AI Readiness For The Whole Organization (Special Edition)

Donna Mitchell [00:00:00]:
Welcome to a special edition of pivoting to web3 with co host Jose Garcia, founder of Talk Coded. Today's timely conversation, AI readiness, the whole organization. Let's get busy.

Jose Garcia [00:00:18]:
Here we go. We have begun. Thank you everyone for coming. My name is Jose Garcia. I'm one of your hosts today. Chat is open, by the way. Feel free to leave in your questions or comments. Don't wait for the Q A to start.

Jose Garcia [00:00:35]:
We're taking your questions as we go. And my fellow house, Donna Mitchell, thank you so much for hosting this event with me. Donna. And we conceived of this. Com, this event a couple months ago. We were just chatting, feeling the breeze was late at night for me and not so late for you, but look.

Donna Mitchell [00:00:59]:
Where we are now. It's exciting. Thank you for having me.

Jose Garcia [00:01:04]:
Oh, thank you so much for coming.

Donna Mitchell [00:01:05]:
I learned a lot from you and I as well.

Jose Garcia [00:01:09]:
So we're going to give, we have to be charming for about two and a half minutes. People sign in. That's our job now. It's just to be charming. And you know, this is also kind of a question that I get. A lot is a question. But some people are like, would say, just go for a quick win. You know, get chat GPT, get Claude, just start running a, just run up a use case in an organization.

Jose Garcia [00:01:41]:
Other people will say, hold on a second. Before you start putting all your company's data through a chat GPT free, a free account, there's a couple of things you need to know. And a lot of companies, a lot of companies, they, they will kind of dive in head first and then a couple months later they encounter problems. So we're going to explore the alternative approach, which is AI readiness. And we're, Don and I were learning as we go. We're not the experts on this, but we have five experts which are going to be bringing on one at a time. Feel free to bring in any questions. I'm going to be kind of like focused on the speakers.

Jose Garcia [00:02:29]:
Donna's going to be monitoring the chat, keeping us on time. How are you? How do you feel about that, Donna?

Donna Mitchell [00:02:37]:
I'm doing well. I'm waiting for some hellos and some questions and bring them in open.

Jose Garcia [00:02:46]:
I think we're gonna like, we're gonna bring on our first speaker and.

Speaker C [00:02:51]:
I'm.

Jose Garcia [00:02:51]:
Gonna, you know this. A lot of people will say, hello, Peter, hello, Jesus, Hello, Mary, hello, Mr. Patterson and Bernie, I think we're going to. And I'm, I'm actually in Valencia, Spain. I'm really glad the Internet came back on just Time for this webinar and our next, our first speaker. And actually a lot of people will be saying, we're saying on LinkedIn we need people who kind of can reach out to regular people and educate them about AI and help them get on speed with literacy and readiness. And we're bringing him on. And now.

Jose Garcia [00:03:44]:
Hello, Jason. Thanks for coming.

Jason Padgett [00:03:46]:
Thank you, Jose. Thank you, Donna. It is more than an honor to be here today. What an exciting topic.

Jose Garcia [00:03:52]:
And I think I met Jason in an after party, what it was two months ago and then we started chatting. And I wonder, you do some really interesting stuff where you go into your local community and host AI literacy, AI readiness trainings. And the people who come to your events, they're a lot different from the people who normally you normally hear from on LinkedIn. I wonder if you could maybe introduce yourself a little bit and tell us about what you're doing.

Jason Padgett [00:04:31]:
Sure, I'd love to. Thank you. So my name is Jason Padgett. I'm located in West Lafayette, Indiana. Lafayette, Indiana, which is right next to Purdue University. So we do have a lot going on around the technology space. But I have watched enterprise really embrace AI at a much more rapid pace than the general public and small businesses, nonprofits. And so I have, I've created some literacy programs.

Jason Padgett [00:04:58]:
One of them was through a lifetime learning association and the majority of the people were 60 to 75 years old. We did an eight session course on just everyday AI. They were so inquisitive. It was so, it was just, it was really enthralling to see how these lifetime learners wanted to learn how they could leverage this as well as avoid the consequences of some of the doomsday stuff that we hear. And, and I would say to anyone who's on this today, like if I had three pieces of advice for you. One is be playing with these tools, be learning how to use these tools. The second is come to the after party because after parties are a great place to network. Jose really introduced me to it.

Jason Padgett [00:05:38]:
And then the third is give yourself a round of applause because this is a great topic to be exploring. You know, Microsoft just put out their new 2025 Jobs Index report and it was very different from the 2024 one. In the 2024 one, it showed that most of the use was shadow use among employees with enterprises, some spending hundreds of thousands of dollars just to do AI, to do AI. And only about 15% of those experiments working out very well. I think that not doing, just doing AI to do AI is what all of our speakers Today are probably going to talk about, because if you're not AI ready, which doesn't mean buying more software for your company, it means teaching your people how to ask questions and process problems in a different way. And if you probably already have plenty of people in your company who are just using their own ChatGPT account, as Jose mentioned at the beginning of this, that can be dangerous because they may be putting your company's IP in there. But I think even more important is when you shine some light on that shadow use, what you really get is an opportunity for all those people to start collaborating around these tools and to become the force multiplier for your organization. We're going to hear a lot about AI readiness, which has a lot to do with literacy, policy and projects.

Jason Padgett [00:07:03]:
You've got to get those three things off the ground. And finally, I would say the one thing that I didn't really understand when I first got into AI is that responsible AI is more about data governance than it is anything. You know, if we're, if we're just using an out of the box chat GPT, a large language model, then we're, the information we're pulling from is like everything from the Internet. So it could be what, what an expert like Donna had to say about Web3 and it could also be what some 14 year old who wrote an article on Reddit said about Web3. And it really doesn't know the difference. So when it comes to what data needs to inform the AI models that you're using, that can be extremely important. And I know that, I know it's a lot to take in, but it's also like we're on the cusp of a new frontier. Right? To me this is like, this is the ability to skill up everyone and to empower your employees to be able to be the best they can possibly be and make your organization shine.

