Risk, Resilience, and AI in the Supply Chain with Yossi Sheffi

[00:00:00] and responsible companies are building resilience.

[00:00:02] But they are building resilience in a very smart way.

[00:00:05] For example, a big technology company,

[00:00:08] every product that they buy,

[00:00:10] they look at five levels of resilience.

[00:00:12] Level one, they just look all over now with AI

[00:00:15] and find out who else is supplying this product

[00:00:18] aside from their current suppliers.

[00:00:20] Level two, they take a sample from two or three of the others.

[00:00:24] Level three, they want to have several samples time just to see this consistency in the part.

[00:00:30] Level four, they start taking test runs. Level five, they actually use a second supplier.

[00:00:35] Depending on the situation, they calibrate themselves, but they're always ready to move

[00:00:39] and they have the information. Hello and welcome to the next level supply chain with GS1US,

[00:00:45] a podcast in which we explore the mind-bending world of global supply chains,

[00:00:50] covering topics such as automation, innovation, unique identity, and more.

[00:00:55] I'm your co-host, Reed.

[00:00:56] And I'm Liz.

[00:00:57] And welcome to the show.

[00:01:00] Reed and I just had a great conversation with Yossi Sheffi,

[00:01:04] who's the director of the Center for Transportation and Logistics at MIT.

[00:01:09] I was lucky enough to be able to see him speak to a classroom of executives.

[00:01:15] A few weeks ago, and he just has such a great background in all things supply chains.

[00:01:20] He's written several books and Reed and I just couldn't have had more fun chatting

[00:01:25] with him. He mentioned something about supply chain managers really were the heroes of the past

[00:01:31] several years. And that just hit close to home because you hear so much about what the supply

[00:01:36] chain went through and really became a kitchen table conversation. So hearing his perspective on

[00:01:42] that was really fun. And also how organizations need to either invest in resilience or in additional insurance

[00:01:49] because things are going to continue to change.

[00:01:52] We really hope you enjoy this episode.

[00:01:54] Thanks.

[00:01:55] It is so good to be with you, Yossi.

[00:01:58] We are so excited to have this conversation.

[00:02:01] I was lucky enough to be in Boston last week at MIT and listen to you

[00:02:07] speak a little bit at the executive education supply chain forum that y'all hold. And we

[00:02:14] are just in for a treat with all things supply chain resiliency and what you're seeing both

[00:02:20] past and what you're seeing that's going to be happening in the future. So just

[00:02:25] real quick, just quick baseline, tell us a little bit about yourself, your

[00:02:29] background and the role that you have at MIT. I've been at MIT 49 years, so by now

[00:02:36] I know where the bathrooms are. I did my master PhD in civil engineering but

[00:02:40] really all my courses were in the operation research, more engineering systems. I worked at the beginning on urban transportation

[00:02:48] planning, used network theory for urban transportation, got frustrated when

[00:02:52] people when I thought I had good ideas but went to mayors all over the United

[00:02:56] States, nobody implemented my ideas. So I started with the private sector, started

[00:03:01] work with truck lines and started a company and then many Tracklines still today are using software that I wrote and then I started working with

[00:03:10] Shipper, the beneficial owner of frame on how to use transportation and then on procurement

[00:03:15] and all the issues and then got to end to end supply chain basically including from the mind

[00:03:20] to distribution to the claiming stuff. So I'm at MIT.

[00:03:25] I'm a professor of engineering systems.

[00:03:27] I am heading the MIT Center for Transportation Logistics.

[00:03:31] We just celebrated our 50th year.

[00:03:34] I was not there all 50 years.

[00:03:36] Even though I've been the last 32 years or so,

[00:03:39] the center is called the largest interdepartmental center

[00:03:42] at MIT.

[00:03:43] It has several areas we do of course research.

[00:03:46] MIT centers are founded in order to do research. The center has the research many, many areas.

[00:03:52] We provide our own graduate degrees. We have a group of industries that are kind of our partners,

[00:03:58] what we call supply chain exchange, fish decomposing, trying to keep it

[00:04:01] manageable size, that we do a lot of activities with many of

[00:04:05] them, I are students, work with our students, and we have five centers around the world that

[00:04:10] we set up in Colombia, in Spain, in Luxembourg, in Malaysia, and in China, and we are in shape

[00:04:17] negotiating a new center now. We're talking about it when the time comes.

