WEBVTT - Mobileye CEO Amnon Shashua Talks Humanoid Robots

0:00:02.520 --> 0:00:07.320
<v Speaker 1>Bloomberg Audio Studios, podcasts, radio news.

0:00:07.880 --> 0:00:11.600
<v Speaker 2>Mobili yesterday had a pretty key announcement a new deal,

0:00:12.119 --> 0:00:14.360
<v Speaker 2>a deal to the tune of nine hundred million dollars

0:00:14.440 --> 0:00:18.680
<v Speaker 2>combining cash MOBILEZ shares to buy Menty Robotics and the

0:00:18.680 --> 0:00:20.520
<v Speaker 2>dealers expected to be closing their first quarter of twenty

0:00:20.520 --> 0:00:22.439
<v Speaker 2>twenty six. We're still in gains up three times percent,

0:00:22.480 --> 0:00:24.120
<v Speaker 2>where significantly off.

0:00:24.000 --> 0:00:24.760
<v Speaker 3>Of our hives.

0:00:24.880 --> 0:00:28.000
<v Speaker 1>As the market digests what this deal really means. We've

0:00:28.000 --> 0:00:30.000
<v Speaker 1>got the perfect person. We're asking what it means. I'm

0:00:30.040 --> 0:00:33.080
<v Speaker 1>non Shashow has here with us Mobile CEO. Is wonderful

0:00:33.120 --> 0:00:34.080
<v Speaker 1>time and time with you, Sam.

0:00:33.960 --> 0:00:35.479
<v Speaker 3>Wonderful to be with your Carolinaded.

0:00:35.800 --> 0:00:39.360
<v Speaker 1>Why is it right for you to go from autonomous

0:00:39.440 --> 0:00:41.680
<v Speaker 1>driving into the world of robotics.

0:00:41.440 --> 0:00:45.880
<v Speaker 4>Well mobilizing AI company in the field of autonomous driving

0:00:46.159 --> 0:00:50.279
<v Speaker 4>driving assist the full spectrum of a computer vision and

0:00:51.240 --> 0:00:56.520
<v Speaker 4>AI to control cars and human ideopolitics has been recognized

0:00:56.520 --> 0:00:59.080
<v Speaker 4>in the past two or three years as a complementary

0:01:00.920 --> 0:01:03.960
<v Speaker 4>and Mobili wants to It's not only another growth engine,

0:01:04.520 --> 0:01:07.400
<v Speaker 4>but from a technological point of view, if you are

0:01:07.520 --> 0:01:10.400
<v Speaker 4>an actor in this area of physical AI, you want

0:01:10.440 --> 0:01:14.160
<v Speaker 4>to then extend it to the full scope of physical AI,

0:01:14.720 --> 0:01:18.040
<v Speaker 4>and I believe that robotics, humanoids, they have a great future.

0:01:18.600 --> 0:01:20.640
<v Speaker 4>Now maybe it's a long term plate that it's not

0:01:20.840 --> 0:01:23.000
<v Speaker 4>you know, a year from now, but I believe it

0:01:23.080 --> 0:01:30.200
<v Speaker 4>has a significant, exciting future. And from Mobili there's synergetic

0:01:30.520 --> 0:01:35.640
<v Speaker 4>areas in infrastructure, in AI, in software, in AI talents.

0:01:35.760 --> 0:01:37.640
<v Speaker 4>So from Mobiliz this is a great move.

0:01:37.959 --> 0:01:40.520
<v Speaker 1>I mean, of course, in many ways, people say you

0:01:40.800 --> 0:01:45.280
<v Speaker 1>think that because you helped found menty your son's work

0:01:45.800 --> 0:01:50.480
<v Speaker 1>at MENTI help me get comfortable with that because many

0:01:50.520 --> 0:01:54.480
<v Speaker 1>have thrown such accusations. Elon must sway sometimes previous acquisitions and.

0:01:54.440 --> 0:01:55.000
<v Speaker 3>Not liked it.

0:01:55.280 --> 0:01:58.160
<v Speaker 4>Well, it's called the related party transaction. Has been done

0:01:58.200 --> 0:01:59.880
<v Speaker 4>in the past and will be done in the future,

0:02:00.680 --> 0:02:04.160
<v Speaker 4>and there are ways to handle it, like being recused from.

