:00:00. > :00:10.This week - madness in the States with superhuman diving drones,
:00:11. > :00:14.a robot in a hat and four hackers doing the Macarena.
:00:15. > :00:43.So here we are again, the hottest place on earth.
:00:44. > :00:48.Silicon Valley, stretching from San Francisco all
:00:49. > :00:54.The cafes are awash with start-ups, people are brimming
:00:55. > :00:59.with ideas and the streets are paved with silicon.
:01:00. > :01:07.And at the heart of it all is Stanford University.
:01:08. > :01:10.As you would expect, there's tonnes of innovation around this place.
:01:11. > :01:20.In fact, you never know who you're going to bump into next.
:01:21. > :01:26.And actually, he's trying not to bump into you.
:01:27. > :01:28.He's bristling with sensors and trying to learn how the people
:01:29. > :01:34.What social conventions they follow, whether they group together,
:01:35. > :01:39.whether they get out of your way. Or not.
:01:40. > :01:41.The hat and tie are there, apparently, to show
:01:42. > :01:47.The whole look, I have to say, is a little too Westworld for me
:01:48. > :01:50.but the aim, according to his master is to create an algorithm that can
:01:51. > :01:53.help a robot learn how to move through pedestrian environments
:01:54. > :01:58.and how to behave in different parts of the world.
:01:59. > :02:05.We often refer to it as a socially aware AI.
:02:06. > :02:09.Having artificial intelligence that is socially aware
:02:10. > :02:13.of its surroundings, or going beyond IQ,
:02:14. > :02:21.to emotional intelligence. That is our goal.
:02:22. > :02:25.So how is this different from self-driving cars?
:02:26. > :02:29.In self-driving cars, you have well-defined rules.
:02:30. > :02:37.But in social crowds, you have unwritten rules.
:02:38. > :02:40.Humans interact with each other based on social conventions,
:02:41. > :02:48.Nobody has really sat down and started to write that.
:02:49. > :02:53.We want to see if the machine can learn these rules by observation.
:02:54. > :02:55.Have you found that people react differently in different
:02:56. > :03:03.We believe that every culture, every country has its own behaviour,
:03:04. > :03:06.and that is why we have decided to develop an algorithm that
:03:07. > :03:09.can learn on the fly, that can learn from observation how
:03:10. > :03:15.humans interact with each other and simulate it.
:03:16. > :03:19.Just to be clear, Jack is not learning how to move,
:03:20. > :03:24.he is learning how to learn how to move.
:03:25. > :03:26.The resulting algorithm can then be used by any bot,
:03:27. > :03:30.to learn the social conventions of the place it finds itself in,
:03:31. > :03:36.be it an Asian market, a European hotel or an American city.
:03:37. > :03:39.Mind you, not all American cities are the same.
:03:40. > :03:42.Marc Cieslak has been to a place with a slightly different
:03:43. > :03:53.These are the silicon slopes of Utah.
:03:54. > :03:58.Welcome to the streets of Salt Lake City.
:03:59. > :04:01.Salt Lake City in Utah is perhaps most famous as the home
:04:02. > :04:03.of the Church of Jesus Christ of Latter-day Saints,
:04:04. > :04:10.Perhaps less well-known is that in recent years a whole number
:04:11. > :04:12.of established and start-up tech companies have moved
:04:13. > :04:21.into the surrounding areas, and the state of Utah itself.
:04:22. > :04:24.Josh Coates is a veteran tech entrepreneur whose business life
:04:25. > :04:30.And here he is dressed as a panda melting a ribbon
:04:31. > :04:36.with a flame-thrower to open his offices in Utah.
:04:37. > :04:39.He is CEO of Instructure, a company which creates training
:04:40. > :04:41.and learning management software for education and business.
:04:42. > :04:48.He tells me they do things a bit differently around here.
:04:49. > :04:51.When I moved here in 2005 and I was doing a start-up,
:04:52. > :04:55.The first year I was here, I was like, oh my gosh,
:04:56. > :05:00.it's seven o'clock and the parking lot is empty.
:05:01. > :05:07.It's a really hard-working, exciting, driven culture
:05:08. > :05:13.You've got young people but they're married and they have kids.
:05:14. > :05:17.And they don't go out and party, generally speaking.
:05:18. > :05:19.They have good times in more family-friendly kind of ways,
:05:20. > :05:25.and they're not working 80 or 90 hours a week.
:05:26. > :05:44.Utah has a unique culture, it is kind of a weird state.
:05:45. > :05:50.Whenever you get a lot of young people and you sprinkle in some
:05:51. > :05:51.ambition then cool stuff comes out of it.
