:00:00. > :00:00.on five people on Thursday night in which acid was sprayed
:00:00. > :00:00.on their faces in order to steal their motorbikes.
:00:00. > :00:07.Now on BBC News, it's time for Click.
:00:08. > :00:15.The cyborgs are coming, the eyeborgs are watching,
:00:16. > :00:47.the bar staff are serving, and Lara photographs a banana!
:00:48. > :00:49.This is Adam Jensen, star of the video game
:00:50. > :01:01.Set in 2027, the poor chap has to undergo extensive
:01:02. > :01:02.cybernetic modifications after being severely injured.
:01:03. > :01:05.Well, just ten years before those events might occur,
:01:06. > :01:20.that plot line doesn't seem that far off.
:01:21. > :01:22.For years now people have been body hacking,
:01:23. > :01:24.giving themselves extra abilities and, as our understanding
:01:25. > :01:26.of robotics has advanced, so has our creativity.
:01:27. > :01:29.Meet Rob Spence, like the cyborg in the video game,
:01:30. > :01:36.It doesn't have Terminator vision like this, yet,
:01:37. > :01:40.Inside a prosthetic eye, which is an odd shape,
:01:41. > :01:42.they're not a sphere, a prosthetic eye, they're actually
:01:43. > :01:48.Inside that is a battery, a video camera and a video
:01:49. > :01:51.transmitter all attached to a circuit board so they can
:01:52. > :02:09.The camera is turned on and off with a magnet.
:02:10. > :02:13.It doesn't look at all comfortable, is it in anyway comfortable?
:02:14. > :02:20.The first consideration that looks the most uncomfortable,
:02:21. > :02:23.it looks like a 90s iMac, you can see all the goods inside.
:02:24. > :02:26.Like the battery and the wires, but that's covered by smooth
:02:27. > :02:31.I don't have open wires and batteries, you know.
:02:32. > :02:43.That kind of made my stomach drop a little bit when I saw that.
:02:44. > :02:46.Rob damaged his eye when he was nine and in 2009 began exploring
:02:47. > :02:50.As a film-maker himself, he was fascinated with the idea
:02:51. > :03:06.It's like an absurd toy for a one-eyed film-maker.
:03:07. > :03:11.I used to watch the Bionic Man when I was a kid, The $6 Million Man.
:03:12. > :03:14.I had the action figure, you looked through the back of his head,
:03:15. > :03:19.I was looking at my Nokia flip phone at the time I was like -
:03:20. > :03:23.That's in fact who I called, I called Nokia.
:03:24. > :03:26.They said - well, we'll call the camera module people in China.
:03:27. > :03:30.It's very small, it's very challenging.
:03:31. > :03:34.It does visual dropouts, which is the visual language
:03:35. > :03:36.of all video from the future, including Princess Leia
:03:37. > :03:45.Since the initial prototype, Rob and his engineers have gone
:03:46. > :03:50.He now has one eye that glows red when it films and another camera eye
:03:51. > :03:55.I get calls from and emails from mom's whose kid has just lost
:03:56. > :03:59.an eye because it's some sort of fun thing to show a kid this maniac
:04:00. > :04:03.running around on videos and glowing red eye cameras and stuff.
:04:04. > :04:08.They're now looking working on ways to transfer the technology to other
:04:09. > :04:12.We're doing 3D scans of those now and then that creates a space that
:04:13. > :04:16.you can take into software to map on the technology that we're
:04:17. > :04:20.Some people golf, I like to make fake eye cameras and, you know,
:04:21. > :04:45.Right, that's the eye upgraded - now let's do the rest of the body.
:04:46. > :04:50.MIT's media lab is home to some of the most innovative tech
:04:51. > :04:53.research in the world, but there's one room here I find
:04:54. > :04:59.The mission of this lab is probably one of the most
:05:00. > :05:01.important goals of our time, they're trying to essentially
:05:02. > :05:05.They want to make it so that if you lose a limb,
:05:06. > :05:09.it won't have any impact on your quality of life and they're
:05:10. > :05:25.So we work on everything from creating new motors and designs
:05:26. > :05:28.for ankles and knees and artificial joints, all the way to marrying
:05:29. > :05:31.these biomechatronic devices with the human body through novel
:05:32. > :05:35.Evidence of this work can be seen with people like Ryan Cannon,
:05:36. > :05:38.complications after a broken leg left him needing an amputation.