[00:04:22] All right, we would like to jump in today. The first topic of supply chains is really around resiliency.

[00:04:29] I would love to get your perspective on this

[00:04:32] because what we've found in the past,

[00:04:34] especially COVID really highlighted this,

[00:04:36] like supply chains are always focused on efficiencies, right?

[00:04:41] Where can we just squeeze everything out of it

[00:04:43] and make it so efficient, but it doesn't take much to throw a little wrench into the system and have it go haywire.

[00:04:50] And I think COVID really taught us that we need resiliency in order to be efficient. And I would love to just get your perspective on what resiliency really means and why or why it's not as important as we may think so.

[00:05:06] So first of all, just to have fun, let me take exception to some of the things you said.

[00:05:11] You said, and please, every little thing can throw supply chain of kilter.

[00:05:16] That's just not the case.

[00:05:18] I mean, a huge pandemic that affect the entire world and closes everybody and affect the

[00:05:23] man and closes factories, yes, can affect supply chain. But supply chain does small disruptions all the time.

[00:05:29] Relatively small. When I said small, some of them are pretty big. And yet supply chain

[00:05:35] worked through them. It's the first time this kind of massive disruption that got from the

[00:05:42] boardroom to the kitchen table when people started feeling it personally.

[00:05:46] This was really the first time.

[00:05:48] We had Katarina, we had Japan, we had many, many big disruptions,

[00:05:53] and there were here and there shortages.

[00:05:56] People did not feel it.

[00:05:58] So, first of all, it was a huge disruption effect of supply chain.

[00:06:02] Let me say second,

[00:06:04] supply chain managers

[00:06:05] were the heroes of the COVID-19 pandemic.

[00:06:09] Think about it.

[00:06:10] For example, things kept going despite supply chain.

[00:06:14] People didn't cause COVID-19.

[00:06:16] They had to respond to it.

[00:06:18] And it was massive.

[00:06:19] It was massive in the sense that not that nobody could

[00:06:21] anticipate it, nobody could have prepared for it.

[00:06:23] And I'll explain why in a minute. But think about it when I said people who are here think about the food supply chain.

[00:06:29] From one day to the next in March 2020, all restaurants were closed, all school

[00:06:35] were closed, all industrial park were closed. This is half the food in the supply chain go to these

[00:06:40] places. And so if we start getting to, in commerce, this is very different.

[00:06:46] As you know, when you send stuff to a restaurant or industrial park, at a university, it is sent

[00:06:52] on toilets with 50-pound sacks. It's not sent in small little packages that have all the ingredients

[00:06:59] of them and how many calories and all the... So, the machinery is not even there to do it. There's not enough machinery to do.

[00:07:06] And while restaurants were closed,

[00:07:08] as I said, supermarket in the study,

[00:07:10] the man was going through the roof.

[00:07:12] Yet, let's get real.

[00:07:14] Nobody went hungry.

[00:07:15] So yes, sometimes a certain cut of meat

[00:07:19] was not available or certain eggs

[00:07:22] were for one week, less availability.

[00:07:24] And the skies were falling.

[00:07:25] People are so used to having plenty of everything

[00:07:29] at anything they want, that the sky was falling.

[00:07:32] And I took exceptionally at the time

[00:07:35] with some of my reporter friends.

[00:07:38] People talk about the United States and mid shortages.

[00:07:42] The Boston Globe and the New York Times.

[00:07:43] Mid shortages come to the United States.

[00:07:46] I know the Boston Globe,

[00:07:47] but you know, I say, what are you talking about?

[00:07:48] Do you even know that the United States

[00:07:51] is one of the largest fore-exporters of meat in the world?

[00:07:55] Even realized that we have more meat

[00:07:57] than we know what to do with?

[00:07:59] Yes, because we could distribute something,

[00:08:01] some plants were closed,

[00:08:02] but it's not like there was no protein.

[00:08:04] So you didn't have certain cuts of meat yet, other cuts of meat.

[00:08:07] It's just a ridiculous statement.

