Ep 13: How To Use AI + Data To Understand Behavior


Episode Categories:



Did Google just declare war on OpenAI? Well, some of the things they announced yesterday, you might think so. So my name is Jordan Wilson. I'm the host of everyday AI the Daily Livestream podcast and newsletter, helping everyday people like you and me keep up with what's going on in AI and how to actually use it. Very excited to have our guest today, carrie Sullivan, the founder and CEO of Culminate strategy group. Carrie, thank you for joining us.

Karrie [00:00:36]:

Thank you, Jordan. I'm happy to be here.

Google's New Magic Eraser

Jordan [00:00:38]:

All right, so let's just get straight into it. Let's do a quick rundown, and normally we bring you news from all over the AI world. Today, we're just focusing on Google. They had their big I O event yesterday, which is their yearly event, where they announce hardware and software updates. We're just going to be talking about the AI specific software updates.

So one of the first things it should be, I think, pretty cool to play with, at least something called Magic Eraser. So they announced this some big updates to Magic. Eraser essentially allows you to if you have a photo that's cut off, if you're trying to get a photo of your kid riding a bike and half the bike is cut off, you can essentially move it over. And it's going to use kind of technology called in painting to fill it all in. Carrie, what are your thoughts on this piece, magic Eraser?

Karrie [00:01:31]:

I love it. It's a lot of fun. And it sounds like it's going to be a lot easier than some of the other image generation and image fixing AI tools because hopefully it's a little better quality too, because you end up with weird hands and fingers and things like that when you start to extend photos.

Google Bard New Updates and Integrations

Jordan [00:01:52]:

Absolutely, yeah. You always got to count the fingers. Right, so the next big update out of the Google IO conference, bard. So their kind of language model got a big update called Palm. We'll get into that in a later show, but I just want to run down just some of the new things coming to Bard. So Bard is essentially, if you haven't used it, it is the Chat GPT for Google. So let's just run down so it's going to be directly integrated into Google Docs and Google Sheets coding. It's going to natively support 20 different programming languages.

You can export to what's called Replit or Google's collab. So you can take code and take it straight into a development kind of playground. There are plugins so announcing plugins, which I've been waiting hey, open AI, I've been paying for chat GPT premium for months, still don't have plugins. And Google is removing the waitlist on Bard. So essentially hundreds of countries throughout the world are going to be getting all these updates within a week or so. So many updates could do a whole show on this. But Carrie, what's your quick take on all these different updates to Bard.

Karrie [00:03:06]:

I love the idea that Google is starting to do some native integrations. I think it's a really smart way for them to start to really think about what that operating model looks like for them going forward. Because obviously AI and OpenAI and even Bard poses an existential threat to their core business, which is search. So I'll be really interested to see how they actually start to shift that operating model into something that's not quite search related specifically.

AI Watermarking: Necessary Caution or Potential Danger?

Jordan [00:03:40]:

Yeah, hot topic for another day. I've been saying this for months. Search as we know it is going to change. Yeah, it's going to be wild. So something else that's going to change completely is images. So in Bard, you can work directly with Adobe Firefly, which is their kind of AI image generating, but they're going to start watermarking AI content, at least for Photos, we know for sure. Carrie, how is that going to change at least public perception of art and AI art?

Karrie [00:04:13]:

That's a great question because you and I were talking a little bit before the session, because I think it's a good thing to start to do some watermarking. We still have to proceed with caution because just like we saw with social media, there's a whole lot that we don't know about AI and unintended consequences and how people might use it to manipulate. I was just reading a story from Matt Carmichael this morning from Ipsos, and he talked about the majority of people at about 70% that are really worried that they're going to see misinformation spread even faster online, that they won't know what's produced by humans versus AI. I think there probably needs to be some standards. I don't know what those standards are, and I don't know how to check if people are using AI to rip those watermarks out. It's a tough one.

AI-Generated Emails: The Future of Communication?

Jordan [00:05:15]:

Yeah. A lot of tough things to navigate. But hopefully something that's easy is this new help Me Write feature, which is essentially bringing chat GPT, but bringing Bard into your Gmail, where it can automatically read the context of an email and you say, Help me write this, and it helps. So it's almost reading your mind. But that's kind of where I wanted to get to. Carrie, with Culminate, because that's something you guys are kind of helping, using some machine learning and AI to kind of do the same thing. So tell me a little bit about what you all are doing at Culminate strategy.

