[Sixstrings]

Google Buys Boston Dynamics, Gets a Bunch of Scary Robots

Andy Rubin, former CEO of Android, became talk of the tech town two weeks ago when he said that Google would be starting its own robotics company. Now, he's making news again as he quietly announced over a tweet that Google bought Boston Dynamics, one of the major robotics companies, on Friday the 13th.

While the name Boston Dynamics might not ring a bell, some of it's projects might. The company specializes in autonomous animal-like robots that have an uncanny ability to get from point A to point B. SandFlea can leap 30 feet into the air and land on a building's rooftop, while Cheetah can run morethan 29 miles per hour, faster than Olympic sprinter Usain Bolt.

The company also has robots that resemble humans. Atlas is a walking robot that can navigate rocky and uneven terrain. Some of the finalists in the DARPA Robotics Challenge received an Atlas robot to tinker with.

Google did not reveal how much it paid for Boston Dynamics, nor did it say what the ultimate goal of the robotics division was.
http://abcnews.go.com/Technology/google-buys-boston-dynamics-acquires-robots-play/story?id=21233076

It's odd, I'd been wondering for a long time why Google or Apple (with their enormous cash on hand) don't get into robotics. Robotics and AI are one of the next big waves, after the Information Age.

So then i see this bit of news, and I did some more reading, and Google has been doing this for a while now. They've bought multiple robotics companies. Combine this with everything else they do -- driverless AI cars, they have the world's largest computer network, now add that to cutting edge robots and we can see where this is going.

Boston dynamic robots are really amazing. You can see some vids in that article link -- one of their robots gallops like an animal would over varied terrain, up to 29 mph.

What's it mean for the future? Well, for one thing, even less employment for human beings. What happens to a McJobs economy when a robot can man that frier and register? McDonalds has already hinted about this in the face of recent strikes, that right now they could just go automated checkout and be rid of cashiers. My local Walmart has started this too, automated checkout -- and I love it, it's so nice, no more lines no more slow cashier. Yet.. what does that mean for the economy? So many jobs lost.

Admittedly this debate has been going on for a LONG time. Robots in car factories in the 70s. And people worried about ATMs cutting teller jobs. We've had gads of job destruction since, via ever greater tech-enabled efficiency. It's like: first you have a bunch of stores with lots of business owners and employees, then you have just Walmart and big boxes, then one day you just have the big box store mostly automated, no $8 employees at all.

It's an old debate, going back to the origin of "sabateur" and dutch farmers throwing their shows into the windmills. The old argument was that new jobs and types of work get created. However, I think there is consensus now that we really do have a problem emerging -- tech innovation and levels of efficiency are so high, at such extremes, and forever increasing -- that we've already crossed that line where more jobs are destroyed than created.

This is coming in the future, this wave of AI robotics, and I'm really curious to see how it will shake out.

My question for the thread is, what happens to the "McJobs" economy when AI and robotics can do even those jobs? How can consumer capitalism sustain itself if the consumers have no jobs?

Take Google's driverless vehicles -- that right there, something as simple as driving a taxi or shuttle van or bus or tractor trailer truck and the myriad post office and UPS and contruction driver jobs -- that's all actually a massive employer in this country. With driverless vehicles, *poof* all those tens of millions of jobs are redundant.

[Outcast_Searcher]

I don't believe that robots will ever compete, because they will never think.

You can believe what you want, including in invisible unicorns. It doesn't mean you know what you're talking about.

Today's robots don't think. Yet they do things like paint cars more reliably, faster, and consistently better than humans can. Thus auto painters at Toyota, as an example, are ALL robots. And oh by the way, such robots now have multiple levels of safety redundancy to ensure they shut down if some dumb human gets too close (ignoring all the safety signs, for example).

Chess programs don't think. They search. (I know, since I've written a couple of programs and have followed the computer chess software and hardware progress for decades). Human chess players, including the top chess grandmasters scoffed at the idea of chess programs beating humans for decades, even as the computers made steady progress.