[00:08:10] And we had a lot of these, not eggs, not these, not that, come on.

[00:08:14] So by and large, first of all, it took a massive, massive disruption.

[00:08:20] Not every little thing gets the supply chain out of work.

[00:08:23] In fact, many supply chain managers make sure they don't.

[00:08:27] And they have all kinds of other ways to deal with it.

[00:08:31] I love the fact that you're keeping us straight.

[00:08:33] I really appreciate your perspective because we can.

[00:08:36] We can over rotate and we can over emphasize.

[00:08:39] And you're really calling out that it took a massive, massive disruption.

[00:08:44] And we still survive through it.

[00:08:46] I have some theories on some other aspects, but I really appreciate your perspective on that.

[00:08:50] And this is why we're excited about this show as well. Let's talk about the resiliency side.

[00:08:56] What's your perspective on that? Okay, so resilience has a problem. There are two things that people

[00:09:02] are focused on resilience and sustainability. And the reason that this is a problem is because despite what everybody

[00:09:09] are, they don't minimize cost. There'll be a hit to short-term cost, whether it'll be

[00:09:13] more sustainable or more resilient and have more inventory, go to multiple suppliers,

[00:09:18] whatever you do to create resilience or whatever you do to create sustainability,

[00:09:22] make sure that you have only certain of home from certain places, it will raise costs. So now the question is for

[00:09:28] us, unless everybody does it, the people who don't invest in resilience will have

[00:09:33] a short-term competitive advantage until, let's say, there's another massive

[00:09:37] disruption. Think about it, the tenure of CEO in the United States is what three

[00:09:42] years, four years are best. They all say, hey, as long as I get my bonus,

[00:09:47] everything is fine.

[00:09:48] Three to four years I'll take my chances

[00:09:50] that nothing will have to.

[00:09:51] It's very hard, and especially since there are also

[00:09:54] international competitors who don't play the game.

[00:09:56] It's a fundamental problem.

[00:09:58] I should say that companies do think about it,

[00:10:01] and responsible companies are building resilience.

[00:10:04] But they are building resilience in a very smart way.

[00:10:07] For example, a big technology company.

[00:10:10] They have five levels of every product that they buy.

[00:10:14] They look at five levels of resilience.

[00:10:16] They look at level one.

[00:10:18] They just look all over now with AI, all over the world,

[00:10:22] and find out who else is supplying this product aside from their

[00:10:26] current suppliers. Level two, they take a sample from two or three of the others. Level

[00:10:32] three, they want to have several samples time just to see this consistency in the part.

[00:10:38] Level four, they start taking test runs. Level five, they actually use a second supplier. So they're always, depending on the

[00:10:46] geopolitical situation on disruption, they kind of move between these levels. So they don't go all

[00:10:52] out there, multiple suppliers, because multiple suppliers cost money, because you have to split

[00:10:56] your buy between multiple suppliers, you get less discount, it's more administration to deal

[00:11:01] with multiple suppliers. So they don't go all out immediately to max two or three supplies.

[00:11:06] But depending on the situation, they calibrate themselves.

[00:11:09] But they're always ready to move and they have the information.

[00:11:12] So that for me is a way of doing it smartly.

[00:11:15] Rather than simply say, we need a lot more inventory.

[00:11:18] Companies today cannot just have a lot more inventory.

[00:11:22] It's expensive.

[00:11:23] Furthermore, another thing that I got at my

[00:11:25] friends in the media, especially New York Times and similar media, Washington Post,

[00:11:31] many times, they write that these greedy supply chain managers are trying to save on inventory

[00:11:38] in order to save money. I said, come on, guys. just in time, the old Toyota, Manofic Toyota production system was there to get quality of the product.

[00:11:48] Originally, there's nothing to do with money.

[00:11:51] It's to get quality product.

[00:11:53] Turns out, to get quality product, you have to just in time system, so for example,

[00:11:59] you don't have a pile of stuff that if a part goes wrong,

[00:12:01] you just go to the pile and take another one.

[00:12:03] No, no, no, you go out there and check why there's something that's wrong and fix it at the source.

