WEBVTT - How AI Can Make Manufacturing More Efficient

0:00:00.120 --> 0:00:03.920
<v Speaker 1>These sees. Bloomberg Business Week with Carol Messer and Tim

0:00:03.960 --> 0:00:21.520
<v Speaker 1>Stentovic on Bloomberg Radio, The Whole World worry. We know,

0:00:22.400 --> 0:00:25.000
<v Speaker 1>we keep quoting our conversation with Kathy Would, but you know,

0:00:25.160 --> 0:00:27.480
<v Speaker 1>she's all in when it comes to innovation disruptions. Has

0:00:27.520 --> 0:00:29.600
<v Speaker 1>got a new research report and it really is touching

0:00:29.720 --> 0:00:33.640
<v Speaker 1>on um she's calling big ideas, but it's touching on

0:00:33.720 --> 0:00:36.239
<v Speaker 1>so many things that we touch on a lot and

0:00:36.360 --> 0:00:41.479
<v Speaker 1>as of late, and that includes something like artificial intelligence. Tim. Yeah,

0:00:41.520 --> 0:00:44.280
<v Speaker 1>and look, I think what's interesting too is that the

0:00:44.320 --> 0:00:47.000
<v Speaker 1>news that we got a little earlier today, Carol, uh,

0:00:47.040 --> 0:00:49.480
<v Speaker 1>that you know, you have a series of companies continuing

0:00:49.520 --> 0:00:54.040
<v Speaker 1>to invest in AI, with Google coming out and saying

0:00:54.040 --> 0:00:56.120
<v Speaker 1>that it's making a big investment there. So and we

0:00:56.160 --> 0:00:58.080
<v Speaker 1>feel like on all the earnings calls, like everybody, oh

0:00:58.120 --> 0:01:00.720
<v Speaker 1>my god, Google yesterday, the first sentence from Sunder Pacha

0:01:00.720 --> 0:01:03.480
<v Speaker 1>I was all about AI. So you know, we we

0:01:03.480 --> 0:01:05.200
<v Speaker 1>we see that there's a buzz around it right now,

0:01:05.240 --> 0:01:07.160
<v Speaker 1>and we see that there's this euphoria. It's led to

0:01:07.200 --> 0:01:10.880
<v Speaker 1>some wild swings and stocks like Buzzfeeds C three AI.

0:01:11.040 --> 0:01:13.600
<v Speaker 1>Buzzfeeds saw triple digit gains in certain days because of

0:01:13.640 --> 0:01:17.119
<v Speaker 1>its association with AI. But what about when it comes

0:01:17.160 --> 0:01:20.800
<v Speaker 1>to deploying AI for industrial uses. We've got a great

0:01:20.800 --> 0:01:23.880
<v Speaker 1>guest with us. Alice Globus is CFO at Nanotronics. It's

0:01:23.880 --> 0:01:25.720
<v Speaker 1>a company that says it uses AI to make the

0:01:25.720 --> 0:01:29.280
<v Speaker 1>manufacturing process more efficient. She joins us via zoom from

0:01:29.360 --> 0:01:31.560
<v Speaker 1>New York City. So, Alice, I think a lot of

0:01:31.600 --> 0:01:34.760
<v Speaker 1>people are familiar with what chat gpt can do. Um.

0:01:34.800 --> 0:01:38.400
<v Speaker 1>I had a friend, uh write a review for me

0:01:38.600 --> 0:01:41.800
<v Speaker 1>in chat gpt and it actually was like almost good

0:01:41.880 --> 0:01:44.639
<v Speaker 1>enough for me to actually use for my own self review. Yeah,

0:01:45.040 --> 0:01:47.680
<v Speaker 1>it was pretty amazing. I didn't use it um, but

0:01:47.760 --> 0:01:49.480
<v Speaker 1>he was tempted to use it for his own reviews

0:01:49.560 --> 0:01:51.160
<v Speaker 1>that he has to has to write as we get

0:01:51.160 --> 0:01:53.760
<v Speaker 1>to review season. What about when it comes to manufacturing

0:01:53.960 --> 0:01:56.400
<v Speaker 1>and making the manufacturing process more efficient? What's the tech

0:01:56.440 --> 0:02:00.440
<v Speaker 1>that you have at Nanotronics. Yeah, Well, just like to

0:02:00.480 --> 0:02:02.840
<v Speaker 1>say thank you Carol and Tim and Bloomberg for having

0:02:02.840 --> 0:02:05.840
<v Speaker 1>me here. With all the excitement about AI, a lot

0:02:05.880 --> 0:02:08.120
<v Speaker 1>of people don't realize the uses that it has on

