125 comments

[ 0.24 ms ] story [ 183 ms ] thread
I think it's time for whoownsbostondynamics.com
(comment deleted)
A company too ahead of its time?
A neat reminder that most Mergers and Acquisitions are money losing and that robots on legs are a neat idea but mostly useless
Legs are an over-engineered solution if your only use-case is a nice, flat surface.

Outdoors, or even indoors with random flights of stairs thrown in, they make sense as a robust and effective solution.

That’s why I said mostly
So the question is whether “most“ of long term robotics applications are on a simple flat surface or not. I truly do not know, but presumably they have some ideas in mind
False dichotomy. There are plenty of solutions for going over horrendous terrain that don’t involve legs
I know its easy to forget during a pandemic, but the world is mostly outdoors.
And most outdoors is navigable by means other than legs
I honestly think I've forgotten this to some extent. I went out yesterday after a long while and was shocked by how cold it was for January.

I just hope we're not all agoraphobic at the end of this.

Having spider legs could be more versatile, but who would want giant robot spiders hanging around..
In a world made around human bodies, hardly.
It’s a good point

I should’ve said that most of the current solutions are useless

There will probably be good ones in the future

Edit to say: current solutions are mostly useless but as people have quite rightly pointed out there will eventually be some good ones
Makes sense. South Korea, where Hyundai is based, has by far the most robots per manufacturing employee [1]. They will make good use of the technology.

[1] https://specials-images.forbesimg.com/imageserve/5f719cf9f44...

Feels like we said this about previous buyers. All I'm seeing is another sucker to keep the money rolling for more of these sweet viral robot demos that never seem to materialize into much of anything.
Keep in mind that “previous buyers” is a very small sample size, and there are a huge number of contributing variables. It is certainly not guaranteed to work out, but given the value of really nailing a leap in the space, the price does not seem like an unreasonable gamble
Atlas can definitely nail a leap at this point...
The two previous buyers, Google and Softbank, didn't have much presence in robotics, unlike Hyundai. Hyundai recently spun off the automation business to a separate company, Hyundai Robotics, early last year.
True, but this is Hydundai Motor Group, not Hyundai Robotics. The two are separate.

IMO, this is more about Hyundai's autonomous vehicle efforts. A BD robot is almost completely different than an industrial automation robot, the former attempts to use 3D mapping to dynamically generate an unpredictable path around obstacles between two points, while the latter is designed to always follow exactly the same kinematics between hard-coded points. A car with a steering wheel and forward/reverse electric motors is a lot more comparable to a Boston Dynamics robot than a 6-axis industrial arm is.

>> IMO, this is more about Hyundai's autonomous vehicle efforts.

I disagree. As the article points out:

  ..., why not just buy an autonomous car startup instead? Boston Dynamics is more about “robots,” which happens to be dead last on the list above. ...
I think Hyundai's VEX, Wearable Vest Exoskeleton, is inline with what the company is trying to do with BD:

https://www.roboticstomorrow.com/story/2019/09/hyundai-motor...

the sucker is the stock market buyer, but those "suckers" are getting rich
The stock market buyer when the music stops*. The music doesn’t seem to stop these days.
Even if 1 out of 20 gimmicky bots work out or better yet solves a problem another Hyundai platform has makes that purchase worthy.
Well, Hyundai's market cap just went up $10bn on the announcement so the market seems to feel it's a win.
South Korea is also very advanced in using automated weaponry for their military:

https://en.wikipedia.org/wiki/SGR-A1

Soon they may not only have sentry guns, but frickin' bots with frickin lasers attached to their back patrolling the DMZ.

Robot sharks?
Samsung SGR-1 is an interesting product, compared to their usual TVs and phones.
I hope the tech doesn't transfer to the TVs to prevent you removing the adware.
The ultimate test will be retrofitting their robots with chainsaws.
It blows my mind that a chat client is worth over 26 times what Boston Dynamics is.
Well to be fair BD's _first product_ was only released what, a year ago?

If they continue chugging along I can see them generating many, many more billions of dollars of value.

(comment deleted)
A casino is more profitable than a hospital.
As a confirmed capitalist and free market believer, I have to say you've got me. There's not much I can say about that.
As it should be. Hospitals being businesses is weird.
To extend this, a casino that has built a brand has a higher profit margin than a hospital.