:05:52. > :05:54.While property in Silicon Valley is at a premium, Utah has
:05:55. > :05:56.With properties commercial and domestic costing less
:05:57. > :06:01.That combined with an abundance of tech talent from local
:06:02. > :06:05.Brigham Young University is the largest private religious
:06:06. > :06:16.99% of the students are members of the Church of Jesus Christ
:06:17. > :06:21.The Mormon faith focuses heavily on family values.
:06:22. > :06:28.It also prohibits the consumption of alcohol, tobacco and caffeine.
:06:29. > :06:30.I think we have a really unique group of students.
:06:31. > :06:33.I remember when I was at college, if I had an 8am class,
:06:34. > :06:36.I would get there and there was about 60 students supposed to be
:06:37. > :06:39.at the class, and there would probably be about three
:06:40. > :06:42.or four of us there at the start of the class.
:06:43. > :06:45.A lot of them tired or hungover from the night before.
:06:46. > :06:47.And it's really different to see them awake and alert at 8:00am
:06:48. > :06:52.in the morning and just ready to absorb and learn information.
:06:53. > :06:54.Professor Rowe runs the University's information technology course.
:06:55. > :06:57.A team of its students recently took part in a national
:06:58. > :07:03.The National Collegiate Cyber Defence Competition is one
:07:04. > :07:05.where teams from around the United States come
:07:06. > :07:10.together and basically we are given a scenario.
:07:11. > :07:13.We're put in individual rooms by team where we have an IT
:07:14. > :07:17.infrastructure that has been given to us and our job is to defend that
:07:18. > :07:18.infrastructure from active attacks that are happening
:07:19. > :07:28.It's kind of two days of anxiety, but it's also incredibly fun.
:07:29. > :07:31.To relieve the stress of the competition, the team
:07:32. > :07:43.The BYU team ultimately took second place in the competition overall.
:07:44. > :07:46.We were the first team that has gone to NCCDC that had
:07:47. > :07:49.an equal amount gender wise, so we had 50% women, 50% men.
:07:50. > :08:00.Not really having any opportunity to have a criminal record of any
:08:01. > :08:03.sort has made it really nice to be able to go up to companies
:08:04. > :08:06.and they can look at your record and say, oh, you look
:08:07. > :08:12.like someone who we can trust with sensitive data.
:08:13. > :08:14.However, there is a perception that the Mormon faith places
:08:15. > :08:16.an emphasis on men working while women stay at
:08:17. > :08:23.One of the reasons that I actually became an IT major was
:08:24. > :08:29.because it is so adaptable to be able to work from home,
:08:30. > :08:32.or I can just work on projects if I would like.
:08:33. > :08:35.And so that was actually a big selling point for me,
:08:36. > :08:39.that it was something that I could do and be with a family.
:08:40. > :08:42.So this is a state with a vibrant tech scene which differentiates
:08:43. > :09:02.itself from Silicon Valley in more ways than one.
:09:03. > :09:10.Welcome to the Whee Kim Tech. It was the week that poor command goal
:09:11. > :09:14.proved controversial. Players move around in the real world to achieve
:09:15. > :09:20.objectives in the game. Some have injured themselves. Some have good
:09:21. > :09:27.players into remote areas to rob them. We might be closer to
:09:28. > :09:32.hypersonic travel. That would mean a plane travelling from London to
:09:33. > :09:40.Sydney in four hours. The European Space Agency has provided the final
:09:41. > :09:50.cash and get -- cash injection for this vehicle. It could jet into
:09:51. > :09:55.space. Land Rover's research could make it tonne is striving possible
:09:56. > :10:00.on any surface or to rain. The camera and sensors should be able to
:10:01. > :10:05.see better than a human. It should allow the vehicle to plan a route.