[00:12:08] So it doesn't happen again. So what happens because of this? You have high quality, which means

[00:12:13] less warranty cost, less rework of cars when you find the problem after the car was made,

[00:12:18] and reputation for good quality. All of this actually reduced the cost at the end of the day, but this was not

[00:12:25] the driving idea. So again, coming to resilience, resilience is a turn taken from a material science,

[00:12:33] is the ability of a metal to retain its form of ship after deformation. Here, we are talking

[00:12:38] about the ability of an organization to get back to the same similar KPI, whatever it is,

[00:12:44] level of manufacture, level of service,

[00:12:46] whatever is the KPI that they worry about

[00:12:48] after some kind of disruption.

[00:12:49] How quickly can they get back to this level?

[00:12:52] As I said, there's a lot of way to achieve resilience,

[00:12:54] and it's really a multifaceted approach.

[00:12:57] Some companies are worried

[00:12:58] where they should invest in resilience

[00:13:00] or increase their insurance, for example.

[00:13:03] Because they have some similar aspects in

[00:13:07] that you invest in something like you paid the premium and you hope you'll never use

[00:13:11] it.

[00:13:12] Resilience is the same.

[00:13:13] You invest in certain property and processes and you hope you'll never use it.

[00:13:16] You hope you'll never be disrupted.

[00:13:18] However, as I talk to companies on overattend and the investor in resilience is better than

[00:13:23] insurance because several reasons.

[00:13:25] First of all, you can get insurance only against named issues.

[00:13:29] And who knows what will happen that's not named?

[00:13:31] I mean, you could get it against, let's say, flood in certain area or earthquake, taking

[00:13:36] it against your competitor coming out with a better product, nobody will give you this

[00:13:40] insurance.

[00:13:41] There are lots of things that insurance doesn't cover.

[00:13:43] Second thing, insurance by its very nature is an antagonistic process. In fact, some companies in their annual

[00:13:51] report said one of the risks they're facing is they'll have a major disruption and the insurance

[00:13:57] is not going to pay. Either they'll go out on business or just for a few years to pay you.

[00:14:00] In many times, you go to court in order to get insurance to pay. So it's antagonistic process.

[00:14:06] Their insurance doesn't actually help you when you don't fulfill your commitment

[00:14:12] to customers.

[00:14:13] You have a customer and you have a contract with them and you can't do it.

[00:14:17] And the customer will say, you know what, I'm going somewhere else.

[00:14:19] What insurance is going to cover the insurance that cuts?

[00:14:21] So there's a lot of things that insurance doesn't cover.

[00:14:24] Finally, maybe supply chain disruption is a basically create a mismatch between demand and supply.

[00:14:30] So there's usually more demand and supply.

[00:14:33] Resilience, when you build in flexibility, when you build in agility to respond,

[00:14:38] it also helps you when you have a sudden surge in demand.

[00:14:43] Because it has the same elements, but if you build it into your company that people are able to move and anticipate and have a playbook

[00:14:51] What to do in certain cases and you actually drill it then you get the ability to respond better

[00:14:57] So building resiliency can help with just normal supply chain increases in demand, which is a good problem

[00:15:03] Not just when something

[00:15:05] goes wrong. Absolutely.

[00:15:07] Under an earphone. Building resilience, a lot of it is just building flexibility and agility

[00:15:13] and the ability to respond and in a mindset. Honestly, there's an element of mindset.

[00:15:18] You know, I talked to many of my friends that said, if you want to build resilience, make

[00:15:22] sure to hire enough Argentinians or

[00:15:28] People who are used to disruptions who always look over the shoulders

[00:15:34] Well, not just sit there and everything is working and only small waves in the flow But people who suddenly we have the currency is not worth anything or something else happening. They didn't get it

[00:15:40] So I make Argentina, but there are several places like this that things are just moving around and people are used to it.

[00:15:48] Do you think some of the challenges that companies have

[00:15:51] when they start talking about resiliency

[00:15:54] is that they have an idea of what resiliency may mean,

[00:15:57] and then they're going at it maybe in the wrong way?

[00:16:01] When somebody says, I want to do resiliency,

[00:16:03] you first have to define it, right? With you said

[00:16:05] it's a metal, that's where the term came from. But how does somebody start this conversation?