0:02:08.160 --> 0:02:11.560
<v Speaker 1>actual physical things in our life. So one of the

0:02:11.600 --> 0:02:14.360
<v Speaker 1>major problems with manufacturing right now is it's a very

0:02:14.480 --> 0:02:18.239
<v Speaker 1>wasteful industry, and there's a lot of problems with manufacturing

0:02:18.280 --> 0:02:22.000
<v Speaker 1>processes that cause everything from uh a lot of waste

0:02:22.000 --> 0:02:26.120
<v Speaker 1>being manufactured, to energy usage to all these other things,

0:02:26.320 --> 0:02:29.400
<v Speaker 1>sometimes human accidents that really drive the price of our

0:02:29.440 --> 0:02:32.640
<v Speaker 1>goods up significantly more than we actually realize. And then

0:02:32.680 --> 0:02:36.680
<v Speaker 1>there's um companies that have contamination issues that are dealing

0:02:36.720 --> 0:02:39.160
<v Speaker 1>with like thailan all or baby formula that you've heard of.

0:02:39.440 --> 0:02:41.560
<v Speaker 1>And this is where AI really can come in and

0:02:41.560 --> 0:02:47.240
<v Speaker 1>take advantage by helping manufacturers UM optimize their systems in

0:02:47.240 --> 0:02:49.800
<v Speaker 1>a way that humans just don't have the ability to

0:02:49.840 --> 0:02:53.840
<v Speaker 1>do so um more because it's it's overwhelming as looking

0:02:53.840 --> 0:02:56.600
<v Speaker 1>at a manufacturing facility, there's millions of things that are

0:02:56.600 --> 0:03:00.120
<v Speaker 1>happening and being able to assess that and optimize the

0:03:00.200 --> 0:03:03.640
<v Speaker 1>process to have the best product possible. Alice are most

0:03:03.639 --> 0:03:07.200
<v Speaker 1>of manufacturing, even some of the big you know manufacturers

0:03:07.200 --> 0:03:09.160
<v Speaker 1>that are at their global manufacturers aren't they doing this

0:03:09.200 --> 0:03:13.160
<v Speaker 1>already though, using AI to some extent to maximize the

0:03:13.160 --> 0:03:16.320
<v Speaker 1>productivity at their facilities. Yeah, Like I would say they're

0:03:16.360 --> 0:03:18.760
<v Speaker 1>using what I would consider early stages of AI. What

0:03:18.840 --> 0:03:21.480
<v Speaker 1>you're seeing right now is really an AI revolution that's

0:03:21.480 --> 0:03:25.239
<v Speaker 1>happening what you're seeing with chat GPT. They're using a

0:03:25.560 --> 0:03:31.160
<v Speaker 1>relatively cutting edge technology that allows computer to learn the

0:03:31.200 --> 0:03:33.840
<v Speaker 1>way humans learned. And this is what we're doing for manufacturing.

0:03:34.240 --> 0:03:37.120
<v Speaker 1>When our systems are learning like a human does, you know,

0:03:37.200 --> 0:03:41.520
<v Speaker 1>they're penalized and rewarded based on what's happening they learned

0:03:41.560 --> 0:03:44.760
<v Speaker 1>through observation, all based on the end goals. Was for

0:03:44.920 --> 0:03:48.520
<v Speaker 1>most of earth manufacturers that's increasing their yields, but sometimes

0:03:48.560 --> 0:03:51.440
<v Speaker 1>it's reducing their power consumption or other things that they're

0:03:51.480 --> 0:03:54.400
<v Speaker 1>looking to achieve. One thing that we talked about with

0:03:54.520 --> 0:03:59.360
<v Speaker 1>AI is inputs and outputs. What are the inputs that

0:03:59.480 --> 0:04:02.880
<v Speaker 1>you use? Yeah, so it's kind of something that we

0:04:02.920 --> 0:04:05.680
<v Speaker 1>do differently is we actually are not simulating what your

0:04:05.760 --> 0:04:08.840
<v Speaker 1>your factor, your your facility. We actually use your real

0:04:08.920 --> 0:04:13.280
<v Speaker 1>time data that's coming everything from uh, your comput your

0:04:13.320 --> 0:04:16.080
<v Speaker 1>brain of your factory, which is known as a PLC,

0:04:16.320 --> 0:04:18.680
<v Speaker 1>which is kind of takes all the sensors in your

0:04:18.680 --> 0:04:21.320
<v Speaker 1>factory and puts it into one place and we're able

0:04:21.320 --> 0:04:24.840
<v Speaker 1>to take that you know, temperature, humidity, pressure, whatever that

0:04:25.000 --> 0:04:28.640
<v Speaker 1>is those sensors in most cases it's thousands and look

0:04:28.680 --> 0:04:30.920
<v Speaker 1>at them to be able to optimize for the end

0:04:30.960 --> 0:04:35.800
<v Speaker 1>goal of winning the game of manufacturing. So Enter Nanotronics.