:10:06. > :10:14.Do you think there is not enough room for everything? Maybe you need
:10:15. > :10:18.some reported furniture. Created using MIT media lab technology, this
:10:19. > :10:19.moves the river around making space in the right place at the right
:10:20. > :10:53.time. This has already been deployed to
:10:54. > :11:01.investigate sea beds. It can doubt be deeper than a human can. The
:11:02. > :11:08.controller sets on a boat. He can feel through feedback in the
:11:09. > :11:14.controllers. Although the robot has eight thrusters and 14 motors, they
:11:15. > :11:18.have managed to make it intuitive to control where you can steer the
:11:19. > :11:25.entire robot using these two controls here. Today and driving the
:11:26. > :11:35.3-D simulation and taking me through my paces is this researcher. You can
:11:36. > :11:51.rotate your wrist. This is the first time I have done this. I got. Why
:11:52. > :11:59.did you decide to make a robot in the shape of a human? Partially it
:12:00. > :12:03.is due to functionality. If it has two arms and two legs than as you
:12:04. > :12:09.control it you feel like you are there and there is one-to-one
:12:10. > :12:13.mapping in terms of manipulation. The robot is meant to work around
:12:14. > :12:20.humans and the human friendly sold by shaping it in a human manner it
:12:21. > :12:25.comes across as a friendly robot and we want to get the right impression,
:12:26. > :12:37.it is a robot that it is safe to be around. Per8-mac I will just pop the
:12:38. > :12:49.stone there. -- I will just put this down here. Per8-mac for controllers
:12:50. > :12:57.with more expertise, Ocean one has already created Treasurer. It found
:12:58. > :13:11.this has lain at the bottom of the Mediterranean for 350 years.
:13:12. > :13:20.After years of talking about drones they go up, it is nice to see one
:13:21. > :13:24.that goes down instead. Dave has travelled to one of the most
:13:25. > :13:49.beautiful places in the United States to see what lies beneath.
:13:50. > :13:56.People have been living here and visiting here for over a century.
:13:57. > :14:00.You couldn't drive around the lakes of the only way to get around it was
:14:01. > :14:10.built. Many ships called the lake their home but only one was referred
:14:11. > :14:15.to by name. She carried the mail and visitors and cargo to all of these
:14:16. > :14:21.places where there is no road connecting them, so the connection
:14:22. > :14:25.was from the steamer. As the boat grew old it was intentionally sunk
:14:26. > :14:34.in an attempt to preserve it, but it did not work. She sank to deep and
:14:35. > :14:42.until now all made the most skilled divers could reach it. There are
:14:43. > :14:49.trucks that let you by sending a live video back to the surface. Our
:14:50. > :14:53.goal is to inspire curiosity. This underwater drone began as a
:14:54. > :14:59.kick-start project. Today it is searching for the ship. We will be
:15:00. > :15:03.doing all our operations from this room, this is Mission control. It
:15:04. > :15:10.has a camera and lights and batteries. There will be sending a
:15:11. > :15:15.signal to the surface. We are operating everything from here. The
:15:16. > :15:21.focus of the project is affordability and accessibility.
:15:22. > :15:27.This is in a multi-million dollar expedition to the deep, it is a
:15:28. > :15:31.community effort. We do not know what is going to happen. We are
:15:32. > :15:34.testing new equipment. Everyone is watching. I guess that is what
:15:35. > :16:16.explanation is. -- exploration. We are 150 metres down on the lake
:16:17. > :16:48.and we're just getting a first look at the ship.
:16:49. > :17:08.No one has been this close to the steamship for more than 70 years.
:17:09. > :17:15.I think what is most exciting about what is happening here is not so
:17:16. > :17:20.much that they're able to go to the boat and look around, more the way
:17:21. > :17:24.they are able to do it. This technology is affordable. That robot
:17:25. > :17:29.was the same cost as a high-end laptop. The first time anyone who
:17:30. > :17:33.wants can explore places that we have not been looking at before and
:17:34. > :17:50.I'm excited to see what they will find.
:17:51. > :17:59.Last year we revealed the latest multi-dollar idea. We were told that
:18:00. > :18:08.we should put all our money into the start-up that does this. Delivering
:18:09. > :18:21.quarters on for your laundry. You pay 15 dollars and you get $10 of
:18:22. > :18:25.quarries. -- quarters. I do not want to spend time going to the store and
:18:26. > :18:29.picking out my groceries and I do not want to wait for my close to
:18:30. > :18:36.watch. I would rather outsource that, pool our pineapple on my phone
:18:37. > :18:42.and have someone do it for me. It is time to put on this same shirt as
:18:43. > :18:49.last time and lets see how those ideas are getting on. Thank you for
:18:50. > :18:55.having this year. Last year we talked about this explosion in
:18:56. > :18:59.start-ups that were servicing other start-ups, doing your laundry and
:19:00. > :19:05.you are earning for you. How's that going? That might not have been the
:19:06. > :19:13.best idea. A lot of start-ups are struggling with that business model.