[00:16:11] Maybe the number one challenge which is still not quite solved is to begin with how to measure

[00:16:17] risk. What you want to do is resilient to respond to certain risks and risks are so varied and so on over the place and can come from any direction

[00:16:27] that it's very hard to measure risk and because what you want to do in the ideal world is to say

[00:16:33] my risk is 7.3 and I did this measure and it's now 4.2. Great, I reduced my risk. We don't have

[00:16:40] these tools, they just don't exist because it is so common. Maybe no AI method in the future will be able to do it, but right now we don't know how to do it.

[00:16:50] So in that sense, that's part of the main reason.

[00:16:53] We invest in it in some cases on blindfade, or we look at certain risks.

[00:16:58] So for example, look at the risk that the supplier will go out of business.

[00:17:01] So we'll not be able to supply because they'll have a strike or China doesn't allow them to do something. So we take actions for certain risks that we kind of

[00:17:10] can envision. So come back. People should have envisioned the pandemic. But so what? So you envision

[00:17:16] the pandemic. What would you do about it? You start pining a lot of inventory. So one company did it.

[00:17:22] Amazingly, you know, which company did accumulate microprocessor

[00:17:26] chips to Yota.

[00:17:28] After Japan, even though they are the symbol for just in time,

[00:17:31] after the Japan earthquake and nuclear disaster,

[00:17:35] they realized that rebuilding supply of chips take time,

[00:17:38] even though they were not really hurt,

[00:17:40] but they did analysis of everything that they buy.

[00:17:43] It's a building of no plan.

[00:17:44] It's four years. You cannot wait. buy. They said, building a new plant is four years.

[00:17:45] You cannot wait and just increasing capacity, take a year.

[00:17:49] So Toyota started accumulating big inventory of chips.

[00:17:54] When the pandemic hit,

[00:17:55] they were the only company that were building car

[00:17:57] at the same rate for the first eight months of COVID.

[00:18:00] Now, they're after it, they ran out.

[00:18:02] It held them for a while.

[00:18:03] But by the way, at this period period the super general motors to become the largest

[00:18:08] Car sailor in the world because they were still building and nobody else could build so for a while

[00:18:13] They were still built for about eight months a huge inventory

[00:18:15] So there's an example of company that did analysis say ships will be a problem

[00:18:21] So we'll have inventory despite our just-in-time philosophy

[00:18:25] because we are reasonable.

[00:18:26] We don't just blindly follow anything.

[00:18:28] We just do the analysis.

[00:18:30] And they be, that's what I mean.

[00:18:32] There's no following blindly what you also said.

[00:18:35] There's no silver bullet here.

[00:18:37] Do the analysis look where things are, where the risks are.

[00:18:41] There are so many projects of trying to measure

[00:18:43] company risks, general company risks.

[00:18:45] And it's a full-term, it just doesn't work.

[00:18:48] And it says as a person who tried to do it himself.

[00:18:50] Even with something hit, you don't know for how long it's going to hit, what are the

[00:18:54] measures that companies can use to mitigate it, what help they'll get from competitors.

[00:18:59] You know, amazingly, during COVID and under huge disaster, huge disruptions, companies

[00:19:05] were getting help from competitors.

[00:19:07] You cannot build it in.

[00:19:08] So that's why even what will happen, the timing of it, how long it will take, very hard since

[00:19:14] it's very hard to measure risk.

[00:19:16] It's very hard to say, I put that much money into resilience and I reduced my risk by X percent.

[00:19:23] Doing an ROI on this is virtually impossible.

[00:19:25] Yeah, it really comes down to more of best practices and keeping that flexibility and

[00:19:30] agility that you've discussed in here and you've shared a lot of examples.

[00:19:34] I'd like to pivot here just a little bit and move away from the resiliency conversation

[00:19:40] and talk a little bit about technology.

[00:19:43] Artificial intelligence is the buzzword and phrase and term

[00:19:47] and everybody's using it for this, that and the other thing. But what technologies are you

[00:19:52] seeing today? Could be artificial intelligence, could be other things. But what technologies are

[00:19:57] you seeing that are having the biggest impacts within supply chain today and moving forward

[00:20:03] in the next year or two? First of all about generative AI.

[00:20:06] At this point, when everything is said and done,

[00:20:09] a lot more is said than done.