Jason Padgett [00:08:03]:
So I think our first speaker today, after me, Cheryl, is going to talk about the her company Hero AI and how to do more with less and what it's like to really be in the startup space and leveraging AI. And I just, again, welcome all of you and look forward to hearing what everybody has to say today.

Jose Garcia [00:08:26]:
Thank you. By the way, everyone, feel free to put your, I know a lot of people have kind of joined us since I, I said this in the first minute.

Donna Mitchell [00:08:35]:
You got people from Saudi Arabia, Colorado, New York, we've got a house full.

Jose Garcia [00:08:41]:
And people from all over. I think it's like, feel free to put your questions in the chat. We're Going to answer them as we go. We don't, we're gonna, we don't really necessarily have a. Oh, you can also attend the after party where everyone's cameras and microphones are be live and you can actually talk to the speakers. And, and I wonder if. Jason, a lot of people when they're first starting out, they're just kind of like, they kind of stall out. They're like, just ask.

Jose Garcia [00:09:18]:
They'll just use an LLM, like a search engine and, and just use it that way. And there's like, no, you got to keep pushing and develop a use case for your business.

Jason Padgett [00:09:38]:
Yeah, I think it really goes back to design thinking. I'm know, identifying a problem or a workflow that is tedious to your company or cumbersome or that you would like to improve upon and saying, okay, what problem are we trying to solve and what information do we normally use to solve that problem? That's kind of the data that you have to have present in the process. But then I would encourage people ask these large language models like, this is what I want to do. This is what I'm thinking about doing. What do you suggest? It took me a while to build a muscle to just ask the AI what the best way to go about trying to solve the problems. I am. I think brainstorming with them on projects is probably one of the greatest values that I've gotten out of them.

Jose Garcia [00:10:20]:
All right, we're going to bring Jason back. Thank you so much. Don't go away, Jason, we'll bring you back. And bring on. Do you want to introduce Cheryl because you know her better than I do. Is that okay?

Cheryl Cunningham [00:10:45]:
Charlotte, was that question for Donna?

Donna Mitchell [00:10:48]:
Oh, was that question. Was that question for me?

Jose Garcia [00:10:51]:
Yeah. Do you mind introducing?

Donna Mitchell [00:10:53]:
I. Oh, yeah. I thought you were going to do all the introductions. Of course I know Cheryl. Cheryl, of course, is one of my guests on pivoting the web 3. But she has an outstanding background. She's the CEO of Hero AI revolutionizing the enterprise space and communications and real time personalized AI interpretation. She has a lot going on at a very high level and I'm going to let her talk about, about that very quickly.

Donna Mitchell [00:11:18]:
Thank you. Hi, Cheryl.

Cheryl Cunningham [00:11:21]:
Hi, Donna. Thank you for the introduction. So, you know, I've been an entrepreneur for 25 years and I've watched many companies come and go. Many friends of mine have lost many, you know, businesses over Covid, for example. And of course, you know, 90% of U.S. entrepreneurs actually fail. So that's why I started the mission to build HERO in The first place. And my mission for HERO was really to enable entrepreneurs to take their vision from zero to exit or zero to success using AI.

Cheryl Cunningham [00:12:01]:
So yes, I'm very excited about this space and the fact that 50% of the global population by 2027 is going to be an entrepreneur solopreneur, self employed. This is a very, very important space to talk about, you know, in regards to AI readiness and how you are going to leverage AI to enable yourself as a, you know, small one person team, three person team. How are you going to leverage that? The tools at your disposal right now to multiply yourself and become that extension of yourself out in the world.

Jose Garcia [00:12:45]:
Okay, I've got a couple of questions.

Cheryl Cunningham [00:12:49]:
Yeah, right.

Jose Garcia [00:12:50]:
Gerald, please. We're at the start. What are the biggest hurdles that you see companies making when they're starting this journey in AI?

Cheryl Cunningham [00:13:04]:
Yeah, I can definitely speak to that. After 25 years and seven businesses, you know, some of them failing. Business entrepreneurs have the resource problem. Every single one of us has a resource problem. Whether it's enough time, enough money, enough employees, enough customers, enough vendors. Right. We are always striving to find those resources that we need. So right now I think one of the most important factors to realize is that these resources that we used to have to take a lot of time and spend a lot of time trying to reach out in the world and attract the right people in contact, communicate with the right people now takes a fraction of a second if we leverage AI to be able to do it.

Cheryl Cunningham [00:13:59]:
So you know, and you hear this over and over again, if you don't leverage AI, then you are going to be one of those people that fall behind. And it's very true. For example, you know, our company is now building an agentic operating system in which, you know, all we have, we will have a full C suite, we will have a full team of people that we would have had to hire and it would have cost us $11 million to hire a team of 60 people. We're now reducing our employee count to no more than 10. And what used to cost $166 per task that I would have to pay an employee to achieve. And it may not be the results that you're looking for. Not that AI is going to give you the perfect results anyway, but costs right now, a 100% employee only company, $166 per task. $166 per task.

Cheryl Cunningham [00:15:10]:
What is it going to cost in the future when we launch this agentic operating platform? $0.02 per task. So I would say, you know, the challenge now is to figure out not just, you know, what tools are you going to use, but how are you going to use these tools in combination? How are you going to aggregate these tools to become that, that technology stack that you use on a daily basis to accomplish and execute your goals? So a lot of us right now, you know, I, I think from my own opinion is the greatest hurdle is not about learning AI, it's about how do we use it to align with your dreams, how do you build a full system, end to end, that actually serves all the tasks that you accomplish, that you need to accomplish today, and how do you use that to accomplish your, your final goal, your big dream, from zero to that success? I think you're on mute, Jose.