:05:39. > :05:41.What's special about his new robotic leg is that it's doing something
:05:42. > :05:44.the human body can do instinctively, but it's extremely
:05:45. > :05:57.The motor is able to work in such a way that simulates a real
:05:58. > :06:00.It uses on board sensors to interfere whether the leg is,
:06:01. > :06:04.for example, in the air or on the ground and perform actions
:06:05. > :06:06.that to the person feel much more like real walking
:06:07. > :06:08.than they would get from a passive prothesis.
:06:09. > :06:14.For amputees like Ryan such innovations are life-changing.
:06:15. > :06:18.I can move in a more rhythmic, symmetrical way and being able
:06:19. > :06:21.to move in that manner allows me to walk at a faster pace
:06:22. > :06:25.for a longer distance and to do more activities during the day.
:06:26. > :06:28.This is not relying just on straight physics and mechanical design,
:06:29. > :06:41.Not all of the research here is about solving disability,
:06:42. > :06:43.this exoskeleton project is about augmenting humans.
:06:44. > :06:46.It allows the body to use much less energy when running or walking.
:06:47. > :06:49.It improves your ability to walk by 25%.
:06:50. > :06:52.So what that means is, if you were to walk 100 miles,
:06:53. > :06:54.it would only feel to you that you walked 75.
:06:55. > :06:58.We're able to do that today, right and those are devices that
:06:59. > :07:00.I would expect to see rolling out commercially in the
:07:01. > :07:05.We're already beginning to see this kind of technology deployed
:07:06. > :07:08.US retail chain, Lowes, is experimenting with kitting
:07:09. > :07:11.out its staff with exoskeletons, designed in Virginia,
:07:12. > :07:13.which could give their employees more stamina at work.
:07:14. > :07:16.With this in mind, the lab at MIT is now looking
:07:17. > :07:20.to the next huge question - how close are we to the point
:07:21. > :07:23.where people might actually want these kind of prosthetics instead
:07:24. > :07:27.I definitely think that we are entering an age in which the line
:07:28. > :07:29.between biological systems and synthetic systems
:07:30. > :07:53.But what might be some of the drawbacks of having these
:07:54. > :07:57.As there's widespread uptake, that they might only be available
:07:58. > :08:08.to people who have the financial ability to pay for them.
:08:09. > :08:24.Welcome to the week in Tech. This week saw some interesting activity
:08:25. > :08:28.on Facebook, and saw Wiz Khalifa take over as the most watched
:08:29. > :08:37.YouTube video. It has been viewed a staggering 2.9 million times. Elon
:08:38. > :08:41.Musk launched a new vehicle, it is supposed to be more affordable than
:08:42. > :08:53.the previous efforts, which cost $200,000. Faraday future have
:08:54. > :08:59.scrapped plans to build a plant in Nevada, which leaves questions about
:09:00. > :09:03.their new vehicle. And this is not a digital version of the Ministry of
:09:04. > :09:07.silly walking, but this is artificial intelligence attempting
:09:08. > :09:14.to learn how to walk. So far, research is being led in virtual
:09:15. > :09:18.environments, but it could help robots learn how to navigate
:09:19. > :09:28.unfamiliar spaces. And, finally, a former scientist built a super
:09:29. > :09:31.Psycho which can fire a jet of water at over 200 mph. At least you'll see
:09:32. > :09:34.it coming -- super-soaker. It says - fashion, style,
:09:35. > :09:48.outfit, that's you. Sometimes it's not that easy to put
:09:49. > :09:53.into words what you want to search for online, and that's why companies
:09:54. > :09:56.are working on ways of us being able to take a simple picture and then
:09:57. > :10:02.search using that image. Pinterest is a place
:10:03. > :10:05.all about images and ideas, they've had a form of visual search
:10:06. > :10:08.for a couple of years now, allowing you to focus
:10:09. > :10:11.in on a particular object Through a combination of image
:10:12. > :10:14.recognition and the data points attached to that image,
:10:15. > :10:16.including the hundreds of thousands of boards it may
:10:17. > :10:19.have been pinned to, This January, they upped their game,
:10:20. > :10:36.though, launching Pinterest Lens, a way of being able to search
:10:37. > :10:40.through a photo with no other data And from that search term,
:10:41. > :10:44.it aims to come up with similar There we go, we've got a picture
:10:45. > :10:48.and something is emerging. Right, those are definitely
:10:49. > :10:51.shoes, but they don't Black shoe with blue
:10:52. > :10:54.laces, some men's shoes. So there are two parts
:10:55. > :10:58.to visual search. The first is computer vision,
:10:59. > :11:01.which is a way of translating the information coming