Karrie [00:05:53]:

Thanks. First off, I'm not entirely sure that I'm I mean, you gotta you got to put a lot of trust in a company before they start writing your emails for you, and you got to trust them with your IP, so that's always a challenging thing. But that said, when it comes to emails, if you've got an AI that will help you write email for people and do some of that day to day communication, man, I'm all for it. I think it makes a ton of sense partially because when we write emails, the language that kind of reveals who we are as individuals isn't necessarily apparent. They're very task based and very kind of task focused, really.

So you look like a very results driven person when you're writing an email. For the most part. Other things though, are a little bit different. So we can use language from LinkedIn or resume or keynote speeches and things like that, anything with neutral language. And we can read that language using AI to do a very accurate, very detailed analysis of somebody's psychology using trait based psychology and language.

How Culminate Uses AI as an Assessment Tool

Jordan [00:07:05]:

Yeah. Can you say and again, I think for the most part a lot of Everyday AI's listeners and watchers are people just learning about the technology. So explain a little bit kind of in everyday terms how Culminate is using technology to help its clients.

Karrie [00:07:29]:

Thank you. Yeah, totally. One of the hardest things we have to do when we go through transformation and even growth, M and A, private equity acquisitions, things like that, is change and culture, the technology and design or things like that, even op models, that's pretty easy. Getting the people to change is always hard. That is something that will never, ever change. And it's not because people don't necessarily want to change, they just don't necessarily know how because not everybody's wired that way.

So what we do is use AI like radar, and we're looking for general thoughts and thought process across a population in a company to help those leaders understand how easy or how difficult it might be to either go. In by department or different groups or things like that to find those folks that are going to have a tougher time with change and figure out how to help them make those changes a little easier.

Jordan [00:08:33]:

Sure. Have you seen any shifts? So you've been doing this type of work for a long time. Have you seen any shifts just in the types of companies that you're working with? Maybe now a hunger to use more of this tech and to use more of this AI and to use more of this machine learning to give them insights into their business? Or do you think there's still some hesitation? Is there still some kind of convincing companies that, hey, this can help you grow?

Karrie [00:09:02]:

It depends. And it depends on the leadership team. It depends on how they think and whether it's something they want to use. But the reality is that HR departments have been using tools, assessment tools like Hogan and Disk and Myers, Briggs and others, for years and years and years. So it's not different. The only big difference is that we're not asking anybody to fill out a survey. I'm not having to go and do 1000 interviews across an organization over months.

So we can take that time to market or time to change, as I like to call it and cut it by months or years because we're able to grab big swaths of the organization faster, especially in big enterprise. And literally those kinds of processes would typically take months and years. So we're taking the time out of it, we're taking the pain out of it. So that's not really a hard sell, if you will. The hardest part is getting people's heads around how to do it a little differently. That's really it.

Transform Your Business Success Rate by 6x

Jordan [00:10:09]:

Speaking of this time to change, what do you think has been the biggest whether it's technology or a mindset shift, what's been one of the biggest things that's allowed that time to change? For a company to go through that digital transformation process, what's been one of the biggest things recently that's allowed that time to change to be much faster or much quicker?

Karrie [00:10:33]:

So one of the core principles that we use is it was actually a finding from McKinsey. They discovered using a similar process with a lot of their Fortune 100 clients. And you can read about it in Beyond Performance 2.0. It's really great source. And they found that you can actually increase the success rate of your transformation by six x if you focus on people behavior and mindset. And they're right. They just don't necessarily know how right they are because they're using a survey to gauge those things.

Just using the survey and paying attention to it makes it go faster. And they're saying by six x and it's more successful from an AI perspective, we can grab much bigger parts of an organization. We've scanned millions and millions of profiles across the professional social media, like a LinkedIn, and that includes in total about 30,000 companies. So we can in aggregate tell you what the culture of a company is in about an hour. That's it. I can tell you what they're talking about around the water cooler in general in about an hour.

Jordan [00:11:57]:

Wow, that's amazing to think. Right? Because I think even before some of these AI and machine learning advancements, before the Internet, that's something the time to change or even to figure out what's happening around the water cooler might take months or years. Right?

Karrie [00:12:14]:

Right. Absolutely. And it still gets watered down because people just generally don't want to change.

AI Enables Quick Business Growth for Startups

Jordan [00:12:21]:

That's true. Looking at that big shift and now looking at how much faster machine learning models are, how much more accessible AI is for small and medium sized businesses especially, I think in the past, maybe this was something that only enterprise companies might think of or think to take advantage of. So where do you see this going in the next couple of months, couple of years? Just with all these advancements that are coming. So how can companies look at using some of this data and some of these insights that you're talking about? How is this going to actually change or improve or, I don't know, could it get worse over the next couple of years. Can data come too quickly? What are your thoughts?