Jose Garcia [00:16:32]:
I've done it again. I've got an AI to tell me you've done it again, Jose. That's what you say. THE We've got some. I have some more questions I want to ask, Cheryl, but I've also got a lot of these questions in the chat.

Donna Mitchell [00:16:52]:
Yeah, there's quite a few here.

Cheryl Cunningham [00:16:54]:
Oh, please.

Jose Garcia [00:16:56]:
THE do you mind, Michelle? Do you mind if I give you one of the questions from the chat?

Cheryl Cunningham [00:17:02]:
Please. Happy. Happy to.

Jose Garcia [00:17:06]:
What are the top three best steps a small biz or a solopreneur to take when they're first getting started with AI?

Cheryl Cunningham [00:17:21]:
You know, there is so much noise out there. So much noise. I mean, if you just go down LinkedIn, there's a million tools and you can hardly keep up. So, you know, I'm going to steal a page from my chief AI officer. And what he basically did was to say, okay, you know, lean on the experts out there. Lean on what the majority of developers are using, the tools that they're using, right? To begin choosing. Because these developers, if they're impressed with a tool, then that's. That tells you something, right? If you're not an AI person and not technical person, go with what the experts are using and if they're impressed, lean on that.

Cheryl Cunningham [00:18:14]:
Because when you use those tools, you know that the developers who are so impressed with it will actually keep up with it. They'll maintain it, they'll support it. There'll be a community that's constantly, you know, answering your questions. So in choosing, you know, what tools to use, I would say lean on experts. And then number two is you really want to take a hard look at your mission or your vision and really think about how do you, you know, jot it down in a notebook or outline or use ChatGPT to help you figure this out, Right? Journey Map your vision from zero to where your exit needs to be or your, what your success looks like and tell ChatGPT or whatever tool you're using, whatever LLM you're using, how quickly do you want to get it done? And you know, do a little bit of a spot check on yourself. You know, do you have the resources necessary to get that done in the amount of time that you're looking for based on everything that you need to do from now till then? So, you know, if you're starting out as an a solopreneur with, you know, say $5,000, that may not get you very far based on, you know, what Chat GPT tells you you need to do. So then what if that doesn't get you so far, then you have to figure out, well, okay, great, I only have 5,000. I need to raise more money.

Cheryl Cunningham [00:19:55]:
The next step would be, you know, putting the proposal together in order to go raise funds. Right. So think very methodically. Use the AI LLMs right now to, to help you kind of plan and really drill down on exactly what you need to do first. And then once you figure out what you need to do, find the tools that you need to do every single step with and, and go lean on the experts to find the best tools to do it. Lucky for us, we're actually building out this whole system, so we're, we're actually going to be our own guinea pigs.

Jose Garcia [00:20:35]:
All right, we're going to hear more about that. Thank you so much, Cheryl.

Cheryl Cunningham [00:20:41]:
Thanks for the question.

Jose Garcia [00:20:42]:
And I know you've got a trip coming up, but we're going to be hearing more from you later. Yes, thanks for coming, Cheryl.

Donna Mitchell [00:20:50]:
Thank you, Cheryl.

Cheryl Cunningham [00:20:51]:
Thank you for having me.

Jose Garcia [00:20:54]:
I'm going to. We're bringing. Jason. You're back.

Jason Padgett [00:21:01]:
I'm back. Got my wheel spinning now.

Jose Garcia [00:21:08]:
Sorry about that.

Jason Padgett [00:21:10]:
You're good?

Jose Garcia [00:21:11]:
I mean, I'm fine. You know what? Although I can't hear, I put myself.

G. Kofi Annan [00:21:28]:
On mute for the car.

Jose Garcia [00:21:31]:
Jason, do you want to take a pick up where you left off earlier? We both had a question about how to use AI safely.

Jason Padgett [00:21:42]:
Yes, I actually have that, like highlighted in the, in the questions there from William Jackson, who, who said, we work with a lot of secure environments and I'm concerned that staff might use AI to enhance their reports, but some of these might be sensitive at minimum. You know what, that's an awesome question. It got me thinking about being a teenager, right? And there was like three different kinds of teenagers. There were the teenagers who had like the super strict parents who just waited Till they knew their parents weren't watching. Whether that was still during high school or in college, did go out and raise hell, right? There was the parents. There was a parents who would let all of us do whatever at their house. And we did, you know, we went there and. And again raised.

Jason Padgett [00:22:26]:
And then there were the parents who talked to their kids about responsibilities and encouraged, like this open, honest conversation and gave them some trust. And those seem to be the kids that best navigated both high school and life afterwards. And I'm not sure that AI implementation is all that much different. If you're telling your employees, no, absolutely not. I'm promising you some of them are going to do it. I'd be one of them, right? But they're going to do it without your guidance, without any regulations, without any rules. Whereas if you develop an internal policy and you educate everyone on what that internal policy is and you work with someone, you can even leverage a nearby university. Most of the universities are exploring AI readiness and AI security.

Jason Padgett [00:23:10]:
So you don't have to spend $10,000 to have McKinsey come in and write a safety policy for you, Right? But get a safety policy, get everyone on board, and then start encouraging them to utilize these tools in a safe way and create little sandboxes of their own. And they'll be just like those kids whose parents gave them a little trust. They'll take that trust and they'll build upon it. And your company will be stronger and so will your employees, and your IP will stay saf. I also want to comment really quick on what Cheryl had to say, because I totally agree with her. LM arena is a great website to go to where they compare all the current models. So a lot of different people have different use cases. Some people want to use video, they want to use image.