Well, today, human chess grandmasters are afraid to face a top level chess computer in a straight bare-knuckle tactical brawl. Why? Because they will be beaten by a 15-ish move deep tactical shot the vast majority of the time.

It's about 16.5 years since Deep Blue first beat human World Champion Garry Kasparov in a serious match. The computers continue to get stronger.

So -- computers can't compete because they can't think? And it takes a lot more thinking to play chess at the master (much less grand master) level than to, say, dump some McDonald's food into the fryer.

...

If you want to see outfits like McDonalds greatly accelerate their moves toward automation, let the clowns advocating for a $15 minimum wage win. That should do the trick nicely.

Fast food companies are already seriously looking at automated ordering via touch screen. Personally, as a customer, I look forward to that -- especially if it makes my meal cheaper and means I don't have to interact with rude people who can't speak decent English or call me "dude" and get my order wrong. Is your argument to this step that such touch screen computers can't think?

Just because today's robots can't think does NOT mean that they can't replace a LOT of the low skilled labor. Moore's law's still exists and as a consequence the computers that drive robotics continue to get MUCH more powerful.

I'm sure this bothers your MSNBC sensibilities. Unfortunately for you, in the real world where businesses can and do earn profits, that doesn't matter.

[Plantagenet]

Define "think".

A robot that that can out think a chess grandmaster is pretty frivken smart IMHO. So is a robot that can beat the best human at Jeopardy.

Getting robots to do things like be bartenders or hotel reservation clerks or take over millions of service jobs will be simple now. It's coming

[Quinny]

The whole point about service IMHO is that it's another wonderful human being that serves you.

I quite like this as well.

http://www.youtube.com/watch?v=UsTIMxeO_ng

[dinopello]

One of the characteristics of thinking is learning and being affected by experiences (inductive learning). So two thinking robots when confronted with identical situatons might respond completely differently depending on what their past experiences were. I've worked on systems like this when I worked in machine learning. One of the problems has always been that people generally don't want a machine that is "unpredictable" - which is the definition of thinking I just stated.

[Sixstrings]

If you want to see outfits like McDonalds greatly accelerate their moves toward automation, let the clowns advocating for a $15 minimum wage win. That should do the trick nicely.

I agree with the rest of your post, but that is an invalid point there.

NO WAGE is ever low enough to compete with a robot, once the tech of the robot is cheaper than a human then nothing can compete with it not even Chinese or Vietnamese.

Fast food companies are already seriously looking at automated ordering via touch screen. Personally, as a customer, I look forward to that -- especially if it makes my meal cheaper and means I don't have to interact with rude people who can't speak decent English or call me "dude" and get my order wrong. Is your argument to this step that such touch screen computers can't think?

And that's the paradox. Outcast, you oppose higher living wages yet you don't like the level of service that $8 an hour buys. These poverty wages have an effect on society at large, making us cruder and going backward and it's just all bad.

What good is all this tech if we just wind up with some kind of Bladerunner dystopia techno-poverty?

[Pops]

Outcast_Searcher wrote:
If you want to see outfits like McDonalds greatly accelerate their moves toward automation, let the clowns advocating for a $15 minimum wage win. That should do the trick nicely.

Soon the robots will be designing and building the robots that design and build the robots that supply everything - kind of a capitalists dream. There is only one catch, no one to fill in the demand side of the equation.

1/0=ERR

[Tanada]

Outcast_Searcher wrote:
If you want to see outfits like McDonalds greatly accelerate their moves toward automation, let the clowns advocating for a $15 minimum wage win. That should do the trick nicely.

Soon the robots will be designing and building the robots that design and build the robots that supply everything - kind of a capitalists dream. There is only one catch, no one to fill in the demand side of the equation.

1/0=ERR

What my ex called the ID10T (I D 10 T) error in programming. Back in the 1980's I speculated that the Japanese bubble would not last because by destroying American jobs they were taking the American consumer out of their sales base. Low and behold their bubble popped and the American mega corporations repeated the experiment by off shoring the manufacturing jobs to China. History doesn't quite repeat, but it sure follows the patterns of human nature.