Karrie [00:13:05]:

For some companies, it will come too quickly. And that's the part that gets me a bit excited. Jordan? I love the idea that a very big company's competition isn't just another big company anymore. A big company's competition is a mid market company or even a startup run by a single person in their basement with no employees. Because AI allows for that ability to scale. And it allows you to scale quickly. As long as you're automating the right things, you can scale quickly and let the humans focus on the really important things that drive your business.

I will probably never have a back office. I've automated finance and accounting. That's part of my business model. I try to eat overhead for lunch. I don't want my clients to have to pay for overhead. So I try to cut it out of my business and make things go faster for them as much as I possibly can. So I don't want to charge them for those hours or that pretty big logo on a building. Instead, I want to be able to scale the business and scale for a client as fast as they need us to. And that's really the trick. Can a company leverage tools like AI to meet the market as fast as the market is changing its expectations of what they want from the company?

And I think the companies that will have a tendency to do what they've always done with automation and that's cut people cut cost cut to make the bottom line look better, they're going to lose. And it's the first time they're going to lose with that strategy because they're not going to be able to keep up with market expectation. They're just not.

Jordan [00:14:55]:

So one thing that you just mentioned there is scaling faster. And I think that there's very few people at the organizational level that can actually do that. I think a lot of times that's kind of a top down. But what would your advice be, Carrie? For the everyday person, maybe who is not the CEO or they're not the vice president at a Fortune 500 company. But given your experience of both what you're doing at Culminate and personally for the everyday person, how can they use AI technology or machine learning if they're not very tech savvy? What's a couple of practical things that people can just start doing this week to improve their output or just improve their work life?

Karrie [00:15:44]:

Love that question. It's actually a question that I ask my team every week, partially because I want them to know that I'm not going to replace them with AI. And I want them to be able to share their ideas with each. Other as far as how they're using different tools to either give them some time back and make their lives better or make their service to a client better. So that's just one thing I do as a leader every single week. What are you using, how are you using it? And what would you like to share with the team? So that's one thing you can do, even as a middle manager or an individual contributor. Ask the people around you what they're doing and share those ideas. Get used to collaborating and sharing those ideas with others because it spreads. You can get some time back and you can create a little bit more balance in your day. And that's probably from a big picture perspective. One of the things that I'd encourage, and I know it's not exactly answering your question, so I'm sorry.

Encouraging Collaboration and Innovation in the Workplace

Jordan [00:16:47]:

No. It's perfect.

Workplace Angst: Pandemic's Lasting Impact on Employees

Karrie [00:16:48]:

One of the things that we are seeing in the data is higher levels of angst. There's a lot of people that aren't feeling good at work the way that they see work. That's kind of one of the things that the Pandemic did for all of us. Whenever you go through crisis, you see the world around you differently, you solve problems differently. Any kind of adversity is an opportunity to solve problems and bigger problems and more complex problems in slightly different ways. If you think about it, we all collectively went through this adversity in the Pandemic and it's changed us all in different ways, but probably in similar ways.

But work hasn't changed. Companies aren't keeping up with how employees want to work and how they see work and how they value work. So when you see a lot of the return to the office, hardline hybrid things like that, where managers kind of feel the need to control the people around them, that's a leader problem. Return to work and full of virtual is a leader problem, not an employee problem, because it's not a productivity problem. Employees will find ways to be productive all day, every day, especially if they can focus their time on the people they care about, the places they care about, and the things in their lives that they care about outside of work.


Jordan [00:18:20]:

Yeah. Wow. Such great insights there, Carrie. All of a sudden I look up at the timer and Everyday AI is already coming to a close today. No. So good though. Such great insights. So we're going to have more about what Carrie has been talking about in our newsletter. So make sure you go to your Everydayai.com sign up for that. We're also giving away a year long subscription to the premium version of Chat GPT. Carrie, thank you again for joining us.

Karrie [00:18:54]:

Thank you for having me. I really appreciate it and this is fun.

Jordan [00:18:57]:

All right, so we hope thank you for listening, watching, reading the newsletter, and we hope to see you tomorrow and every day at Everyday AI. Thank you.

Gain Extra Insights With Our Newsletter

Sign up for our newsletter to get more in-depth content on AI