Jason Padgett [00:23:55]:
They were, you know, some of them are code writers. I would say Claude is hands down, right now the best for code. But if you. If you go to LM arena, you can see they do, like, blind tests like the Pepsi Coke challenge back in the day, where people test two models against a prompt, and then they go with the public vote on which one performed best. And that can be a great indicator of which model is performing the best.

Jose Garcia [00:24:19]:
I found that myself. Yeah, you got to toggle two models. Thank you very much, Jason. We're bringing you back for the group hug. Okay.

Jason Padgett [00:24:28]:
Thank you, Jose.

Jose Garcia [00:24:33]:
I almost kicked you out of the webinar.

Jason Padgett [00:24:37]:
That'd be a first.

Jose Garcia [00:24:40]:
No. There we go. Hello, Kofi.

G. Kofi Annan [00:24:45]:
Hello. Hello. Hello, Jose. Hello, Donna. Good to see you all again.

Donna Mitchell [00:24:49]:
Hi there.

Jose Garcia [00:24:50]:
And I think I've hosted you in two webinars now. And Kofi is our resident marketing and AI whiz. And a lot of people, I mean whatever industry you are, even in healthcare, there's a marketing component to it. And I wonder if you can introduce yourself, Kofi, very quickly.

G. Kofi Annan [00:25:16]:
Definitely. Sure. My name is Kofi Annan. My, the company, my company is called the Brand Sensei. And what we do is really help organizations understand how to best to use technologies and digital marketing to best engage the customers or their stakeholders and then really grow their business in an efficient and effective way. So not really getting overwhelmed by the technology, but having the technology serve the goal of the business. So that's. Yeah, that's.

G. Kofi Annan [00:25:48]:
We've been around for a while. We might. I've worked with a number of really big enterprises, but more recently in the past couple of years, I've been spending a lot more time both small to medium sized organizations to help them navigate this digital landscape that's been accelerated by AI.

Jose Garcia [00:26:09]:
It was Kofi. You introduced me to Donna.

G. Kofi Annan [00:26:12]:
Yes.

Jose Garcia [00:26:13]:
You knew that, Donna. That's how you.

Donna Mitchell [00:26:14]:
Yes. And it's the podcast that met Coffee and I together because he's very outstanding.

Jose Garcia [00:26:18]:
In branding and I'm a marketing person too. Now. I finally admitted it. Background.

G. Kofi Annan [00:26:27]:
Wow. That's. That's one of the, the, the things that, that at this stage of the game, everyone with the tools and AI, everyone is of some kind, some sort of marketing or creative person. So that's the good and good and bad things about that. But we could talk about that. But yeah, embrace it and.

Jose Garcia [00:26:48]:
You know the. I want to. We could start us off and about the, the beginning of the marketing journey. How could, how can they. The thing is, I know you, you, you use all the tools, you know all about the tools. People like me, okay. AI and marketing. And then you see hundreds of tools and it's just too much.

Jose Garcia [00:27:25]:
What advice do you have for people who are looking at hundreds of different solutions and they're like, oh, where do I start?

G. Kofi Annan [00:27:35]:
Yeah, no, that's a common theme that I hear among folks that are leading organizations, especially as an individual. Yes. Like even myself, I play with a lot of different tools. People could dabble depending on whether you're on Siri or. There are a lot of things that are available for free now. But as it relates, you know, based on kind of what we're talking about here as far as AI readiness, so what is the impact and the benefits to my company? And I think that's where a lot of people started getting confused and they don't know where to start. Some years ago we came up with a fairly simple model and a simple approach that we found help a lot of leaders think through, you know, which, which tools to focus on and when for their organization. And there's three stages there.

G. Kofi Annan [00:28:24]:
So the first stage there would be, you know, focusing on optimization. So how can, how can you use AI and AI tools to optimize what you're already doing internally? And the key, key term there is internally. So things like your planning, your scheduling, you know, things with Google Calendar, those kinds of things, Gemini, you know, applying those to content generation, workflow, some, some of the kind of internal workflows, communications, you know, a lot of things that we, we've seen even Microsoft incorporate and Google Tools incorporate in their workflow tools. So really starting there first at the optimization, focusing, optimization. So then you could go to amplify. So how can you use the tools and which tools should you use to amplify what you're doing externally with your customers, how you engaging with your customers and those things like segments. So how are you segmenting your customers? How are you getting to know them better, you know, any of those kinds of activities, which tools are best to use that. And again, sometimes it's the same tools, sometimes it's chatgpt, sometimes it's Gemini.

G. Kofi Annan [00:29:29]:
But again with, within the context we found helps people compartmentalize and focus and get the benefit. And then finally that last part is transformation. So once you've done, you've, you've been able to optimize what you're doing, you'll be able to amplify what you're doing with your customers. Then you're really able and ready to use those AI tools to do what everybody is really hoping to do and anxious about doing, which is transforming the business. So building new products, building new experiences, going deeper into analytics and predictive analytics. So I find those, those three stages helpful with all the companies and the groups that I work with to help them compartmentalize and really, really make sure that the AI choices are being made strategically and they're helping their business grow and they're, they're pacing themselves because you can't do all at once.

Donna Mitchell [00:30:25]:
I like to add something to what you were saying as well. Coffee and the previous guest, especially for the audience, when you get into the tools prior that you do want to get the mindset of your organization in looking at the technology, there are some preliminary things that would need to take place. We Won't get into them now. But more importantly, the culture of your organization, the leadership of your organization, and the mindset of your workers. There needs to be a balance between the technology, the organization, and the needs of people.