[00:20:11] We're still in the hype care, but first of all, AI,

[00:20:14] and is it termed that they've been for 50 years?

[00:20:17] I mean, a lot of AI,

[00:20:18] I start from the old expert system and moving forward.

[00:20:21] When you talk about things like machine learning, for example,

[00:20:24] we had it even before the generative AI, system and moving forward. When you talk about things like machine learning, for example,

[00:20:25] we had it even before the generative AI, machine learning is used through a tremendous amount

[00:20:30] of data, trying to find out what's going on in the data, but then it can be led by a person.

[00:20:37] So for example, how does daily AI know how to recognize a cat in a picture? Well, it

[00:20:43] shows them azilial pictures, and then some of them are cat, some of them are not.

[00:20:47] And then the human beings have vicious cat.

[00:20:49] You look at the label, as long as something is labeled,

[00:20:51] you can do.

[00:20:52] That is supervised learning.

[00:20:54] Machine learning use a lot for forecasting.

[00:20:57] It's simply using a very large, much bigger data,

[00:21:01] many more variables, trying to connect with what happens.

[00:21:04] And then you can see what happens, and you can go back and correct and fix it.

[00:21:07] Generally, AI is a new type of AI, we don't get excited, but large language model, we

[00:21:13] always try to shed the ability to write poems to our wives or loved ones or whatever.

[00:21:18] Well, actually, it's being used already.

[00:21:21] It's being used in defined spaces.

[00:21:23] So for example, chatbots. When you talk to your

[00:21:26] favorite cable provider and you used to call them and they say, press one to get these, press two

[00:21:32] to get these, press three, you press three, you get seven other manuals, you press five, you get

[00:21:36] six other manuals, and then you get frustrated, then you scream agent agent agent and somebody

[00:21:41] comes online. Now you actually talk to the chatbot. That's actually a large language model. It's not that large, it's specific language, we're just in technology.

[00:21:49] You talk to them, they talk big, they try to understand what you're saying. After a while,

[00:21:53] you get frustrated, you scream agent agent agent agent and the human comes online. But these

[00:21:57] things are becoming better. The place that they really work at the place when the vocabulary is

[00:22:02] limited. So you go to a dank it donut window to order

[00:22:05] something. Oh, I'm not doing anything. And you talk to them and it works. It works very well. Why?

[00:22:10] Because you don't go to dank it donuts driving window and ask questions like, is there gone?

[00:22:16] You go and just say, I want French fries with the or whatever. I want sprinkled donut. The number

[00:22:21] of words and then senses use is limited. It's easy for them to understand.

[00:22:25] So it works.

[00:22:26] So when you have problems that are well-defined,

[00:22:28] things like even today, a lot of people

[00:22:31] are using a large language model in order

[00:22:33] to go through CVs.

[00:22:36] Companies may have job opening, may have many, many CVs.

[00:22:39] You have large language, and then going through that

[00:22:41] to see what's going on.

[00:22:43] And get the words and the phrases that they really can help them in what they need. You have onboarding a lot of administrative

[00:22:49] processes when you can actually put some framework around it, some define it, call it

[00:22:55] being in the sandbox. So it's not just open-ended, everything goes. These are working well and more

[00:23:01] and more people are using large language models for this. And now the new language models that I actually combine not only language, but video and pictures. Another

[00:23:12] thing that the large language model I use for it is say, what's the probability that the

[00:23:16] supplier will go out of business? It used to be so people can look at Denon Brent Street

[00:23:21] and others, financial statements, but these are backwards looking. They are two quarters old usually by the time you see them.

[00:23:27] So they don't give you a forward view.

[00:23:29] What does give you forward view is if you look

[00:23:32] at the social media, regular media TV,

[00:23:35] and you see a lot of mentions of redundancy

[00:23:38] and projects that don't work,

[00:23:40] and the executives that are living,

[00:23:42] things like this, and about six months before a company goes out of business,

[00:23:46] you can see this.

[00:23:47] You can see the growing, growing, growing.

[00:23:49] Now companies knew it.

[00:23:50] I wrote about it in my first book in 2005.

[00:23:53] Companies were doing it,

[00:23:54] but they can do it for 10 critical suppliers.