If I had $10 million to make something new, I might well build a hospital. I don’t claim deep knowledge of either industry, but I suspect that survivorship bias colors the lucrative casino examples we are thinking of. In contrast, a few hours with Google maps could probably turn up a reasonably profitable site for a new hospital, just based on ballpark supply and demand

Hospitals shouldn't be profitable at all.
Could you add context? What chat client?
Slack was recently acquired for $27 billion, so I assume they're referring to Slack.
Salesforce bought Slack for 27.7 billion dollars.
In this context, many tech chat/social products would fit the bill. Myspace was acquired for ~$600 million, Instagram and Tumblr for $1 billion, WhatsApp for $20 billion. The more recent acquisition was Slack for $27billion.
Same here.

I can imagine Boston Dynamics stepping into arbitrary dangerous situations ranging from construction to rescue to military to mining to anything else.

Slack's tech moat is minor, network effects are not considerable (not enough companies using slack to connect to third parties) and Microsoft is destroying them.

They've been trying to sell their robots to all of those customers for years, but they're still way too expensive for the benefit. As I see it, they're a very risky venture with their value mostly being in a hypothetical bright future.
(comment deleted)
Same could be said of tesla, yet they are one of the biggest companies on the planet now and just made Elon the richest man on the planet.
Yeah, doesn't seem right. I have this same thought when I try to think about the scale of a lot of day to day things. It's bizarre how big something becomes when it gets popular, and how many awesome things struggle.
It's always easy to overvalue products/services that are based on number of users than on actual improvements in a narrow but highly technical field with a super-high barrier of entry.

The number of active users is one of the most enticing metric a VC firm can lay its eyes on.

Edit: For context, the parent comment is talking about Salesforces' acquisition of Slack [0]

[0] https://techcrunch.com/2020/12/01/salesforce-buys-slack

Boston dynamics might indeed advance the state of the art but not be able to capture the value ther advancements bring. Also, how big is their competitive moat? Does, say Uber care if their deliveries are done via Boston Dynamics robots or via some competitor? What if in 10 years some research group makes a new company that can deliver the same dancing robots that Boston Dynamics can?

The concept of how moats work in the context of a social app vs how to monetize it is well understood.

Also I'm not saying that the price of Slack was fair. It's entirely possible that the purchase will be a regretful one for salesforce, similar to the regrets that Tumblr's buyer must have felt.

I guess that means that the buyers think the chat client will take over the world but the robots won't.
Or perhaps would-be buyers do think that maybe robots will eventually take over the world, but not before taking over the company. You surely don't want to be on the wrong side of negotiations that make up the inaugural campaign of the battlebot union.
How much value is created by the chat vs by the robots? How many more users does the chat have?
Yes, "value". This is why making free market dynamics an ideology/religion is insane.
The value created by a new chat application is approximately zero.

This is a problem solved since late 1990s and is now well into extracting the diminishing returns phase. The market for chat applications is nearly zero sum game for a while.

In case of Slack, its continued losses every year is a good indicator of its zero value.

In pondering about this question one should keep in mind that economic value is not the same thing as financial value.
In theory, it’s because the marginal cost of software is close to zero, and margin is the biggest driver of value. In practice most of these companies’ margins are negative.
Me too, though I urge everyone to avoid jumping from a gut feeling to judgements about what each “should” be worth without being explicit about the philosophy that determines your “should”.

There are lots of ways to determine value. I do not think that we are ever going to come to a consensus on the best one, but at least we can strive to understand and be understood.

This is a sale price, so the value is 'money value' and thus a financial issue. The 'philosophy' behind what determines it is quite clear:

What's the balance sheet of the company? What's its profit? What's its potential for growth and rise in profit? How desperate are the current owners?

how about researchers making less than jr dev barely out of school?
How to configure WebPack is worth more than how to compute a Z transform, apparently.
This is why I have a hard time calling web developers "software engineers".

There are definitely some exceptions, but most web devs I have met are glorified technicians stringing together apis while reading medium posts.

... and unless the robots start sharing internet memes, it'll stay that way.
Everything is worth as much as someone else is willing to give up for it.
It blows my mind that Hyundai was actually able to buy BD at a fair price in current environment. I don’t think it would surprise anyone even if they paid $5B for BD.
No real customers, no real products and no real revenue streams.