0:02:04.040 --> 0:02:06.040
<v Speaker 3>Decision, which I did and so forth.

0:02:06.080 --> 0:02:09.040
<v Speaker 4>And you know, my son has a zero point zero

0:02:09.080 --> 0:02:12.120
<v Speaker 4>zero zero zero one percent is just an employee, so

0:02:12.160 --> 0:02:15.720
<v Speaker 4>it's not a big deal. You know, he graduated from

0:02:15.760 --> 0:02:18.560
<v Speaker 4>computer science and he wanted to do some work and

0:02:20.040 --> 0:02:22.360
<v Speaker 4>go work for a company that is very, very strong

0:02:22.400 --> 0:02:25.720
<v Speaker 4>in AI, working on the future of physical AI. So

0:02:25.919 --> 0:02:29.200
<v Speaker 4>it makes sense. But you know that is really immaterial.

0:02:29.639 --> 0:02:34.720
<v Speaker 4>The material part is the scope of AI. You know,

0:02:34.840 --> 0:02:38.840
<v Speaker 4>physical AI includes humanoids, includes autonomous driving, and it makes

0:02:38.880 --> 0:02:42.360
<v Speaker 4>a lot a lot of sense. Mobili over the years

0:02:42.440 --> 0:02:45.680
<v Speaker 4>was always looking for a new growth engine, right for example,

0:02:45.720 --> 0:02:49.560
<v Speaker 4>take computer vision and applied maybe to security areas in

0:02:49.639 --> 0:02:51.880
<v Speaker 4>cameras and so forth. But we never found something that

0:02:52.000 --> 0:02:55.560
<v Speaker 4>really clicked, found something that has significant modes, that has

0:02:55.560 --> 0:03:00.840
<v Speaker 4>a significant term, and humanoids is becoming that click is

0:03:00.919 --> 0:03:01.680
<v Speaker 4>becoming this.

0:03:01.720 --> 0:03:03.200
<v Speaker 3>Area where are you coming and grow?

0:03:03.280 --> 0:03:06.120
<v Speaker 5>You know, I've been talking for years about you know,

0:03:06.200 --> 0:03:09.160
<v Speaker 5>mobilized existing footprint and then it's go to market in

0:03:09.200 --> 0:03:11.600
<v Speaker 5>the future. What's the go to market for mentee?

0:03:12.160 --> 0:03:14.119
<v Speaker 4>Well, we see to go to market in two phases.

0:03:14.600 --> 0:03:22.040
<v Speaker 4>The first phase is structured environments like warehouses, assembly plants, retail,

0:03:23.160 --> 0:03:27.480
<v Speaker 4>where the number of tasks are finite are also understood

0:03:27.480 --> 0:03:30.440
<v Speaker 4>in advance. What are the tasks that the customer is interested?

0:03:30.639 --> 0:03:33.240
<v Speaker 4>A customer will buy a fleet of robots, not just

0:03:33.480 --> 0:03:38.200
<v Speaker 4>one robot, and you can customize per customer that I

0:03:38.240 --> 0:03:42.400
<v Speaker 4>think based on pocs that the mentee has been doing.

0:03:42.800 --> 0:03:44.720
<v Speaker 4>If you look at the demonstrations that have been showing

0:03:44.760 --> 0:03:47.760
<v Speaker 4>like robots picking boxes, moving them from place to place.

0:03:47.840 --> 0:03:49.760
<v Speaker 5>You would essentially just sell the hardware.

0:03:49.520 --> 0:03:51.880
<v Speaker 3>A seller robot either sell or Lisa robot.

0:03:52.200 --> 0:03:52.320
<v Speaker 5>Right.

0:03:54.000 --> 0:03:57.680
<v Speaker 4>We believe that the manufacturing costs in a reasonable volume

0:03:57.720 --> 0:04:00.000
<v Speaker 4>of tens of thousands of robots would be twenty thousand dollars.

0:04:00.160 --> 0:04:03.960
<v Speaker 4>So this gives a lot of flexibility in a business model.

0:04:04.000 --> 0:04:06.080
<v Speaker 4>So this is phase one. We believe it will start

0:04:06.080 --> 0:04:10.360
<v Speaker 4>in twenty twenty eight. Phase two is unstructured environments like

0:04:10.400 --> 0:04:14.480
<v Speaker 4>home use. That is much more complicated because the tests

0:04:14.520 --> 0:04:15.360
<v Speaker 4>are open ended.