0:04:35.839 --> 0:04:37.360
<v Speaker 1>Tell us a little bit about your company, what you

0:04:37.400 --> 0:04:41.719
<v Speaker 1>specifically do against the backdrop of this conversation we're having. Yeah,

0:04:41.720 --> 0:04:45.400
<v Speaker 1>so i'd say we're the first generative AI company for manufacturing.

0:04:45.440 --> 0:04:51.680
<v Speaker 1>We started with early artificial intelligence for inspection, which of

0:04:52.240 --> 0:04:56.560
<v Speaker 1>identifying defects within the manufacturing process. And and that's not

0:04:56.640 --> 0:04:59.800
<v Speaker 1>exactly the most sexy thing, but the reality is when

0:04:59.839 --> 0:05:04.000
<v Speaker 1>you have problems in your materials, especially in industries where

0:05:04.000 --> 0:05:07.240
<v Speaker 1>it's a long time to manufacture something like the semiconductor space,

0:05:07.640 --> 0:05:10.440
<v Speaker 1>the sooner you can identify problems that are in your product,

0:05:11.160 --> 0:05:13.839
<v Speaker 1>the quicker you could either address them in your manufacturing

0:05:13.960 --> 0:05:17.760
<v Speaker 1>process or replace your supplier. Communicate with your supplier that

0:05:17.760 --> 0:05:20.880
<v Speaker 1>there's a problem with that coming in. So that allows

0:05:20.880 --> 0:05:22.960
<v Speaker 1>you to increase your yields and reduce your costs in

0:05:23.000 --> 0:05:27.000
<v Speaker 1>the overall process. Talk to me about customers. Who do

0:05:27.000 --> 0:05:29.440
<v Speaker 1>you have out there right now? You know we're working

0:05:29.480 --> 0:05:33.719
<v Speaker 1>with most fortune manufacturers. We have over two customers ranging

0:05:33.760 --> 0:05:38.359
<v Speaker 1>everything from biotech to semiconductor to automobile industries. How do

0:05:38.400 --> 0:05:42.560
<v Speaker 1>you how do you sell it? That's a good question.

0:05:42.920 --> 0:05:46.599
<v Speaker 1>You know, we've been in in an inspection for over

0:05:46.720 --> 0:05:48.640
<v Speaker 1>a decade, so this is one of those things that

0:05:48.680 --> 0:05:51.800
<v Speaker 1>we started with. I would say the hardest industry to penetrate,

0:05:51.839 --> 0:05:55.240
<v Speaker 1>which is the UM the semiconductor space. You know that

0:05:55.480 --> 0:05:57.839
<v Speaker 1>they are really more lenient to going to some of

0:05:57.839 --> 0:06:01.240
<v Speaker 1>the larger manufacturers, but right now what we're seeing is

0:06:01.320 --> 0:06:03.480
<v Speaker 1>that they're hitting a wall with what can be physically

0:06:03.480 --> 0:06:06.839
<v Speaker 1>done by hardware, and the only way to overcome this

0:06:06.960 --> 0:06:11.080
<v Speaker 1>is really through artificial intelligence. It allows our customers to

0:06:11.120 --> 0:06:13.920
<v Speaker 1>able to do predictive maintenance, to actually have real time

0:06:13.960 --> 0:06:18.440
<v Speaker 1>feeds from supply chain UM distribution like their EARP systems

0:06:18.480 --> 0:06:21.400
<v Speaker 1>such as SAP or one of those systems, to change

0:06:21.520 --> 0:06:26.719
<v Speaker 1>their process based on shipment delays or even people that

0:06:26.720 --> 0:06:31.520
<v Speaker 1>are interfering or you know, fatigue that's coming into the process.

0:06:31.560 --> 0:06:33.960
<v Speaker 1>So these are things that we're able to take into

0:06:34.000 --> 0:06:39.080
<v Speaker 1>account and help optimize. You know, Alice, I kind of

0:06:39.160 --> 0:06:42.000
<v Speaker 1>keep thinking, we've been talking about AI for years. This

0:06:42.040 --> 0:06:44.599
<v Speaker 1>isn't new, it's been around since the nineteen fifties. But

0:06:44.680 --> 0:06:48.400
<v Speaker 1>I do wonder you you know, and you mentioned generative AI,

0:06:48.520 --> 0:06:52.640
<v Speaker 1>which is this idea of generating novel content right like

0:06:52.800 --> 0:06:56.919
<v Speaker 1>chat GPT. Was there something that has happened in the

0:06:57.000 --> 0:06:59.960
<v Speaker 1>last year or so that all of a sudden we're

0:07:00.080 --> 0:07:03.000
<v Speaker 1>talking about AI and rightfully? So I just want to

0:07:03.000 --> 0:07:07.760
<v Speaker 1>make sure that our level of um focusing on it.