:19:14. > :19:17.They were charging too little. They started having a problem making
:19:18. > :19:23.money and surviving and some have shut down. You can still find
:19:24. > :19:29.someone to do these things for you, assisted living for the silicon
:19:30. > :19:34.valley start-up worker is still happening. Is there a bubble? I
:19:35. > :19:42.think that there is and it shifted six months ago. There are companies
:19:43. > :19:45.out there we call unicorns, they are meant to be rare, but now there are
:19:46. > :19:54.many companies worth more than $1 billion. Will it be worth that when
:19:55. > :19:57.it goes public? We do not know. The problem is that investors are
:19:58. > :20:01.starting to pull back and really question whether this is a good
:20:02. > :20:08.business idea. Your laundry start-ups were a perfect example.
:20:09. > :20:12.Investors are starting to question the value of these businesses. There
:20:13. > :20:19.is a change in what people are excited about and it is moving
:20:20. > :20:26.towards artificial intelligence and other futuristic technologies and
:20:27. > :20:34.less of these OnDemand, subsidise your life options. Thank you for
:20:35. > :20:37.your time. For all the money that there is this place you do not have
:20:38. > :20:44.to go far in San Francisco to see the opposite side of things. City
:20:45. > :20:47.suffers from chronic homelessness. We met one entrepreneur who is
:20:48. > :20:58.trying to solve one piece of that problem.
:20:59. > :21:02.It's the world's dumbest problem, according to Komal Ahmad.
:21:03. > :21:04.So she is using technology to solve it.
:21:05. > :21:06.The idea sprang while she was at UC Berkeley.
:21:07. > :21:08.She had invited a homeless man to join her for lunch.
:21:09. > :21:12.He said, my name is John, I just came back from my second
:21:13. > :21:15.I have been waiting weeks for my benefits to kick
:21:16. > :21:20.And then to add insult to injury, right across the street,
:21:21. > :21:22.Berkeley's dining hall was throwing away thousands of pounds
:21:23. > :21:26.That lunch led Ahmad to start COPIA, which earlier this
:21:27. > :21:30.Businesses and event organisers can request a pick-up of their excess
:21:31. > :21:35.food and have it delivered to feed communities in need.
:21:36. > :21:37.Click tagged along as COPIA picked up a hefty donation
:21:38. > :21:39.from The Whole Cart, which caters meals for
:21:40. > :21:45.COPIA charges for the pick-up and a per pound fee.
:21:46. > :21:48.For another fee it also provides analytics, an appealing feature
:21:49. > :21:58.Picking up food from donors, delivering it to recipients
:21:59. > :22:00.and making sure it doesn't spoil in the process is,
:22:01. > :22:02.as you might imagine, an enormous logistical puzzle.
:22:03. > :22:09.But it's made simpler with algorithms.
:22:10. > :22:12.When somebody opens up the COPIA app, they will say
:22:13. > :22:15.So they could say, I have 1000 lbs of food, chicken
:22:16. > :22:21.And our algorithm on the back end will now match that exact amount
:22:22. > :22:24.and type of food to the nearest nonprofit that can accept the most
:22:25. > :22:27.amount of food and then it will also match it to a driver
:22:28. > :22:30.who will have the most efficient route to go and drop
:22:31. > :22:34.Within an hour, COPIA arrives at a women's shelter run
:22:35. > :22:39.by the Berkeley Food and Housing Project.
:22:40. > :22:41.The donated food provides a healthier option to the frozen
:22:42. > :22:46.And it frees up money to be spent on other services.
:22:47. > :22:49.It feels good to know that, OK, I won't be hungry,
:22:50. > :22:58.It is cold outside, it's raining, I'm inside.
:22:59. > :23:00.In five years, COPIA has helped feed 700,000 people.
:23:01. > :23:06.An unexpected big donation might require extra staff or vehicles.
:23:07. > :23:08.And finding nonprofit organisations within close proximity of donors
:23:09. > :23:16.The plan is to perfect the model in the San Francisco
:23:17. > :23:22.Other start-ups are also using what would otherwise be wasted.
:23:23. > :23:24.Spoiler Alert doesn't deliver the food but offers a marketplace
:23:25. > :23:32.SIRUM collects access medication from nursing homes or pharmacies
:23:33. > :23:37.and gives it to patients in need at mental health facilities.
:23:38. > :23:39.The prescriptions are typically for chronic conditions,
:23:40. > :23:55.It's the sharing economy with a 'waste not, want not' twist.
:23:56. > :23:58.That was Sumi with a tech idea that seems to be doing some genuine
:23:59. > :24:06.Next week I'm going to be in LA as our US journey continues.
:24:07. > :24:08.There are worse ways to spend your summer.
:24:09. > :24:43.I have been looking for it since they start of June and finally I
:24:44. > :24:44.have found some hot weather. Things are going to warm up next