:11:02. > :11:03.in through the camera into words. The second is the data
:11:04. > :11:06.set and the data set So with Pinterest Lens,
:11:07. > :11:12.when you point your camera at something in the real world,
:11:13. > :11:15.the computer inside the phone translates that image into text
:11:16. > :11:19.and then it takes the text and it takes the image and runs a search
:11:20. > :11:21.against 100 billion pins on Pinterest to find the ones that
:11:22. > :11:29.seem the most relevant. OK, it knows it's a lemon,
:11:30. > :11:32.that's definitely a good start. OK, I think you scroll through it
:11:33. > :11:36.and some of the results make sense, which is sort
:11:37. > :11:40.of like when you search with words because often you search then
:11:41. > :11:43.and a lot of the things don't make And having come up with those words,
:11:44. > :11:52.I've got a series of recipes So we've got a lemon drizzle cake,
:11:53. > :11:56.a lemon polenta cake. We've now got some
:11:57. > :11:57.artwork of lemons. It's a lot better than it did
:11:58. > :12:01.on my boots and this's probably because this is a very simple image
:12:02. > :12:04.to recognise and understand. Pinterest Lens is also powering
:12:05. > :12:06.Vision, the image search function in Samsung's Bixby
:12:07. > :12:09.which is currently only available And so today we're announcing
:12:10. > :12:14.a new initiative called Google Lens. Google Lens is also
:12:15. > :12:16.due for release soon. The company says it'll be a new way
:12:17. > :12:20.of the computer being able to see and even act on its surroundings
:12:21. > :12:30.whilst you're talking Also working in this space
:12:31. > :12:34.is a chat bot called Glamix, which is a way of photographing any
:12:35. > :12:38.item that you like, sending it to them via Facebook messenger
:12:39. > :12:41.and receiving a response that should tell you where you can
:12:42. > :12:43.buy a similar item. So let's give this a go
:12:44. > :12:47.on my boots to start with. It works with pictures found
:12:48. > :12:54.on Instagram or your phone, eventually allowing you to narrow
:12:55. > :12:57.down results based on price The bot uses artificial
:12:58. > :13:01.intelligence, machine learning and what it calls 'content based
:13:02. > :13:04.image recognition' to search As well as shopping for individual
:13:05. > :13:13.items, it aims to be able to help Making clicking through to items
:13:14. > :13:17.so easy is of course amazing for retailers,
:13:18. > :13:19.but also if you're So if someone passes
:13:20. > :13:23.you by and they're wearing something you really like,
:13:24. > :13:26.you need to be quick. Having spent a while testing both,
:13:27. > :13:34.the results were sometimes surprisingly accurate, and other
:13:35. > :13:36.times, kind of questionable. But it is early days and the more
:13:37. > :13:40.this sort of technology is used, the more data it collects
:13:41. > :13:42.and the more reliable Well, that explains
:13:43. > :13:46.the weird birthday present Now, earlier we looked
:13:47. > :13:56.at human beings attempts to become more robotic,
:13:57. > :13:59.but there's a whole lot of research that's attempting
:14:00. > :14:01.to make robots more human. It's not actually taking place
:14:02. > :14:04.at a robot art school like this, but it's nice to think it might
:14:05. > :14:07.be, isn't it? There is a long way to go
:14:08. > :14:10.in robotics, just picking up all those weirdly shaped everyday
:14:11. > :14:13.objects is still an enormous challenge, requiring a robot
:14:14. > :14:16.to recognise a given object and to decide how
:14:17. > :14:20.exactly to pick it up. But a team at Berkeley says that
:14:21. > :14:23.Dex-Net here is the most effective When not playing with Lego,
:14:24. > :14:27.it's being taught and building up a huge database of 3D objects
:14:28. > :14:30.by its masters. When something new comes along,
:14:31. > :14:33.it uses its 3D sensor to compare it to this list,
:14:34. > :14:36.it then uses its neural network to figure out the best way to grasp
:14:37. > :14:41.it and it is said to get it right The springy legs of this creature
:14:42. > :14:49.were 3D painted at UC San Diego, they're designed to be able to more
:14:50. > :14:52.easily traverse difficult environments, such
:14:53. > :14:56.as disaster areas. As we know, even walking
:14:57. > :14:59.on flat surfaces is still Well, I say ouch, but of course
:15:00. > :15:09.these things don't feel pain. That said, there are those of us
:15:10. > :15:13.who are asking whether even feelings might one day be part
:15:14. > :15:18.of a robot's mind. At the simplest level it makes
:15:19. > :15:23.sense, robots are pretty expensive, you don't want them to run
:15:24. > :15:26.willy-nilly into fire and acid But at a more complex level,