G. Kofi Annan [00:30:55]:
Yeah, totally agree. At the end of the day, as much as we can automate a large portion of what we're doing, and to some extent, that might be the goal, you still need people, you need to still need human input. And those humans need to know both how the technology is serving their particular function, but then they need to, at the end of the day, know how to communicate with the technology. So it goes into prompting, you know, all those kinds of things, which, you know, again, depending on your stage, you could dive deeper in, but the people definitely matter at this stage.

Jose Garcia [00:31:27]:
Hey, keep the questions coming. We. I know it may seem like we're not asking them, but we are keeping them recording. We're all gonna bring it to all of them. Don't get them all now. We're gonna get to them in the after party.

G. Kofi Annan [00:31:41]:
Actually, I like the. I saw a question by Nimal which was, why would anyone work on a tool which makes them redundant in an enterprise? And I know that's. That's again, another common barrier that I find for leaders in organizations as far as AI adoptions. Some leaders want to boil the ocean and want to automate everything. And then other leaders, you know, especially if you're talking about industries, regulated industries like healthcare, which I've spent a lot of time in, some leaders in those industries are more wait and see, and, you know, they're more tepid about that. But again, I think the. What Jason mentioned about the literacy, you know, how at the very least, having your. The people in your organization understand how the tools work within their role helps alleviate some of the stress around replacement, giving them some facility and some say into how those tools get brought into the organization as well, really, again, helps manage that.

G. Kofi Annan [00:32:47]:
In my opinion, it is not the goal of technology to replace everything that humans do. It is. It's the. My ultimate goal with a lot of the organizations that I work with is how can you form that hybrid work situation where the technology is working for you, you're directing it appropriately, and you're both growing with the technology without. Without being replaced. So it is. It is a strategic approach. And Nimal, we have some frameworks that we could, you know, share with you about how to do that, if you're thinking about doing that within your organization.

Jose Garcia [00:33:23]:
We have time for one more question. How are we doing for time?

Donna Mitchell [00:33:27]:
We're doing okay for time. But I really have a question for Dr. Amani or it could be for coffee as well. Anything else there that you wanted to add? Coffee?

G. Kofi Annan [00:33:38]:
No, not, not really. I mean it's, it's, it's, there's, I think using that framework that I mentioned, the OATS framework, optimizing, amplifying and transform helps, you know, like no matter the size of your organization, it's about, you know, staircasing towards the end goal, which is your organization goal. I think that really will help filter out the tools that you need to use and how.

Donna Mitchell [00:34:08]:
And knowing you, you were right on time, you're right on schedule. So thank you so much. So, Dr. Amani, I'd like to introduce you to everyone to Dr. Amani. Dr. Amani Alabad. She's the assistant professor, conversational AI expert and a general AI designer.

Donna Mitchell [00:34:25]:
She's bridging human and machine communications through next gen design. She's doing quite a lot. But I do have a question here and I'm wondering if this falls in your space or not. And I think it does. AI on healthcare and data that needs to be cleaned before we use it, or how is that working in medical or some other areas. Could you give us some insight on the importance of the data, what's happening in those sectors specifically? Because some of the questions that we do have that's coming in and those that I've had initially in the diagnostic diet in the doctor space, the physicians, how is that being received? I think that's really the question from your perspective being, being Dr. Amani, can you give us some insight please?

Dr. Amani Alabed [00:35:11]:
All right, thank you so much for having me, Donna, and thank you for that. That's actually a very interesting question and it does actually touch on my area in that, that sense. So my area is not healthcare primarily, but we do also talk about data readiness. And how is it that we would need to understand if we were to deploy AI systems which basically rely and they are the data dependent, how is it that we would need to structure the data in order to receive good outputs. So in that sense, if we look at organizations employing AI systems with the proper training on good data, then you would see that the AI integration would be to some extent successful. But then again, as I mentioned, they are fundamentally data dependent. So that's why if we were say that an AI system is effective or not effective, that is tied to the quality, the availability and how the data is basically governed. So one of the examples I wanted to share for today is IBM's Watson's for oncology and was an AI, supposedly an AI support tool where you would be able to understand, basically it was supposed to be like that, but it was able to help prescriptions and it was there for cancer treatment.

Dr. Amani Alabed [00:36:44]:
So the promise that was given during that time was that it's there to eradicate cancer, which was a very big promise. And it failed miserably despite the huge investment that this tool was having. So the reason for that is because the data that was this tool was trained on was insufficient. It was trained mostly on synthetic data, on hypothetical scenarios that do not reflect the actual data that we have and in the world basically. So the data was not that great. And it also relied on some of the trainers or some of some of the doctors preferences. So in that sense the data that was received to the AI system and the way that the output was delivered, it was mostly reflective of these preferences of the doctors which were very, it was a very small number of doctors. So in that sense it was not accurate.

Dr. Amani Alabed [00:37:46]:
And that's why it was leading to inaccurate prescriptions, wrong recommendations. And so. And it was to some extent posing a risk because it was giving very wrong medication to people who suffer some kind of the side effects of these medications. So in terms of wasn't, that's why we say that if you were to really employ AI as part of a critical workflow or automating a workflow, a critical workflow in your space, you have to make sure that you have adequate data readiness and you just don't just go for whatever is there. And that's why we would need to be able to say that we don't want to just hop on the trend and just adopt AI just for the sake of doing it. Because that would lead to very severe consequences. On the other hand, if we were to look at what Kotu was mentioning, basically the AI integrations that were making some sense are the ones that were faced. So there was some kind of phased adoption.