I’m not even sure just how much IP BD actually has that is truly unique, this doesn’t seem to be something very easy to patent in an enforceable manner.

BD isn’t something that’s easy to scale and dump on someone else this isn’t the type of company that VCs are really interested in these days.

I can't help but read this with a negative tone.

To put a more positive spin on it: for roughly 1/26th the cost of another chat client we could have twice the scary robots.

There are a lot of similar things that we spend a relatively large amount of time talking about but don't actually cost that much in the grand scheme of things (fundamental science, alternative energy, and presidential elections, to name a few).

Just to give some numbers:

- SpaceX in 2019: 46 billion

- ITER: 22-65 billion

- LHC: around 20 billion

- ISS: 150 billion

(with the caveat that numbers for publicly funded projects are very difficult to pin down given the diversity of funding)

Basically all the projects that are slowly pushing us toward the future we read read about in science fiction are on the order of the cost of a chat client (the ISS being the outlier as one of the most expensive public projects in history). To me that's cause for optimism.

(comment deleted)
I get your point, but

a) Slack has 5x as many employees, global market share, consumer ubiquity, etc. It’s simply a bigger entity (in the same way that Home Depot is a more ‘boring’ business than slack but 15x larger in valuation)

b) Boston Dynamics is in many ways a research lab rather than a for-profit company. In research and the arts, things might be nominally inexpensive but their real value is priceless.

Things have value for different reasons - Home Depot has value for its stores and employees: even though it is relatively conceptually simple to build retail outlets and hire people to staff them it still is a major undertaking that takes time. A place like BD is mainly valuable for the knowledge and capabilities of the existing employees/technology.

Being bought for the users (as with Salesforce buying Slack) vs being bought for the knowledge and tooling (many of Apple’s acquihires come to mind).

If you want to look at it that way then an aerated cold beverage is worth 219 times Boston Dynamics.
There is value in Boston Dynamics, but will take long time till they create mass market products. They will have specialized products for special areas, but those will be small series, where running production is expensive and a high margin isn't possible. (If a human is cheaper than the machine there is little reason to buy the machine)

Value for WhatsApp comes from the fact that it threatened Facebook's business model. If people use other than their products they have an issue. There is also an option for revenue growth (not only ads, but also payment processing etc.) Question is how that will play out with current outcry regarding Terms of service changes ... whether that will have an impact. But for Facebook certainly valuable decision.

Slack and Salesforce is more interesting. There is of course the market - others tried to acquire Slack as well, driving the price up. How much it will bring Salesforce (also in Cross-Sales "hey, want some CRM which integrates well with Slack? We have something!") is to be seen ...

I have no idea whether Hyundai made a good deal for itself, but I think it's healthy to approach any acquisition or valuation with skepticism.

Corporate M&A seems especially affected by misperception and bad incentives. Executives often want to expand their kingdoms, even at the expense of shareholders. They also suffer from "shiny new object" syndrome - buying whatever's sexy.

Potentially worse, there's the "winner's curse" in auctions. About 7 years ago, Joseph A. Bank tried to buy its larger competitor, Men's Wearhouse. Men's Wearhouse then surprised Bank's execs by making a counter-offer. They got into a bidding war against each other, neither side wanting to "lose," with Men's Wearhouse "winning" by buying its smaller competitor for $1.8B. And we know, even pre-pandemic, how that story turned out!

Point is: don't weigh too heavily what other people will pay for a company today... weigh more what your analysis suggests the company will produce in the future (not what other people will pay in the future).

Funny I thought you meant when Facebook bought WhatsApp for $19B.
I blows my mind that a company that's sold / leased a few hundred units (?) worth less than $100,000 each is valued at $1 billion.
FYI, this is from December 11th.
Darn, I thought it was that dance video that sealed the deal.
If they can mass produce that dog that can carry groceries up stairs, I’d get one.

Overall, I’d say it’s a play to make them more attractive to Apple as a tech company.

There was a robot mule a few years back.
I realize Korea is an ally, but this seems like tech we wouldn't want to lose the domestic capability to produce. Or tech we wouldn't want to cede the lead on.

Why would Darpa pour so much money in for this outcome?