[00:23:57] Now companies do it for tens of thousands suppliers

[00:24:01] because they can look at all of them

[00:24:03] and the software will bring up those

[00:24:04] that a person has to look at. of them and the software will bring up those that a person

[00:24:05] has to look at. So these are some examples of things that are happening today. Now in terms of

[00:24:10] technology in general robotics is becoming very important in supply chain. Wherehouses are introducing

[00:24:17] robotic at a breakneck speed. Autonomous tracking is moving along. We have tests that are running in the southern

[00:24:25] United States. They're running in Mexico and Texas. The southern states, when the roads

[00:24:30] are open and the weather is good, they don't run in Boston. So when the roads are crazy

[00:24:36] and the weather is at that point, it still needs a lot of technological improvement,

[00:24:40] but it needs one more thing. Jogs in general don't disappear overnight

[00:24:45] because it takes time for all these technologies to catch up.

[00:24:48] Take autonomous flying.

[00:24:50] Today 787 is a drone.

[00:24:53] It can fly by itself.

[00:24:54] Not too many of your listeners will go on an aluminum

[00:24:58] tube flying at 35,000 feet when nobody is at a week.

[00:25:01] Nobody is at a stick.

[00:25:03] It will come, people believe it will come,

[00:25:05] but it will take a lot of psychological adjustment. Similarly, it will take some getting used to

[00:25:12] to a tunnel struck going behind you on a highway at 60 miles per hour with no driver. That will

[00:25:18] take some getting used to and getting comfortable with it. It's not clear it will happen and Congress

[00:25:24] not clear that we'll go for it. So we'll see. The interesting thing in the United States

[00:25:28] transportation regulation are state by state so some states may go for it some

[00:25:32] state will not but it's not gonna be America wide not anytime soon. As you see

[00:25:37] again I think I mentioned before, long combination vehicles are not enough

[00:25:41] everywhere in the United States. Some states allow it some state they don't.

[00:25:44] We'll see how all this develops.

[00:25:46] But is it?

[00:25:47] Robotics, we talk about a drone delivery,

[00:25:50] started to help now that stayed a lot of problems

[00:25:53] with all of this technology and the acceptance.

[00:25:55] And drone delivery actually does not work well

[00:25:58] in the southern part of the United States

[00:26:00] because people shoot at the drone.

[00:26:02] They look like moving targets.

[00:26:04] It's hard to get people to respect the drone.

[00:26:06] But it's being used, it's being used in Africa

[00:26:09] for medical supplies, works very well.

[00:26:11] There are experiments going on with drones

[00:26:14] that are flying off vehicles,

[00:26:15] and the vehicle goes to surfaces.

[00:26:18] Drones does their final, final delivery.

[00:26:20] We'll see how that works.

[00:26:22] All of this that you just said is fascinating.

[00:26:24] It'll be interesting to see acceptance. You had mentioned that about having a pilotless plane and

[00:26:30] going home from Boston last week. I thought, no, whether it can fly itself or not, there best be

[00:26:36] two humans in the cockpit, just in case. By the way, two humans are going away. All the manufacturers

[00:26:44] experimenting now with one pilot.

[00:26:46] They don't believe they can get two pilots out, even though technologically they can.

[00:26:50] But they believe they can get one pilot out and the other will be AI.

[00:26:54] This has been so great.

[00:26:56] We're running a little bit low on time.

[00:26:58] There's two questions that Reid and I ask all of our guests as we're finishing up our

[00:27:03] podcasts.

[00:27:04] I'll ask you the first.

[00:27:05] I'm excited to hear your answer from a technology perspective,

[00:27:10] whether it's in your personal life, whether it's work.

[00:27:14] What is the favorite technology that you're seeing

[00:27:17] or using that's out there today?

[00:27:20] It must be communication technology.

[00:27:22] Must be the fact that I go and trip anywhere in the world and I can talk to my wife and see her real time.

[00:27:29] I'm still one of those who used to go to Europe and had to make a telephone call home.

[00:27:34] I had to go to a bank of telephone and some operator would go on and say,

[00:27:38] wait two hours in this and we'll call you the idea that you can walk everywhere in the world

[00:27:42] and talk to my kids, to my wife, see them in real time.