Dr. Amani Alabed [00:38:55]:
The data was not that major. I would say it was not changing fundamentally. How is it that we would do our businesses or how is it that we would do our work? And that's why it would say here that the system that we are employing should have a deep understanding of the workflow. So if you think here of IBM, you would see here that it didn't have quite a good understanding of what is it that the clinicians would usually have. So prescriptions would happen on a case to case basis. And this is something that is very complex for a system to also understand. So the data was unstructured and that's how it's reflected in the output, I would say.

Donna Mitchell [00:39:38]:
Oh, thank you. Jose, did you have a question? I didn't mean to jump in there. Right there, but I had to get it while I could. Nice meeting you. Can't hear you.

Jose Garcia [00:39:55]:
Oh, you can't hear me?

Dr. Amani Alabed [00:39:56]:
No, we can't.

Speaker C [00:39:57]:
And now.

Jose Garcia [00:39:59]:
Sorry.

G. Kofi Annan [00:40:00]:
Hello.

Jose Garcia [00:40:01]:
The what could speak to the kinds of problems people are having. I know we spoke before this in past that so that they've started. A lot of organizations failed. What kind of like cautionary tales and pitfalls can you warn them about?

Dr. Amani Alabed [00:40:26]:
Yeah, so we talked with Donna, we just touched now on the data deficiencies, which is a very, very important pitfall, I would say basically trying to understand having the companies understand how the AI would be managing the entire data life cycle and making sure that they have a robust data management practice and also a data strategy. Which ties us back or brings us back to the very main point is that a lot of companies would not have a strategic vision or leadership commitment, which is something that is very important. Whenever you are saying that we would want to introduce AI in our organizations. So you would see here a lot of companies focusing on integrating AI literally in everything and just not focusing on the key operations that would make a difference. So say we've seen here thousands of use cases of AI. Some of them would be, for example, generating social media posts, creating this kind of emails or so on. So some of the things that are, maybe, I don't know, scheduling a meeting which would be good for you as an organization. But these are not core practices or these are not core applications that would drive revenue.

Dr. Amani Alabed [00:41:48]:
They were, they're not there to have a high return on investment. So in that sense, whenever we are focusing on AI integrations that are not tied into strategic goals, we would see here that people will not see the return on investment properly. So that's one. So if we were safe, for example, if we were to move, how is it that we would integrate AI into core business practices? Say for example, using AI in marketing, but instead of us just trying to automate social media captions and so on, we're using it to, for example to score leads. Say this is a sales qualified lead or this is a lead that would really bring me money to my business. And in that sense we are trying here to see the value, we're bringing in more money because we're using the capabilities of AI into basically a good use. And that's basically one. So we need to at least identify A couple of use cases, and I did mention it before, and that's a report, I mentioned it to you with the Boston Consulting Group, saying that businesses that don't see a lot of return on investments are usually spreading themselves way too thin.

Dr. Amani Alabed [00:43:03]:
So they're applying AI into 6.1 use cases rather than 3.5 main use cases. That would be driving the profit. And usually when we're thinking of AI integration, we're thinking of how is it that we are always throwing it on the IT department and we're looking at data that is being siloed, different departments and so on. So we're just not working collaboratively in an environment and a company to be able to say, well, this is how we would be leveraging AI properly. So we need to make sure that there has to be leadership in that sense that is driving this force rather than just throwing it on the employees and just asking them just to deal with it. So it has to come from the top. And then they would be leading the change throughout.

Donna Mitchell [00:43:53]:
So. Dr. Amani, can I ask you a quick question? I'm sorry, Jose, I gotta jump in. If you have a nonprofit organization, it may not fall in your camp. It may be in Brian's camp a little bit more. But if you have a nonprofit organization, they got agencies, they're siloed. There's data everywhere. It's old school.

Donna Mitchell [00:44:11]:
Is there an application or is there. What are your recommendations? Let's not even go there so we can keep it short and sweet. What would you recommend to that organization that's more on the behavioral side or in the mitigation of risk and more in human services? Do you have any suggestions or how do you see AI playing out there? Possibly.

Dr. Amani Alabed [00:44:34]:
Just thinking of it right now, I would say if it's more on behavioral data, we would just focus basically on AI applications that would help us understand what makes people move. So here I would be thinking of maybe applying AI to understand the behavior of these end users interacting with this nonprofit organization and trying to understand maybe their preferences, their behaviors and so on. Or if seen a lot of applications of AI having synthetic data, at least to some certain extent. So you don't want here to say that you want to apply again, AI into everything, at least just identify what is it that's driving your nonprofit organization to capture, as I said, proper leads, maybe, or getting you these clients or these people that are willing to engage with your business and then just really use the capabilities of AI in that specific area.

Donna Mitchell [00:45:34]:
That was very helpful. Thank you. Thank. Thank you. Thank you very much. Thank you. Thank you. I'll save the rest for Brian.

Jose Garcia [00:45:46]:
Brittany.

Jason Padgett [00:45:47]:
Brian.

Jose Garcia [00:45:47]:
Hello.

Donna Mitchell [00:45:49]:
Hey, Brian.

Jose Garcia [00:45:50]:
Hi.

Speaker C [00:45:51]:
How are you? I'm unmuted now, right? Yeah. Okay. Well, you know, I think we've covered a lot of topics today and I've seen some people, you know, popping some questions in Q and A, which are some really deep philosophical questions and some things that will be, you know, needing a greater discussion. I encourage people that are interested in those kind of questions to stick around for the after party because that's where we see a lot of those discussions happening. And as, as Jason said, you know, he joined one of our webinars a while ago, we met him through an after party and I remember myself, I was like really impressed by a question he asked and reached out so we'd start a conversation. I think that's a great way to get involved in this space. There's no bad questions when it comes to thinking about AI readiness and thinking about what I think draws the larger discussion together, which is what I would say is AI governance. So if you think through the kind of, just the, you know, progression of topics that we've seen today, we started with Cheryl talking about the real needs for small businesses, entrepreneurs to kind of utilize AI to one, create, you know, value that would take much longer to create in the kind of normal circumstances of how business does start up as well.

