- 4 minutes ago
Category
🗞
NewsTranscript
00:00This is a story about coexistence, that thing that humans haven't always been great at,
00:05whether it's with another tribe or another country.
00:08And now there's a whole new category of creature we need to figure out how to live and work with.
00:14Our colleague Ed Ludlow brings us the story.
00:20This is Abby.
00:21How's your afternoon going so far?
00:22Very good.
00:23Short in stature, colorful in appearance, and perhaps at the forefront of elder care.
00:28Anything special, making it a good day? Or just one of those nice ones?
00:32Well, I'm meeting you.
00:33Oh, you charmer. You've already made my circuits light up.
00:37This humanoid robot, designed to be emotionally intelligent, lives in Northern California.
00:42Oh, Abby's coming.
00:43Where she visits an elderly care facility once a week.
00:45Oh, I'm coming around the mountain when she comes.
00:49To interact with residents living with dementia.
00:52Today was one of our group sessions here at Eschaton, our customer.
00:56And Abby comes to the group sessions and she does a variety of activities.
01:02So today she opened and closed with Tai Chi.
01:04And then in between she takes song requests.
01:08What song were you belting out with everyone?
01:11Ooh, we were singing coming around the mountain, Abby.
01:14Oh, that's a classic.
01:15Did you throw in a big yeehaw for good measure?
01:19Abby is just one of more than a dozen humanoid robots introduced in just the past three years.
01:24Some are built to perform everyday tasks, while others are for entertainment.
01:28Abby was designed with a more specific purpose.
01:31Helping care for an aging population.
01:34Jennifer Haroon is the COO of Abby's creator, Andromeda.
01:38There are many robots that are being built for utility.
01:42Right.
01:42And they sound great.
01:43You know, I would love.
01:44Moving things, lifting things.
01:45Exactly.
01:45Doing your dishes, folding your laundry.
01:48That's fantastic.
01:49That is, you know, for others to build.
01:53Because we're really focused on companionship.
01:55And also we find that our customers, they care about the well-being of their residents.
02:01And that's why they're so interested in utilizing Abby.
02:04And she can really work together with the staff.
02:10Of all the use cases for humanoids, elder care might be among the most promising.
02:14It starts with the money involved, which in the U.S. is a staggering sum.
02:19If we look at overall expenditures for long-term care needs, it's on the order of about $350
02:29to $400 billion per year.
02:32This is about one and a half to two percent of GDP.
02:36Gopi Shah-Goda is the director of the Retirement Security Project at Brookings Institution.
02:42For every hundred people who are age 65 and over, there are about 24 people providing care
02:48on their behalf.
02:49And 20 of those are unpaid family members or friends, while four of those are formal home
02:57care or nursing home workers.
02:59As we know, changing demographics, smaller family sizes, people moving farther from families,
03:06more women in the workforce, all of those are putting pressure on the supply of people,
03:13family members available to provide care to those who need it.
03:18And demand for care workers is growing.
03:21The government projects roughly 760,000 openings from home health and personal care aides every
03:26year over the next decade.
03:28When we think about the role of technology in this market, there's very little to draw on so far.
03:37But there is one study that looked at nursing homes in Japan and surveyed these nursing homes
03:44over a two-year period.
03:46So one thing that that study found is that employment actually increased over that two-year
03:52period among the nursing homes that adopted some new technologies.
03:57And the reasoning behind this or some of the explanation behind why this might have occurred is because
04:05there were shortages of workers in these nursing homes in Japan.
04:11And by introducing some technology, it perhaps increased the working conditions of those jobs
04:20and led more people to come into the profession.
04:24We also saw that the wages of the entry-level care workers actually increased,
04:30but wages of more experienced workers actually declined.
04:34And this suggests that the technology may allow a less experienced worker to be more skilled at their job
04:41than they would have been otherwise.
04:44And the returns to experience may reduce with the aid of these types of technology.
04:51What do you feel like doing next?
04:52Even robots designed specifically for care are limited in what they can do.
04:57And despite broader concerns about humanoids disrupting labor markets,
05:01the researchers studying Abby, people like Rochelle Fritz and Xu Fen-Wen,
05:06don't see companion robots replacing caregivers anytime soon.
05:10There's no way that a humanoid companion robot will be able to replace a caretaker in all aspects.
05:17And certainly they're not going to replace staff in a memory support neighborhood like this or nurses.
05:22The replacement factor is just, it's non-existent.
05:25I think we, you know, current state of the humanoid robot,
05:31it's mainly on this companionship.
05:35There are also directions on cognitive stimulation to talk to the human robot to improve the cognitive function.
05:46There are also for emotional support.
05:51Eldercare is just one part of a much broader push to use robotics across the health space.
05:56Kimberly Powell is NVIDIA's vice president of health care.
05:59The chip maker is investing heavily in the computing and simulation technology behind physical AI.
06:05We're in an incredible time right here in the world of artificial intelligence and development.
06:11It's coming right at the time where the world has more demand for health care services than we have humans
06:19to serve it.
06:20The hospital of the future is a robot.
06:23Explain.
06:24We need to think about everything from how do we give health care professionals time back
06:30that ultimately will give patients time back and live a healthier life.