Even DARPA’s hindsight isn’t 20/20.
I wonder if we overestimate the "value" of replacing soldiers with machines...
Number one american export is USD. They don't seem to be interested in doing much beyond that..
That won't last long. Everyone else wants that role gone. Crypto, RMB, ...
Wasn't the previous owner Japanese anyway (SoftBank)?
Considering that American big tech has now conquered the world, they (the US govt) probably don’t want to appear nationalistic.
Because DARPA invests in high-risk research! They pour lots of money into lots of failures. It’s the nature of research.

DARPA gave up on Boston Dynamics in 2015 - BigDog and it’s like were just too noisy and finicky for use in a combat situation. I believe all the BD robots they purchased are in storage. And it’s not like Boston Dynamics was ever an extension of DARPA. The DARPA collaboration (along with the rise of YouTube!) made Boston Dynamics famous around 2007-2008, but the company has been around since the 90s and always had other projects and clients.

In general, robotics technology does not belong to the US, and as Boston Dynamics’s technology has applications in medicine, manufacturing, and humanitarian aid, this nativist greed is particularly inappropriate.

> this nativist greed

That's presumptive of you.

Do you really think Hyundai is going to use this tech for humanitarian aid, or are you just fishing here?

I want a strong US to counter the fact that the number one economy in the world is a fascist state. A state with more engineers than us and that can quickly duplicate our tech sector.

I'm happy that it's our ally buying this tech, but the US needs to be building more of this technology, not less. If we let our lead be in technology like stupid chat apps, that's a frivolous market with no moat.

> Do you really think Hyundai is going to use this tech for humanitarian aid

No, but like Ford or Mercedes or any other company like then, I believe they would sell this tech to humanitarian aid organizations that might use it.

> I want a strong US to counter the fact that the number one economy in the world is a fascist state.

If you don’t want people to say you’re motivated by nativist greed, maybe try having ideas that aren’t nakedly greedy and nativist? At the very least your nihilistic zero-sum America-first attitude is morally indistinguishable from Trumpism, even if in your head it sounds more noble and strategic.

First of all, China has the largest population in the world. They should be the biggest economy in the world, regardless of how evil their government is.

Second of all, if you were truly motivated by a noble desire to stop global fascism, you should be glad that South Korea has this technology instead of the US! The point should be that democratic countries counter China’s power, not that it should be US v. China. In particular, not only is South Korea more directly threatened by Chinese authoritarianism, but their government is far more stable and democratic than the US. The US has only been a democracy since 1965 and, unlike South Korea, our democracy is rapidly decaying.

But it seems that your nativism is leading you to greedily assert the US should be the most powerful country in the world because something something China fascists, despite the fact that we just saw an attempted fascist takeover of Congress by US citizens. It’s deeply hypocritical and indefensible.

> your nihilistic zero-sum America-first attitude is morally indistinguishable from Trumpism

I voted for Biden, Warnock, and Ossoff. Just because you're left of me doesn't make it okay to apply the "Trump" label so broadly. It's not Trumpism to have some degree of nationalism (reverence for our demoratic institutions, multiculturalism, values, economy). Nor is it to worry we went too far with globalism and created a system that gave rise to today's CCP.

We should want for a better, more equitable America. We should want the same for the world. I think we agree on this.

> If you don’t want people to say you’re motivated by nativist greed, maybe try having ideas that aren’t nakedly greedy and nativist? At the very least your nihilistic zero-sum America-first attitude is morally indistinguishable from Trumpism, even if in your head it sounds more noble and strategic.

I live in America and I want growth. Are we not allowed disappointment over the fact we're hemorrhaging technologies and capabilities?

If you believe this is a zero-sum game, then do you want America's lead to be averaged out over the rest of the world?

I believe all people are equal. I want democratic nations to grow. India, in particular, is poised for great growth. But that doesn't mean I want our engines to stall or level out.

I do not believe all nations or systems of government are equal. China, Russia, and our ally Saudi Arabia are detriments to their people and peace worldwide. These players shouldn't be equals. I would love for their citizens to emigrate and gain freedom.

> The point should be that democratic countries counter China’s power, not that it should be US v. China.

I would agree with you if we had a global democratic union. The moment all nations sign up to protect personal liberty, create a strong worldwide economy with ample opportunity, upward mobility, and commit to defending this in unison -- sign me up. But that isn't going to happen.

Japan and Vietnam have the correct ideas about China. We need to sell them lots of weapons. Europe, on the other hand, doesn't know where its allegiances lie. And the rest of the world is willing to partner with China to get a better deal.