Speaker C [00:47:15]:
It allows these businesses to innovate quicker and get to scale quicker and, you know, most likely develop a competitive positioning relative to other startups in their same vertical or space that they're competing in. So I think that's one set of messages that we have around AI. However, for big companies, enterprises that have already been struggling with, you know, other integrations, technology integrations, they have immense tech debt already, right. And are barely keeping up with the workflow demands of new technology integrations. People have a slightly different take on how they can embrace and incorporate AI. We're talking about integrating either third party tools or tools that are developed in house within these larger organizations. We still come back to the same questions of one, AI readiness, how do you know? And some of the questions that people brought up around data, right? If your data is not ready, you're not ready for AI. So we can't ignore some of the more common challenges of technological innovation, digital innovation, which go back to data pre preparation, ensuring that you have data that already supports some of the core business solutions and business challenges that you need to face.

Speaker C [00:48:42]:
I think Jason brought up some interesting questions or answered some questions about how do we evaluate the various tools out there and gave the example of LLM arena as one tool that does that. There's lots of tools out there that have leaderboards that you can check out daily. They're updated all the time when a new model comes out. What I encourage people to pay attention to though is if the data aligns with the specific use case that you're looking at for your industry. So some of these are more generic, some of them are tailored to financial tech, some are tailored to healthcare, some are tailored for customer service. And you really just need to kind of scroll down into the data and kind of understand what's being evaluated. The most common leaderboards are only talking about the accuracy of the models. And if you've been following this for a while, you see that most are within the same 5% range of accuracy.

Speaker C [00:49:41]:
Accuracy is almost consistent across models right now. So if you're comparing models, you want to look at other features and usually that's when we get to use case specific things. We talk about the role of, you know, kind of what other data points might be pulled into and then what risk might be opened up. So, you know, I think Dr. Amani brought up an interesting question here too, or, you know, situation from IBM history and history, early history with WatsonX and one of the interesting things about IBM and Watson now is IBM made the decision to not scrap Watson entirely from the failure in the healthcare arena. But they re envisioned what IBM, Watson could be and they pivoted to AI governance. So it's actually one of the tools out there that is platform agnostic, it is model agnostic. And it is a way of kind of looking at any workflows of AI LLMs that you have that you're developing and kind of puts in place these kind of governance elements for the automated governance tooling.

Speaker C [00:50:58]:
That's important as we move forward in developing AI. Now, is it something that, you know, small businesses can afford? No, it follows that same kind of model of IBM, but it just also shows us kind of instructively what we need to be focusing on, which is AI governance to assure that we're aware of the risk that we're exposing our businesses to managing those risks safely and, and innovating and evaluating for the future. Again, I think, you know, Dr. Ramani brought up another interesting point of businesses spreading themselves too thin and therefore not always realizing the value from AI. I agree. And I think that, you know, and I think Jason touched on this too and Kofi by saying, you know, you can't just do AI for AI sake, right? You do need to think about specific use cases. What are you trying to do out the gate? Is it customer service oriented? Is it some other kind of way of utilizing data that you haven't used before and pushing it forward in more predictive AI fashion while you're developing a generative AI application? So there's lots of ways that we can still be innovative, move forward with AI while doing it safely and doing it with a governance first perspective. I think there are some other themes that we can pull through here from our conversation today.

Speaker C [00:52:35]:
Just dropping back and looking that big picture. One, you need stakeholder involvement across your organization for AI to be successful. It can't be a top down. My CEO wants AI. Let's scramble to do it as quickly as possible and hope for the best, right? It's not going to work. We know that you need to develop it with a use case specific focus, but you also need to bring those stakeholders in very early across your business. You need business analysts, right? Not just developers, not just the engineers. You need data scientists, you need marketing, right? You need someone that represents your end point customer, right? So in healthcare, that's patients or patient advocacy groups or someone that represents that viewpoint.

Speaker C [00:53:27]:
It also, if you're targeting healthcare providers, you might need to bring in various healthcare providers, doctors, whatever, as you're kind of envisioning what these solutions could look like. So you need stakeholder involvement early and often. That's a critical part of the AI governance process. You also need to think about how are you integrating AI tools within the workflow, the existing workflows, right? If you haven't done that part of your analysis during this AI governance and frankly readiness assessment process, right? You're assessing the readiness of the whole organization. If you haven't done those pieces and you just plop AA down into a specific workflow where you haven't done the research and involve the stakeholders, it might fail. Right? And you're not getting the ROI that you plan for because you've just inserted it in the wrong place. Right? And you know, maybe this is an ERP solution where you're looking at, you know, trying to improve upon screening processes and tooling so that you have HR involvement more efficient, you're still going to need two points in that workflow, most likely where you have that AI solution integrated, maybe even three. Right? So it's not just going to be at the external recruiter point in time.

Speaker C [00:54:43]:
You'll need the HR involvement, you'll need your finance team involved and thinking through what budgets that do people have to support the workers that are being screened for Hired, etc. So there's plenty of considerations that involve thinking through the integration and planning it effectively. And then finally I think a theme that's come through here is really about change management, right? This is why AI appears frightening to people. People are like, it's going to take my job. I really don't think so. Some jobs perhaps, some like, you know, and they're jobs that people don't want, frankly. Who wants to just read contracts every day and compare version to version and do that kind of thing? Lawyers don't want to do that. They want to do higher level work.