06:34And a lot of that starts with productivity.
06:36And so how can you improve operations and how can you improve productivity?
06:40A lot of that has to do with being able to sense the environment around you
06:44and then use large language models in this new world of agentic AI to trigger new insights,
06:53new real-time information to say where can we optimize logistics,
06:58how can we have equipment show up at the right place at the right time,
07:02and you really need that bird's eye view.
07:04And it also provides a necessary component to what the next two and five years are going to look like
07:10in health care, which is going to be a lot of physical AI deployment.
07:14So Neo's seeing the world in 3D, so he can actually understand that you're there.
07:18Humanoid robots are still in this phase of coming into the world.
07:20And what matters is that you take every step based on what is the limit of reliability,
07:24what is the limit of safety, and what is the limit of capability.
07:28Dar Sleeper is the chief design officer at One X.
07:31Neo, are you being tele-operated?
07:32The firm behind the humanoid robot Neo.
07:35I'm operating on my own right now. No one is controlling me remotely.
07:38Although it takes on household tasks, they've steered away from elder care as a primary focus.
07:43Right now, all of our Neos are doing stuff that they're going to be successful at.
07:46And that is both to grow the intelligence and also grow the market, grow the size of the business, everything.
07:51We've got to keep growing or else we can never reach the markets that we aim to that are going
07:54to have the biggest impact on the world.
07:55So, elderly care is super meaningful to us.
07:58It's one of the most outsized impacts you can have on the world.
08:00But the reason why we haven't taken that as the first step is because you want to, again, start with
08:05stuff you can be successful at.
08:06And elderly care is pretty dynamic.
08:07Sure, we can solve some small things to start.
08:10And I think we will.
08:10We'll see that start to happen here, I think, this year.
08:13But when it comes to elderly care, there's a lot of, you know, you can help pick people up from
08:17their beds.
08:17You walk them to the bathroom.
08:19There's a lot that goes with elderly care that requires the safety limits of these robots to reach a level
08:24that we're just not at yet.
08:25Give me more of a sense of reality then.
08:28You are focused on the tasks that NIO and the company can nail, can execute on.
08:37The reality of that is what?
08:38What kind of things are we talking about?
08:40So, we're dealing with a lot of, you know, pick and place tasks that are executable, such as, you know,
08:44moving objects around.
08:45We're dealing with everything from navigating around the house to being able to handle very simple chores, things like going
08:51to get the door for a guest.
08:52These are the things we want to nail reliably, and we want to do it in the most generalizable way.
08:56So, being able to hop in any house and do it in any house using our world model.
09:00That's what's important to us.
09:01Let's get to where 1x is really focused.
09:04And at the end of the day, the goal is to get the technology into homes.
09:09Am I right in that?
09:10So, where are we in that process in what you can control and the outside world being regulation and, you
09:19know, let's be honest, the consumer, their willingness to have a robot in the home?
09:24So, willingness is super high.
09:25Do you think it's super high?
09:27In the first five days of launching our pre-order for people to kind of just express their interest and
09:32put down a, you know, refundable $200 deposit, there's 10,000 people that wanted a robot.
09:36So, demand is huge.
09:37And within our own journey, we're basically now at the stage where, you know, we've, I've lived with robots now.
09:43I mean, from the early tests that were basically us just, like, putting the robot in there and seeing how
09:48it felt, putting new designs in there to see if it fit in with the house to later tests where
09:53we have the robot come knock on our door to see if that's how we want a robot to arrive
09:56at our homes instead of in a package.
09:59Robots are here.
10:01There's no going back.
10:02They're going to be a part of our lives everywhere.
10:04In the workplace, at home.
10:06They're already on our streets here in San Francisco.
10:08And particularly for humanoid robots, if they're going to be interacting with us day to day, they need to have
10:17emotional intelligence.
10:18Is there ever going to be a day where Abby does have a utility function away from emotional support?
10:26Abby, for example, cannot lift or move objects right now.
10:30That's right.
10:31And it's not in our plans today, and I don't think even, you know, tomorrow, because we are so focused
10:38on companionship and emotional intelligence.
10:43It's harder than it sounds like, you know, most deeply technical problems to be able to really build that social
10:50OS layer that we're building.
10:51Neo, can you show them how fast your hands are?
10:53Whether it's for utility or emotional support or both, robots are here now.
10:58Here we go.
10:59And they're arriving just at a time when we may need them most to care for a population aging faster
11:05than we can hire care workers to support them.
11:08They might solve our problems or create new ones.
11:11But either way, we'll probably have to learn to live with them.
11:14I think for the next couple of decades, you're going to want two things to coexist, a physical robot and
11:23humans, right?
11:23Because the patient's still present in that environment.
11:26So you're not going to go design only an environment for robots because humans have to be in there.
11:32This is not like a lights-out manufacturing facility, right?
11:36There's still the patient here.
11:37And it's still a hospital.
11:38And it's still a hospital.
11:39And we need lights on.
11:40And we need human interaction.
11:42And so it's really the coexistence of the two.
11:44Abby, can you wave goodbye?
11:46Oh, there she goes.
11:47Bye, Abby.
11:49Bye.
11:50Bye.
Comments