:15:27. > :15:33.we're looking ahead to a time when robots might interact with us
:15:34. > :15:36.on a more personal level as companion robots for the elderly,
:15:37. > :15:39.for those who are sick or are in pain and perhaps maybe
:15:40. > :15:42.they need to understand the similar sort of experience and perhaps
:15:43. > :15:44.develop something like Pain is not just about us
:15:45. > :15:48.saying ouch, there's an emotional element to this
:15:49. > :15:50.as well, isn't there? So are we actually talking about
:15:51. > :15:56.programming some kind of emotions We don't really understand
:15:57. > :16:05.what emotions are in human beings. Like you say, you might assume
:16:06. > :16:08.there's some sort of phenomenon that So hypothetically if we developed
:16:09. > :16:13.systems that worked like pain, might emotion develop off the back
:16:14. > :16:16.of that as well? There are those robots that
:16:17. > :16:18.do look so life-like, the Boston Dynamics' Big Dog
:16:19. > :16:21.and the walking robots, we actually feel quite sorry
:16:22. > :16:24.for them when they fall over or even When those videos were released
:16:25. > :16:29.online the reaction was like - oh no, you're bullying them,
:16:30. > :16:31.don't hurt them. They don't at this stage have
:16:32. > :16:34.that technology at all. There's no suggestion they do feel
:16:35. > :16:37.pain, but the human reaction So is that going to inform
:16:38. > :16:41.how we behave towards Is that where you're looking
:16:42. > :16:45.at applying our sympathies? I mean, I think science fiction
:16:46. > :16:48.model of a human-like entity There may be more kind
:16:49. > :16:53.of cute models we've seen already of robots that,
:16:54. > :16:56.sort of, pull on our heart strings in a more child-like way and there's
:16:57. > :16:59.those that suggest that we shouldn't have anything that looks human-like
:17:00. > :17:02.at all because it's disingenuous, it's cheating and it's tricking us
:17:03. > :17:05.into treating them like they're The doctor thinks that appearing
:17:06. > :17:11.to feel pain may make us treat Of course what many people
:17:12. > :17:18.are worried about is how much respect the robots will have for us
:17:19. > :17:22.and, most of all, our jobs. Last week Caterpillar invested
:17:23. > :17:26.seven million in this Australian Now robots can build houses
:17:27. > :17:34.at the rate of 1,000 bricks an hour. Ambition in the area
:17:35. > :17:36.is huge and for the first time out of the lab,
:17:37. > :17:39.ETH in Switzerland is working on much more ambitious structures
:17:40. > :17:42.like this undulating wall which has Now an increasingly robotic
:17:43. > :17:52.workforce raises a number of issues and along with the worry
:17:53. > :17:55.of what jobs will actually be left to us in the future,
:17:56. > :17:58.there is another one. Fewer workers earning a wage,
:17:59. > :18:02.means fewer workers paying income tax on their earnings and that means
:18:03. > :18:09.less money going into the economy. Now some tech brains,
:18:10. > :18:12.including that of Bill Gates, are calling for a robot tax
:18:13. > :18:15.to counter that and Cat Hawkins went Almost everyone in the world
:18:16. > :18:24.who works pays tax on the money they earn, but at this restaurant
:18:25. > :18:27.in San Francisco there are no waiting staff
:18:28. > :18:32.and robots plate the food. That work is currently not taxable
:18:33. > :18:35.and politician Jane Kim is now looking into how this is changing
:18:36. > :18:40.the city's economy. So what we're seeing
:18:41. > :18:43.is after automation that you can hire less people in order to deliver
:18:44. > :18:46.products maybe quicker But it's one of the questions
:18:47. > :18:52.that we have, it's true this is really convenient,
:18:53. > :18:54.but at what cost? It's not just restaurants, this
:18:55. > :18:59.picture is now seen across the city, from hotels and hospitals
:19:00. > :19:01.to the latest addition to the autonomous family,
:19:02. > :19:05.self-driving cars. Policy makers have noticed, every
:19:06. > :19:08.time a robot take as human job, The research is showing us that jobs
:19:09. > :19:16.are going to get lost over the next ten years and if before
:19:17. > :19:18.the Great Depression we could have predicted
:19:19. > :19:21.what would come afterwards, if government could have prepared
:19:22. > :19:23.for the job loss that occurred, That is the level at which we are
:19:24. > :19:28.looking at potentially over the next ten years,
:19:29. > :19:31.in terms of job loss Estimations of how many jobs will be
:19:32. > :19:37.wiped out vary widely from study to study,
:19:38. > :19:42.but a recent report especially has It's estimated that robots
:19:43. > :19:46.will replace 37% jobs in the United States