0:04:16.440 --> 0:04:18.360
<v Speaker 3>You need to continuously learn, so.

0:04:18.320 --> 0:04:20.640
<v Speaker 4>It's not that you can prepare yourself to all the

0:04:20.680 --> 0:04:24.839
<v Speaker 4>possible tasks a home in a home use, so you

0:04:24.920 --> 0:04:27.480
<v Speaker 4>need new technology. So it's not that the same technology

0:04:27.520 --> 0:04:29.839
<v Speaker 4>you use for structured environments. And this is one of

0:04:29.839 --> 0:04:32.720
<v Speaker 4>the things that I showed in my keynote, a technology

0:04:32.839 --> 0:04:34.960
<v Speaker 4>that Minty, and this is why Minty is called Minty

0:04:35.200 --> 0:04:39.240
<v Speaker 4>from mentory. The robot is watching passively watching a human

0:04:39.360 --> 0:04:43.200
<v Speaker 4>showing a new task, and the video goes up to

0:04:43.240 --> 0:04:45.359
<v Speaker 4>the cloud. In the cloud, there's a foundation model that

0:04:45.400 --> 0:04:49.000
<v Speaker 4>takes the video, moves it into a simulator, trains over

0:04:49.040 --> 0:04:51.960
<v Speaker 4>the simulator. Today it's a few hours in the future,

0:04:51.960 --> 0:04:53.120
<v Speaker 4>a few minutes goes back to.

0:04:53.040 --> 0:04:54.640
<v Speaker 3>The robot and the robots perform the test.

0:04:54.839 --> 0:04:56.840
<v Speaker 5>Let's go back to your core business and on this

0:04:56.920 --> 0:05:01.000
<v Speaker 5>week in video showed us a full stack solution autonomous driving.

0:05:01.680 --> 0:05:05.200
<v Speaker 5>You know, they make arguments that their software and hardware

0:05:05.200 --> 0:05:09.200
<v Speaker 5>competencies will make them win. Your evaluation of what they're

0:05:09.240 --> 0:05:11.920
<v Speaker 5>offering against what you plan to do costs for mile

0:05:12.000 --> 0:05:13.919
<v Speaker 5>basis how worried you are.

0:05:14.600 --> 0:05:15.239
<v Speaker 3>I'm not worried.

0:05:15.279 --> 0:05:20.040
<v Speaker 4>Competition, I believe is always good. It kind of ignites

0:05:20.120 --> 0:05:24.159
<v Speaker 4>the market. In my keynote, I had Christian Zenger, the

0:05:24.200 --> 0:05:29.320
<v Speaker 4>CEO of Folkswagen Autonomously Cars, chairman of Moya, come to

0:05:29.360 --> 0:05:31.840
<v Speaker 4>the stage and talk about one hundred thousand vehicles in

0:05:31.839 --> 0:05:35.359
<v Speaker 4>the next eight years, one hundred thousand robotaxes in the

0:05:35.360 --> 0:05:39.480
<v Speaker 4>next eight years. Talked about that we are almost there

0:05:39.520 --> 0:05:42.560
<v Speaker 4>in removing the driver, going driver less in Q three.

0:05:42.960 --> 0:05:44.920
<v Speaker 5>Yeah, but are they using mobili Yeah, of course, this

0:05:45.040 --> 0:05:48.279
<v Speaker 5>is clad, iron clad, This is a mobilize.

0:05:48.279 --> 0:05:51.680
<v Speaker 4>There are one hundred id bus vehicles in multiple locales.

0:05:52.160 --> 0:05:56.800
<v Speaker 3>All of them is mobil mobilized tack and we.

0:05:56.760 --> 0:06:00.159
<v Speaker 4>Are really on the way of removing the driver and

0:06:00.200 --> 0:06:05.400
<v Speaker 4>a very cost effective solution. Our sensor set is you know,

0:06:05.720 --> 0:06:10.400
<v Speaker 4>ten thousand dollars twelve thousand dollars. Compute is very very

0:06:10.440 --> 0:06:14.680
<v Speaker 4>cost cost optimized mobilized chips are one tenth of the

0:06:14.720 --> 0:06:19.280
<v Speaker 4>cost of competing chips. We developed our imaging radars which

0:06:19.279 --> 0:06:22.320
<v Speaker 4>are the best on the planet, those imaging radars. So

0:06:22.480 --> 0:06:24.960
<v Speaker 4>I feel very confident that we are on our way

0:06:25.000 --> 0:06:26.240
<v Speaker 4>to a new growth engine.