Nationalism isn't Trumpism. The world still needs a strong America.

> despite the fact that we just saw an attempted fascist takeover of Congress by US citizens.

I shuddered at this too.

We need to punish Trump, the enablers in law enforcement and positions of power, and all other parties that made last week's events possible. But to think a bunch of hillbillies stood any chance of destroying our democratic institutions is a bit much.

I agree that this should be called out as an example for now and forever, but do not use this as evidence that America is a fascist state. The results of the election prove otherwise.

> First of all, China has the largest population in the world. They should be the biggest economy in the world, regardless of how evil their government is.

This is the root of our disagreement. Totalitarian regimes should not prosper.

If there was only one government in the world, I'd want it to be democratic. But democracy won't win unless we fight for it. Democracy isn't the default. It doesn't get to exist just because we hope for it. So far, it's just been a blip on the radar. It could disappear in a moment if we don't defend it.

Europe exists because of America. If America stops protecting it, Russia's influence will take it over.

China is the next great enemy. Same story.

> In general, robotics technology does not belong to the US

why? What do you mean?

This alongside the recent Apple car rumours makes Huyndai quite forward-leaning.
I wonder if there is any connection - is there some use apple cars could get out of boston dynamics technology?
Most likely not: those japanese companies make a ton a different products in a ton a various domains, contrary to most american/european companies, so they can do cars and robots and ships and pens and whatever else without many links.

Also it's most likely that Apple will use Hyundai as they did use Intel for some time, until they make their own.

I wonder how this relates to Hyundai's partnership with Apple that was just announced and caused a massive rally of their stock price.
Surely leverage via stock

(Edit: this story was Dec 11th, so clearly not.)

Looks like the Koreans are gonna be the first to get a combate robot.
This article was posted almost a month back ~ Dec 11th 2020. I wonder why no media captured this!
HN thread Hyundai to acquire Boston Dynamics [1] from one month ago. The article [2] from The RobotReport states:

> In early November, Bloomberg first reported that Hyundai was in talks to acquire Boston Dynamics from Tokyo-based SoftBank Group. The Japanese conglomerate is selling off non-core business assets after it was hit hard by a series of soured bets, including WeWork and Uber. In September 2020, Softbank sold Arm to NVIDIA for $40 billion.

[1] https://news.ycombinator.com/item?id=25363981

[2] https://www.therobotreport.com/hyundai-acquires-boston-dynam...

Is Hyundai trying to make a major move here? This now makes a lot of sense coupled with their pending partnership with Apple.
How exactly? You mean something like a self riding horse?
Considering their current market cap, it surprises me that Tesla or Space X didn’t pick up BD. Seems like a natural acquisition for a rather low price.
>Considering their current market cap, it surprises me that Tesla or Space X didn’t pick up BD.

Perhaps companies are leery of accepting Tesla's inflated shares as currency in any deal. SpaceX is still continuously raising cash, so I doubt they have a billion dollars to spend willy-nilly.

If they were, they could still sell them off immediately.
Usually, there are lock up periods for those, plus the uncertainty between when the operation starts being discussed, an agreement is made and shareholders approve. What will Tesla be worth in a year?
Can you explain why? Neither needs dancing robots, and both seem capable of building things themselves if they need it?
Robotics seem like a natural extension of space travel or transportation to me. Or maybe I just associate crazy high-tech stuff with Musk.
Tesla wants self driving cars and space shuttles that control themselves. Robot dogs are not that helpful to either. Maybe for Mars rovers, but I assume he thinks he can do better.
Can't wait for them to attach one of those Samsung autonomous sentry guns to the back of Spot and start patrolling the DMZ with it.
Congrats! Saw a demo of the "Cheetah" in Cambridge about a year and a half ago and it drew a massive, captivated audience. Even if they simply seek to dominate the live robotics sports & entertainment market with their dancing androids the acquisition would be worth many multiples. But I suspect their vision goes beyond Hollywood, or even post-Covid automat style restaurants ;)

Interesting that it was Hyundai and not a domestic partner. Still has to pass trade scrutiny. But the subtext here is that American Manufacturing is slow to adopt robotics vs Asia. This could be a wake up call to a sea change in the culture

Manufacturing in America - A View From The Field

https://workofthefuture.mit.edu/wp-content/uploads/2020/11/2...