ williesleg

Wait till those fuckers get hacked by china.

⬐ amelius

Yes, this is possible with machine learning today.

⬐ arnaudsm

Boston Dynamics do not use machine learning, but good old control algorithms (ZMP probably)

⬐ bobajeff

I'm pretty sure they do. I remember it being mentioned in a video about them or the robotics industry. At least for the robots with legs.

⬐ iandanforth

Some things to note here.

1. It's quiet. This is a roboticists way of saying, "Hey, I'm efficient and well made!"

2. There are no 2x/3x/5x disclaimers on the video. This means that these movements are real time. Standard for BD but still a flex in the robotics world.

3. Pinch points, pinch points everywhere. This thing is not inherently safe to be around. If you don't trust the perception and safety software don't get near this thing. Also it could probably smash you.

4. They illustrate the tug capabilities (pulling the extensible conveyor) but omit the bit where a human has to connect the robot to the conveyor.

5. The first bit shows an end-effector with dozens of suction cups picking up fairly heavy loads. That's a lot of vacuum to generate. I'd really like to know the runtime of this thing before it has to recharge.

⬐ bobajeff

When watching the video the first thing I noticed was how perfectly stacked and uniform in size and shape the boxes were. It hinted at being able to lift a specific non-box object (spot mini) but I feel it wasn't a realistic representation of actual cargo loads.

Also, unloading is only part of the problem. Loading is more like a game of tetris. Only you have to make sure the packages don't crush one another, topple over, and use the space efficiently. Keeping in mind that some packages are heavier/lighter/bigger/smaller than others or oddly shaped (think tires, weights, chair etc.).

⬐ squarefoot

"but I feel it wasn't a realistic representation of actual cargo loads."

Completely unrealistic since all boxes were empty [0], and there's a cut when it lifts the only heavy object, which also is done at much lower speed. It's a demo, although still very impressive. I hope one day to see BD selling robots in kits.

[0] Easy to infer from how they move when picked up or fall on the conveyor belt; also well loaded boxes would easily break or at least strain if picked up by one side.

⬐ Doxin

As to point number 4, I think the idea there is to have the bot permanently connected to a conveyor like that, and use them to unload or load trucks. It's a pretty useful use case, even if it does mean the bot can't roam the factory floor.

⬐ polote

> The first bit shows an end-effector with dozens of suction cups picking up fairly heavy loads.

That seems like a light load, about 1kg

⬐ shellfishgene

I assume if they can build a robot to recognize and pick up individual boxes they can find a way to have it find, connect and disconnect itself from the conveyor...

⬐ bhuirlylg

That extensible conveyor looks pretty ghetto.

Surely production ones will be more robust and have automatic linkage.

⬐ jabroni_salad

Every walmart in the country has that extensible conveyor for unloading trucks.

⬐ freeflight

It's not ghetto at all, conveyors like that are in use all over the logistics sector.

⬐ nimbius

as someone whos worked alongside caged articulated robots, i can think of nothing more terrifying than a multi-axis, unrestricted pick and place on wheels. when a robot crashes, it generally does so with an astounding amount of uncontrolled speed and power.

https://www.youtube.com/watch?v=7FwdMjYUyKM

https://www.youtube.com/watch?v=Cpg_3syMT_U

stretch seems like the fastest way to the goriest obituary. its also noteworthy to state most robotic lines can be powered off in an emergency at several points. Try convincing a fire team to enter your factory to extinguish a blaze with a dozen flaming Boston Dynamics robots wandering the floor in various states of operation.

⬐ iandanforth

You're not wrong :) Do the robots in those videos use hydrolics?

⬐ mam2

there are hundreds of safe robots with the proper software. it does need to be a priority for the teams working on it

⬐ BoiledCabbage

The other concern is that they've starting to be deployed via policing - and all of the consequences that will entail.

The laws aren't yet prepared for this and it will cause the same types of impacts on society. A lawmaker put forward banning arming them, which seems like the bare minimum [1] - but what happens when they're used to "herd" protestor? Is that a good thing or bad?

We as technologists keep repeating the same mistake, of developing new tech, ignoring the impact it will have on society, and then being shocked at the negative response by broader society.

It's disappointing to see almost no discussion of the impact of tech here on HN, just superficial discussion of the tech itself.

As a software engineer, what matters is not the code you write, but the impact is has on the customer. It takes most engineers many years to learn this.

As a technologist, the exact same thing applies.

[1] - https://www.wired.com/story/new-york-lawmaker-wants-ban-poli...