Introduction: Prioritizing AI in Business

Today's talk will really be about like how to prioritize right within your

business so how many of you use AI right now in any form like okay great and how many of you find that it can do a lot of things but like it it's it requires massaging right like it's never like perfect right and like so how many of

you are just didn't said founders but how many of you are business owners entrepreneurs have a side hustle?

How many of you do anything like that? Okay, and how many of you are in a leadership position in a company at the moment? Okay, perfect.

And how many of you are like students? Curious. Okay, great.

So essentially, the goal of this talk, those of you who have experience in business, and those of you who work in companies or will, one of the things you'll see is that companies are very confused about where to prioritize with AI.

They're running around with lots of different ideas about what to do. And things are getting a little bit messy and things are not exactly working the way we want, right? Costs are going up, results are not clear, right?

So this talk is really about how to find the one number in your company that will help you figure out where you're going to focus, okay? It's going to be a very different perspective on how to look at business. Okay.

About the Speaker

So a little bit about me, and I promise this will be the most boring slide on the presentation. So I'm a computer guy. I was formerly at Deloitte, where I was the National Lead for the Natural Language Centered Processing.

We scaled that to 30 practices before ChatGPT. We scaled that to 30 practitioners. We grew a lot of contracts. We cut delivery time by half.

I've also architected the AI platform for a Forbes 30 under 30 startup, like over 10 million ARR, and Peter Norvig, director of research at Google and author of one of my favorite textbooks on AI, also endorsed my architecture.

Why the Right Metric Matters

Case Study: Blockbuster vs. Netflix

So let's look at an example of a company and why numbers are very important, okay? So the company we're talking about is Blockbuster, okay?

So they had about 800 million of revenue coming in from one metric. And this metric was how many people return their movies late, right? So I don't know if you guys have ever rented movies. You know, we live in the internet age.

But, you know, it's like a library. Like, sometimes you don't return movies on time. So they would penalize people on returning late. Now, this brought in a lot of revenue for them. them.

Netflix went to them at a particular point of time and said, like, you can buy us out for, I think it was approximately 50 million, okay? And Blockbuster said, no, why should we buy you out? You are like a rounding error, right, on our balance sheet.

But what happens 10 years later? Blockbuster is totally bankrupt and Netflix has taken over the market, right? So what happened? Like, how did a company that was bringing you this much money go bankrupt?

Well, my answer, my argument is very simple. It's because the metric they prioritized, right, which is pure revenue by any channel, was anti -consumer.

And when consumers found something that was more aligned with their interests that didn't penalize consumer behavior, consumers shifted platform, right? We see this again and again.

So all of your metrics have to be aligned with your customer, right? That's the most important thing. You don't want to have any other metrics.

So we want to try to identify one metric in your company that's going to be the most customer -aligned, okay? And that's going to create the most long -term recurring revenue for you. Okay, so let's kind of proceed with that.

So, yeah, so a great number going up is not the same as the right number, okay?

More Examples: Kodak (and Sears)

So Kodak invented, we know, the first digital camera. They didn't prioritize it because most of their revenue was coming from the non -digital. side, right?

Similar story with Sears. But let's look at, like, another, like, simple example. Let's

A Simple Analogy: Restaurants and Repeat Customers

take, like, a restaurant. Like, we've all eaten at restaurants.

If you ever talk to a restaurant owner about how their restaurant is doing, they'll give you a variety of answers, like, oh, like, Google reviews, or we're very busy on, like, Fridays, but on Tuesday afternoons we're not so busy, or one of my staff left. So it's like a confused haze of metrics, right?

But the most important metric for restaurant is actually how many consumers come back to your restaurant and how often they come back and if you talk to restaurant

owners most of them are not prioritizing this they're prior they are lost in a jungle of metrics okay I'll give you another example of this that I like a

Focus and First-Principles Thinking

Case Study: Southwest Airlines’ 10-Minute Turnaround

lot so yeah so this is Southwest Airlines okay so essentially they were were going to run out of money okay they have like at one point they had like four planes and they have like you know a hundred some dollars in the bank now obviously they can leverage debt when they have assets like planes but they were going to go bankrupt so what's the one metric that prioritize does anybody

know okay so what Southwest realized is because the only way to make their models successful like they had to have enough planes to take people back and and forth. The only way to do this is they had to have their planes land and leave in 10 minutes, okay, where most plane companies are taking an hour.

So how did they do that? They did the opposite of what a lot of the airlines industry does.