0:06:26.279 --> 0:06:28.479
<v Speaker 3>With the robotax your striking deals.

0:06:28.920 --> 0:06:30.479
<v Speaker 1>We understand as us all to make it that you

0:06:30.480 --> 0:06:33.719
<v Speaker 1>struck a deal with the nine million cars for safety software.

0:06:34.520 --> 0:06:38.599
<v Speaker 4>Okay us automaker that are three to four no pick one.

0:06:42.240 --> 0:06:47.000
<v Speaker 4>But I think that this nine million units is a

0:06:47.000 --> 0:06:50.080
<v Speaker 4>big news because we are talking about the evolution of AIDAS.

0:06:50.240 --> 0:06:52.599
<v Speaker 4>So aid US today is a front facing camera. Maybe

0:06:52.640 --> 0:06:56.200
<v Speaker 4>you have a front facing radar. This nine million units

0:06:56.279 --> 0:06:59.599
<v Speaker 4>is called surround AIDAS. You are taking a front facing camera,

0:06:59.720 --> 0:07:04.279
<v Speaker 4>four parking cameras, multiple radars, a very strong value proposition

0:07:04.400 --> 0:07:09.159
<v Speaker 4>to the customer, and it replaces multiple ECUs in the car.

0:07:09.360 --> 0:07:11.720
<v Speaker 3>It replaces the aco F parking AEO OF for.

0:07:11.880 --> 0:07:17.800
<v Speaker 4>Driving monitoring system, hands free driving on highways, safety.

0:07:17.880 --> 0:07:19.760
<v Speaker 3>Functions for the next five years.

0:07:19.880 --> 0:07:22.440
<v Speaker 4>So this is the evolution of AIDAS and it's high

0:07:22.520 --> 0:07:25.920
<v Speaker 4>volume We have two customers in the past year, Volkswagen

0:07:25.960 --> 0:07:28.640
<v Speaker 4>and this US customer nineteen million units.

0:07:29.240 --> 0:07:31.480
<v Speaker 3>This is big AI twenty one.

0:07:32.160 --> 0:07:35.560
<v Speaker 5>We're hearing that in Vidia Misto Loong is quite interested

0:07:35.600 --> 0:07:38.600
<v Speaker 5>in another one of your companies. He's here this week.

0:07:38.720 --> 0:07:40.640
<v Speaker 5>What can you tell us about how talks are going?

0:07:41.040 --> 0:07:42.880
<v Speaker 5>Did you manage to speak to Jensen while here?

0:07:43.200 --> 0:07:45.840
<v Speaker 4>You know, the Israeli press is not as tight as

0:07:46.120 --> 0:07:48.480
<v Speaker 4>the US press, So don't believe anything you're reading the

0:07:48.560 --> 0:07:51.680
<v Speaker 4>Israeli press. No, no, that there are talks, talks and

0:07:51.840 --> 0:07:55.160
<v Speaker 4>Vidia talks with others, but no, it's nothing even close

0:07:55.200 --> 0:07:58.040
<v Speaker 4>to talking about it in the press.

0:07:58.680 --> 0:08:01.760
<v Speaker 1>How do you think all I investors would feel about

0:08:01.760 --> 0:08:07.040
<v Speaker 1>an acquisition of AI twenty one by.

0:08:05.120 --> 0:08:09.000
<v Speaker 4>An It has nothing to do or I twenty one

0:08:09.120 --> 0:08:13.320
<v Speaker 4>is working on foundation models and agents, has nothing to

0:08:13.360 --> 0:08:17.040
<v Speaker 4>do with the activity of Mobilize. Activity of Mentee Mobili

0:08:17.240 --> 0:08:17.800
<v Speaker 4>is physical.

0:08:17.840 --> 0:08:20.680
<v Speaker 3>AI. You're a busy man. You're a busy man.

0:08:20.880 --> 0:08:23.560
<v Speaker 5>Busy man for the time here at C s Las Vegas,

0:08:23.560 --> 0:08:27.360
<v Speaker 5>Amon Shashua mobil I C e O. Thank you very much.