Speaker C [00:55:28]:
Right. And so there's plenty of what we call back office automation that could be done more efficiently through AI and free up the resources for people, the humans, to be doing something more valuable for the organization and probably more exciting for their career. There's one example of this that I think I just was thinking about the other day and that's around LLM ops. So LLM operations, separating that from AI operations. And you know, one of the things we don't talk about enough is the intersection between AI and cybersecurity. Right? We think about the risk involved in AI, but when we talk about bringing that together and thinking about how do we make sure that we're using AI and not opening ourselves up to new risk that we can't manage. This is an important point of intersection and I think it points to the need for at least medium sized and larger businesses to have a team focused on LLL operations. So for example, I hadn't thought of this before, but everyone develops their own prompts when they're using generative AI within a workplace setting.

Speaker C [00:56:36]:
Those prompts actually need to be treated like you would a software product and they need to be retired maybe after a two week cycle because if you don't, it's really opening up your threat matrix to potential problems if you're not retiring your prompts. And the prompts degrade in performance over time. We already know that model degradation of performance is a challenge. So you know, just thinking about that, which I know is kind of like a big, you know, down the, down the cascading flow of people like oh my gosh, I wasn't even thinking about something like that kind of problem. But it's a very practical problem that I think points to the need for new jobs. Right. People would be doing focused work on things that we hadn't even thought about the need for yet. Certainly a company does those kinds of things on the cyber security level, but they're not doing them on the AI level yet.

Speaker C [00:57:33]:
And I think that's an important thing thought process to think about. You know, it's something that I'll have to think about for my other business as we're scaling up is how we, you know, make sure we have enough humans to kind of do that important function as we're treating AI solutioning like you would a software product, right. And that's. It's early days and everyone's kind of experimenting with different ways of handling the level of risk and things like that. So for those that are kind of more on the panicky side of oh my gosh, it's going to take my job or feeling overwhelmed by learning about AI, everyone's in the same spot, right? And I think it's early days, we're learning lessons, you know, Cheryl, I think importantly said there's, there's ways to approach us, to think about. We're freeing up time, we're freeing up resources. And if we do this with an AI governance first kind of mentality and a readiness assessment mentality, I think we're able to position our businesses for success in a way that we can't do if we're not doing it from this kind of planful, strategic way.

Donna Mitchell [00:58:47]:
Brian, I really like that because at the end of the day, it's really about enhancing the organization, enhancing the human being, enhancing as we go forward. And it's really not a threat. You do have the good, bad and ugly, no matter what we do in the world today. But it's the enhancement of the productivity, the efficiency, the minimizing of costs, making sure the deliveries are. They should be on supply chains. But you do have to take your time, take your moment and do the analysis of the organization.

Jose Garcia [00:59:18]:
Thanks, everybody. And by the way, when the web ends, there's an option to go to a questionnaire that will take you to a list a of resources from all the speakers. There's also a link to the afterparty in that page. We've got. We're wrapping up. Can we have a final word, Jason?

Jason Padgett [00:59:42]:
Me? Yeah, I was actually just thinking that. Yeah, thank you. Excellent information from everyone. I was actually kind of relating this to one of my favorite AI scientists, which is Dario Amade, who is the lead CEO at Anthropic, the makers of Claude and Dario writes a lot of papers on this subject and he seems to be less hyped and more concerned with safety and alignment. And he just wrote a paper about interpretability and the fact that AI scientists, before these things become too powerful, need to understand what's going on underneath the hood. And I think that relates to business as well. Like we're such a consumer society before we start integrating and releasing agents in our businesses, we need to be utilizing this technology so that we can at least interpret how it works at a basic level as it begins to work at a more advanced level.

Jose Garcia [01:00:31]:
Thank you, Cheryl. Final word.

Cheryl Cunningham [01:00:35]:
The future is not going to be built by those who simply have ideas. Right? You, you can't just chase your vision without a head. You gotta systemize it, you gotta purposely engineer it and then you gotta move, make it move even as you're sleeping. That's, that's the future. AI is going to work as you're sleeping. So don't try to run your business without, you know, knowing the direction that you're headed in. And realize that you have to have everyone's buy in, not just the executives, but the colleague that's sitting next to you. And so when you make your decisions and choose the type of system you know, you want to make sure to choose a system in which everyone's going to benefit from whatever technology it is that you want to bring on.

Jose Garcia [01:01:36]:
Healthy.

G. Kofi Annan [01:01:40]:
There we go. Don't be scared of technology. This is not, there's really not anything new. The implementation is new. And really at the end of the day, it's about building up your organization and helping you reach your goals and choose the right technologies for those. Forget about everything else.

Jose Garcia [01:02:02]:
Thank you, Amani.

Dr. Amani Alabed [01:02:06]:
Right. Invest in your skill force. I would say that invest in the people that are, you are expecting to use AI because if you do not do that, your investment will not be worthy. If you're expecting people to just use a tool that they're not trained on, as Jason was saying, they will play around and they will mess with it. So make sure you invest in your people.

Jose Garcia [01:02:37]:
The after party now. Thanks everyone for coming. Thanks to all the speakers. And so you've got the link there, it's on the LinkedIn event page or you can go to the questionnaire and it will take. You have a list of all the resources from the speakers. There's also a link to the after party there. Thank you very much. See you in the after party.

Donna Mitchell [01:02:59]:
Thank you everybody.

Jose Garcia [01:03:00]:
Oh, by the way, you don't have to have your camera and microphone on if you don't want to in the after party, you don't have to have your camera microphone. It's your choice. Take care.

Speaker C [01:03:09]:
Thanks everyone.

Donna Mitchell [01:03:10]:
Thank you everyone. Thanks for tuning in to pivoting to Web3 podcast if you are a developer, innovator or AI expert expert with insights to share or if you're looking to partner, let's Connect. Visit MitchellUniversalNetwork.com and be part of the conversation. Want more content? Check out my playlist for more episodes. See you next time.