[00:27:46] That, for me, is personally fundamental change.

[00:27:49] But there are many.

[00:27:50] The e-commerce delivery, it's so damn convenient and so unsustainable.

[00:27:55] But it is what it is.

[00:27:56] Everybody can murder it.

[00:27:58] Lot of fun technologies.

[00:27:59] You know, everything that you get on your phone or the app makes life easier in some

[00:28:03] ways.

[00:28:04] Some ways too easy.

[00:28:05] In some ways, now some ways weird.

[00:28:08] When I look at some of the young students,

[00:28:10] who may sit in a restaurant, in a group of six or eight,

[00:28:14] and they don't talk to each other, they're all on the phone,

[00:28:16] it's just weird for me to see this.

[00:28:18] How people are wedded to their phone,

[00:28:20] and it's get very weird when people are losing personal comfort.

[00:28:24] And maybe it's just me old folks talking because I still value talking to people.

[00:28:30] You may have just answered our next question, but in your 49 years in this business, what's

[00:28:34] something that's just blown your mind?

[00:28:37] Just made you look at the world differently.

[00:28:39] My mind gets blown routinely when I visited my colleagues labs. For example, I just have a colleague, a oil thing, who does clothing that had basically

[00:28:49] the truss meat.

[00:28:50] The possibility are endless because, for example, they can do just like the Bose earphone noise

[00:28:56] cancellation.

[00:28:57] This kind of material can do noise cancellation.

[00:29:00] When it hears noise, the consumer is on the opposite wave, just like they're talking about taking all of the airplanes and all the internal stuff and airplane and pulling the material in.

[00:29:10] So you won't hear anything. You won't hear the engine ever. Let's build into the material.

[00:29:14] It's also built the material. He works with a military because it transmits stuff.

[00:29:18] So he worked with a Marine who jumped into cold water and see how the body reacts. It transmits all kinds of information. It's mind-blowing staff to me.

[00:29:25] I have a colleague who does works with breast cancer.

[00:29:29] To give you an idea, she works with

[00:29:30] Massachusetts General Hospital,

[00:29:32] considered the best hospital in the country.

[00:29:35] When you see a picture of a breast

[00:29:37] and telling you if in seven or eight years

[00:29:40] from which you're developing to a breast cancer or not,

[00:29:42] the best people in MGH, the best radiologists,

[00:29:45] are correct 8% of the time.

[00:29:47] She is not correct 100%, but she is correct 20% of the time.

[00:29:51] You see that it works.

[00:29:52] It's better, it's not 100%, it's not even close to 100%.

[00:29:56] She is an expert in AI.

[00:29:57] You start seeing how things are changing

[00:30:00] and how radiologists will start using it,

[00:30:04] as a tool, they don't use it now, but

[00:30:06] use it as a tool going forward and say, okay, maybe I should look again. Maybe I should look

[00:30:11] more closely. So their colleagues will do amazing stuff. And this is for me, one of the

[00:30:16] delight of being at MIT. Yeah. And just the quick hit that I got last week. And I got to go to the cave,

[00:30:25] which is so cool to be able to see different routing

[00:30:30] and different scenarios, just boom.

[00:30:33] It was really cool and you get to see this every day

[00:30:36] at the forefront of technology.

[00:30:38] And it was just fascinating.

[00:30:40] So thank you so much.

[00:30:42] It's been great talking to you

[00:30:44] and just hearing your insight

[00:30:45] on everything that's going on in the supply chain technology.

[00:30:48] AI, I love what you said about just being mindful

[00:30:51] and being ready for change and having that mindset.

[00:30:55] I think that that could help us all,

[00:30:57] not just in our work lives, but in our personal lives too.

[00:31:00] So thank you so much. It's been great.

[00:31:02] Thank you very much for having me.

[00:31:04] Bye bye.

[00:31:06] Thank you for joining us on this episode of the next level supply

[00:31:10] chain with GS1 US.

[00:31:12] If you enjoyed today's show, please subscribe to our feed

[00:31:14] and explore more great episodes wherever you get your podcasts.

[00:31:18] Don't forget to share and follow us on social media.

[00:31:20] Thanks again, and we'll see you next time.