:19:47. > :19:49.by the early 2030s. So the biggest concern
:19:50. > :19:52.is mass job displacement, lack of true, meaningful,
:19:53. > :19:55.high wage work. We are already seeing a decrease
:19:56. > :19:59.of that in San Francisco where we have the fastest growing
:20:00. > :20:01.income gap in the country and a wealth gap that is akin
:20:02. > :20:08.to the country of Rwanda, and so we have a shrinking
:20:09. > :20:12.middle-class and we have this growing imminent threat that
:20:13. > :20:16.many of our meaningful, working-class and even
:20:17. > :20:19.middle-class jobs may go away At Cafe X, again a human worker has
:20:20. > :20:25.been replaced by a robot. An Americano with milk,
:20:26. > :20:29.served by a robot. Now, the human has a different role,
:20:30. > :20:33.advising on coffee beans and showing customers how to use the tablet
:20:34. > :20:37.to operate the robot. The owner is not sure about the idea
:20:38. > :20:41.of a tax on the replacement. I guess I find it a little odd
:20:42. > :20:44.because what robots are supposed That means it allows a shift
:20:45. > :20:51.in labour from doing highly repetitive, low productivity tasks
:20:52. > :20:57.to more useful things. So in order to have this machine
:20:58. > :21:06.operate, there has to be a lot of engineers on software,
:21:07. > :21:09.hardware and manufacturing to build Jobs like this require training
:21:10. > :21:12.and that's what Supervisor Kim wants If you're a childcare worker
:21:13. > :21:17.or you're an in home support services worker,
:21:18. > :21:20.working with a senior or individual with disability,
:21:21. > :21:23.you often work three or four hours So one of the ideas was,
:21:24. > :21:30.why not tax robots and invest in these poverty jobs and make them
:21:31. > :21:33.truly living wage This would mean a robot tax
:21:34. > :21:38.potentially subsidising low paying, but essential jobs,
:21:39. > :21:41.so that the human employees Currently, many people are working
:21:42. > :21:46.but not earning enough to live, leading several politicians around
:21:47. > :21:49.the world to float the idea This would be expensive
:21:50. > :21:55.for governments and Supervisor Kim is suggesting an automation tax
:21:56. > :21:59.could be a solution. If there's one thing that
:22:00. > :22:01.San Francisco is known for, it's leading the conversation
:22:02. > :22:06.on technology and innovation, but as harder and harder questions
:22:07. > :22:09.are asked about automation and what this really means
:22:10. > :22:12.for people's jobs it seems appropriate that this city,
:22:13. > :22:14.which has added so much to the problem, is also grappling
:22:15. > :22:21.with what could be the solution. But the rise of robotic workers
:22:22. > :22:25.is playing out on a global scale and San Francisco is not the only
:22:26. > :22:28.place trying to lead In the EU, a proposal to tax robot
:22:29. > :22:36.was voted down earlier in the year and one of the Commissioners who did
:22:37. > :22:40.so says robots will create more They are worried because they say
:22:41. > :22:44.robots they will take their jobs, Progress always created more jobs
:22:45. > :22:50.than progress used to destroy. The train is moving and speed
:22:51. > :22:55.is high and now it's up to us to be on that train or to stay and to wave
:22:56. > :23:05.to the leaving train. Concerns about automation replacing
:23:06. > :23:08.human jobs has been felt sense the Industrial Revolution and more
:23:09. > :23:11.recently workers in the manufacturing industry
:23:12. > :23:13.have seen jobs disappear As the issue of a robot tax
:23:14. > :23:20.begins to spread further, a fundamental question still needs
:23:21. > :23:22.to be answered - In the context of robots of course
:23:23. > :23:29.automation is much broader They gave this definition
:23:30. > :23:36.more than 100 years ago. Politicians can no longer
:23:37. > :23:38.ignore the robots creeping into the workplace and while many
:23:39. > :23:41.of the big questions are still being thrashed out,
:23:42. > :23:44.it's clear that the issue of robot workers is becoming more
:23:45. > :23:53.and more of a political one. Yeah, really interesting
:23:54. > :23:57.issues, aren't they? That was Cat Hawkins
:23:58. > :24:00.and this's it for this week. You can follow us on Twitter
:24:01. > :24:02.@BBC Click throughout the week and like us
:24:03. > :24:05.on Facebook, too. Thanks for watching
:24:06. > :24:33.and we will see you soon. Some decent, dry, and also for some
:24:34. > :24:37.sunny weather around this weekend. But there will be a lot
:24:38. > :24:40.of cloud around at times, threatening some rain,
:24:41. > :24:42.particularly on Saturday. And throughout Saturday,
:24:43. > :24:45.the air gets warmer and more muggy.