0:07:08.240 --> 0:07:11.040
<v Speaker 1>I mean, we see companies obviously doing big deals like Microsoft,

0:07:11.080 --> 0:07:13.440
<v Speaker 1>and we see Google doing a much smaller deal, but

0:07:13.480 --> 0:07:16.760
<v Speaker 1>nonetheless these are big names involved in it. But I mean,

0:07:16.920 --> 0:07:20.320
<v Speaker 1>are we rightfully focusing on it right now? Because there's

0:07:20.320 --> 0:07:24.200
<v Speaker 1>some new development that has made it much more useful

0:07:24.320 --> 0:07:29.760
<v Speaker 1>in our world. So, actually, the big breakthrough in some

0:07:29.840 --> 0:07:34.640
<v Speaker 1>of this generative AI technology happened in and I think

0:07:35.000 --> 0:07:37.200
<v Speaker 1>why it's taken so long to actually get into the

0:07:37.200 --> 0:07:38.840
<v Speaker 1>mainstream is that there's been a lot of R and

0:07:38.920 --> 0:07:42.840
<v Speaker 1>D that that's surrounding it. And it's honestly a combination

0:07:43.000 --> 0:07:46.480
<v Speaker 1>of human acceptance of AI and willing to take the

0:07:46.560 --> 0:07:50.800
<v Speaker 1>chance of giving over control to a black box system.

0:07:50.880 --> 0:07:52.960
<v Speaker 1>And that you mean in the corporate world right when

0:07:53.000 --> 0:07:55.960
<v Speaker 1>you talk about corporate world for yeah, and the corporate

0:07:55.960 --> 0:07:59.280
<v Speaker 1>world for sure, and then you know, coming with with

0:07:59.400 --> 0:08:04.080
<v Speaker 1>AI for chat CHYPT. These techniques have been developing since

0:08:04.520 --> 0:08:07.000
<v Speaker 1>seen there was a scientific paper that it really stems

0:08:07.040 --> 0:08:10.000
<v Speaker 1>from this, But you know, there's been advances in speed

0:08:10.040 --> 0:08:14.360
<v Speaker 1>and internet uh speed connectivity. These things also help increase

0:08:15.040 --> 0:08:20.680
<v Speaker 1>adoption of AI within just a social setting. Alright, we

0:08:20.720 --> 0:08:22.600
<v Speaker 1>only have forty seconds left here. What was it like

0:08:22.640 --> 0:08:26.600
<v Speaker 1>to work with Neil de grasse Tyson? Just quickly? Which

0:08:26.680 --> 0:08:30.280
<v Speaker 1>you did do you're an astrophysicist, just quickly? What was

0:08:30.320 --> 0:08:34.160
<v Speaker 1>it like? You know, the it changed my life. I

0:08:34.480 --> 0:08:36.679
<v Speaker 1>always say that everyone should encourage everyone to get their

0:08:36.720 --> 0:08:41.320
<v Speaker 1>PhD in astrophysics because it shows that, you know, problem

0:08:41.520 --> 0:08:44.600
<v Speaker 1>is too small to fix. Um. You know, we're trying

0:08:44.640 --> 0:08:47.680
<v Speaker 1>to solve some of the world, the universe's largest problems.

0:08:47.679 --> 0:08:50.360
<v Speaker 1>And when I look at you know, something like manufacturing

0:08:50.360 --> 0:08:54.440
<v Speaker 1>and be able to have carbon negative manufacturing facilities across

0:08:54.440 --> 0:08:57.679
<v Speaker 1>the world being optimized through it actually seems like an

0:08:57.679 --> 0:09:01.000
<v Speaker 1>attainable future that we can we can do. And you know,

0:09:01.080 --> 0:09:04.679
<v Speaker 1>it was definitely an incredible experience. Well. I love the optimism,

0:09:04.880 --> 0:09:08.920
<v Speaker 1>uh and I like the idea that no problem is insurmountable.

0:09:09.040 --> 0:09:11.600
<v Speaker 1>So um, really a great way to start to wrap

0:09:11.679 --> 0:09:14.120
<v Speaker 1>up our Friday. Alice, thank you so much. Alice Globe

0:09:14.160 --> 0:09:18.800
<v Speaker 1>is Chief financial Officer Antotronics via zoom from New York City,