FIR 93: Applying Lessons From AI's Combat With COVID-19 !!

11:22
 
シェア
 

Manage episode 270850732 series 1410522
著作 Grant Larsen の情報はPlayer FM及びコミュニティによって発見されました。著作権は出版社によって所持されます。そして、番組のオーディオは、その出版社のサーバから直接にストリーミングされます。Player FMで購読ボタンをタップし、更新できて、または他のポッドキャストアプリにフィードのURLを貼り付けます。

We know that AI is being used in certain healthcare scenarios and impacting business supply chains. But, what are the REAL lessons, and do they apply to me and my business??

Hey, this is Grant thanks for joining another episode of ClickAI Radio. So I've been thinking about could there be some lessons learned from AI's use during COVID-19.

Could those lessons be applied to help our business? So to do that, I wanted to first start at what are some things that AI is doing to help combat COVID-19. I saw a report not long ago from the World Economic Forum, they were talking about some ways in which they've seen it applied. And it's areas that of course, if you're watching the news at all you're fully aware of, obviously, AI is being used in the healthcare industry, certainly with research labs and hospitals. They're using AI and machine learning, which is a component of AI in multiple ways. One of those ways that's interesting is in the use of chatbots. And sometimes we love to hate those things, right? Because we really, really just want to talk to a human, but the AI is informing these chat bots. And in one particular group, it's a startup company out of France. They're called Clevy.io.

Is the website, they are a French startup company. And they're based on the Amazon AI platform. Anyway, they launched a chatbot to make it easier for people to find official government communications about COVID-19. It's already handled over 3 million messages. So that's, that's pretty amazing, actually, when you think about the load that that took off of humans, right, in terms of having to manually do that, or the real question is, and what I don't know is how effective was that communication? Right? Did the AI really help?

When you think about with your own business think, okay, could I leverage a similar approach? Right, can I apply AI for my chatbot support? Alright, so that's one area. The other area, of course that we see AI being involved with is in the evaluation of the tons of data, right the the health care providers and the research they're, they're faced with this exponentially increased volume of information about COVID-19. And the reality is, AI thrives in that sort of environment where there's lots of data. It's difficult for us humans to derive insights from something like that. And so as a result, Amazon launched this search, it's actually really cool. It's called the cord 19. Search, CR d dash 19. Search. It's this website that's powered by machine learning that helps researchers to quickly and easily search for research papers and documents and answer questions like, you know, when is the salivary fire load highest for COVID-19? I don't even know what that means. But these researchers do, I think it's awesome that they can leverage AI and machine learning to do something like that. So you think, Hey, we got to have something like that for small to medium business owners, right?

We ought to have a place where you could go and be able to ask a lot of questions about running your business. And what it's doing is it's leveraging a bunch of documents, best practices of business. Right? That would allow you, huh? Remember that? I said that Okay. All right. So let's look at another way that AI is, is helping out during COVID. This has something more to do from a business process, a value chain issue, right? It's really to help avoid the disruption in the food supply chain. You know, some food processors and governments obviously need to understand the current state of agriculture, right? And what's happening in the food in the food chain. Lots of organizations and people are involved in solving that problem. So there's this agritech startup company called mantle mantel labs. They're also an Amazon customer.

So what they're doing is they're offering this AI driven crop monitoring solution, but what's cool about it is they're offering it to retailers free of charge free of charge for a period of time during COVID. You know, that's obviously to take some of the pain, the financial pain off, but also to deliver the capabilities in there as well. It's amazing how this technology works it, it assesses the satellite images of crops to flag potential issues to farmers and retailers early on, so you can get notification early on about any impacts to the supply chain. I was talking to a supply chain group, this is probably two months ago now. They were talking about how when the COVID stuff first hit, they provide supplies to some of the major restaurants that have global reach. They ended up shifting the use of their supply chain to help to help manage and store extra supplies like toilet paper and things things like that. Right which is amazing, right?

And so AI and mL, it's helping both in the healthcare and research labs. It's helping in supply chain, obviously, and in other areas also. So the question is, Hey, can we learn some lessons from this? Well, alright, let's look at another example here. The MIT Technology Review shared an experience that they saw that they took note of. They said, Hey, in the first 10 days, when the pandemic hit, the Amazon, top search hits dramatically changed, right? It started to of course, include things that people started buying in bulk that Normally we'd not been buying in bulk. So things like toilet paper sanitizer, masks and so forth. Right. And so the interesting part about this is where the AI models are in place, and are trying to optimize, oh, how can I make sure I get the right kinds of goods to people.

People are typically searching for phone case chargers and things like that. Those were like in the top, top range, suddenly all that is within a matter of days, gets just turned upside down. And now Hey, people looking for toilet paper and sanitizer and masks. And so the AI models actually had had not been prepared for that it had an effect on the AI models in place. Now, this is really an important lesson because the AI models had to be updated to reflect the change in behavior. Now, this isn't surprising to me, right? This is actually, this is actually the way in which we should deal with our AI models. Gartner, who's a leading research and technology research firm, advisory firm, they they shared a report was this is some time ago, four or five months ago, where they said, Look, there's many teams that are implementing AI, but not many that implement a monitoring solution.

In this report, they had this maturity model, and they talked about, how mature is a company if you use an AI this way, whereas you're in another level of maturity if your company another way, and the net net of it was that if you're doing some monitoring, and then making adjustments to your AI, you actually are in a better position to pivot your company. Right?

So that starts to become a critical aspect to creating what we call an evergreen profit for your business. See, the reality is, machine learning models are intended to support some level of flexibility, of course, but they need to be revisited and reshaped over time when it gets when a deviates too far from what the model knows to be, quote unquote, normal, right? Then of course, it needs to be modified. So here's the key thing, right? One of the best things you can do is you can incorporate frequent refinements to your AI models, whether your AI models are telling you how your customers are buying your products, or how you're providing service to them or what their refund levels are.

All of that changes over time because of things that are outside of their control, in this case, a recent pandemic, but it could be other things could be competitive pressure can be a shock to their own supply chain or your own supply chain, whatever it is. The key is to connect your business information to your AI models, right that that starts to become the key. And then to implement a monitoring process. The monitoring process keeps an eye on, Hey, how are these AI predictions going? And is the new data we're getting and the way that our customers are working with us? Is that changing our aim our AI models and predictions, so you got to be prepared to adjust the AI model so you're getting the best insights and the predictions. Okay, so in summary, what are the two lessons we can learn from AI's use during COVID 2019. There's two here.

First one is working with fewer resources at your disposal as a small to medium business owner, you want to leverage AI to give you the business insights and the predictions that you need to help you use your resources even better. That's lesson number one, right? COVID-19 has definitely reinforced the need for that.

Number two, lesson number two, set up the continual monitoring and modification to your AI models, right? Because you don't know when toilet paper and sanitizer metaphorically are going to jump in priority. And then you need to be able to pivot quickly as business, right?

Those unforeseen things, you want to have the AI be apprised of this and so put in place regular like monthly every other month monitoring of what's changing in your data and therefore what should your AI models be giving you additional insights and guidance on? Okay, that's it for now.

Thank you for joining and don't forget to subscribe to ClickAIRadio.com.

98 つのエピソード