So they got rid of business class, they got rid of first class, they got rid of meals, they got rid of like all luxuries, okay, they got really good at baggage handling okay and then there are the only Airlines in the

United States that has made profit every single year like they have been profit like for 47 years straight okay no other Airlines has come close to this okay so

this is the power of engineering your business from first principles around like what is the core metric of your business okay so again very different idea than what a lot of people do and you know Steve Jobs, Bill Gates, Warren

The One-Word Secret: Focus

Buffett they've all been asked this question if you could describe the secret of your success in one word what would that word be anybody know the answer yeah so that word is focus okay so how do you focus right you need to

focus on the things that are going to drive long -term money into your business right you need to define this one number okay so that's kind of the main point

here so you want to first find the number then you want to protect it like you don't want that number to get worse and then you want to figure how you're

going to improve it okay so if for your business if you think about in the next 12 months what's the one number that would make the biggest difference right if you can start to identify that that's going to start to give you the pathway way forward for how you can grow your company whether you're the owner of the

business whether you work in a department of a company what is the most important metric in your department right and if you're looking to to be a leader in your business one day this is how you being bring clarity to a company

okay so I give you another point that's very important right so there's a term

Avoiding AI Sprawl: “Don’t Boil the Ocean”

in consulting that we use called don't boil the ocean right and what is what what like the full saying is like, don't boil the ocean for a handful of salt, right? You don't need to, you can boil like a cup full.

So how do you not boil the ocean, right?

So like, here's one of the downsides of AI. AI can do a lot of things. And it's very easy.

Those of you who program you can, you can create 10 pieces of software in a day, you can create when you when you used to use five different pieces of software, now you can use 100. Do you know the problem with that?

The The problem with that is you don't know if it's correct or not, and you don't know how to maintain it over the long term. So you can end up with way too much of this going on.

Why “Human in the Loop” Can Still Fail

And one of the biggest problems with AI is it doesn't know when it's wrong. So people say, well, we'll solve that, we'll put a human in the loop. The problem with that, we've done studies on this, this is from Amazon.

What happens over time is when the AI is mostly correct, and when it's wrong, it's not exactly wrong, it's sort of wrong. humans start rubber stamping that more and more and more so what happens is over time your approval rate goes to almost 100 % so all of the human in the loop reviewing the so -called AI processes are basically auto approving

everything so the solution to this is not more AI processes the solution to this is not automating the business department by department with a hundred KPIs with all sorts of fancy dashboards.

The solution to this is to focus and prioritize by finding one metric and designing the business around that. Okay. So it's a very different way to look at a business in conjunction with AI. This, I believe, is the solution to the problem of like, you know, AI slop, essentially, where we get mass like abundance of things that are not

How to Find Your One Number

Starting Point: A Business Audit

exactly right so how do you do this okay so one of the things that we do as an AI companies we go into companies and tell them what to do where they are right so we start with a business audit okay and whenever I do business audit mostly what

I find is founders and business owners they're usually not correct about how they should proceed they usually come up with all the things they want to improve in the business and they don't even work together there's no prioritization there's no sense of

complexity there's no sense of how much we'll have to get into the business and talk to all the people that interrupt things and how much old processes are tied into new things and how making a change could break something like they don't have a sense of complexity okay so if you're in

a company or if you're doing ai work for uh for people like figure out where you will focus like so the audit is to help you start to identify what are the crucial moving parts of the business and then how do we connect those dots in a way that we're going to be able to optimize for that

one metric right like how do you go downstream from that one metric again with southwest it's like our one metric is the plane has to be off the ground in 10 minutes and then what are the moving pieces, well, there's food, there's luxury seats, there's like all these other things that interrupt the speediness of that happening, right?

And how you then shift that, okay?

Reducing the Cost of AI: Grants and Funding (Canada/Ontario)

So here's another good news. If you're a business owner in Canada, there's actually government money that can help support AI transformation for your company.

Whether you're a big company or a small company, this can apply for everybody even a self -employed entrepreneur if you're incorporated you can get some of these funds okay so essentially there is money

for the audit itself there's about there's about 50 % of 15 ,000 the government will cover for this money for implementing at about 50 % of 150 ,000 these are in Ontario there's also R &D it's one of those programs is called

or tread you might have heard of it there are others but it can cover about 67 cents on the dollar for salaries including a salary you pay yourself if you're a business owner and essentially this is stuff that you could claim in the past two for the past 18 months so you might even be able to claim it for

the last year and then to be able to claim it you can claim it every year for R &D span technology span investments that you're doing in the company so you can significantly reduce your costs for doing all of this and you can usually

stack these there's also if you're in manufacturing there's a grant in Ontario for that there's clean technology grant if you're investing in new buildings things like this there's grants for training hires both new and old and that can be up to a hundred percent and that could include AI training for your staff

even in a big company so a lot of that's available and there's a national AI fund that's going to be rolling out okay so one of the things i encourage you to do is if you own a

business or if you're in a decision making capacity in the business and you want a free look at what grants you'd qualify for like just scan the qr code um i work with karen who is sitting right

there in the audience um uh karen if you want to stand up and say hi if that's possible um so So that's Karen, okay, she's really wonderful and she will be able to walk you through specifically what funds you'd qualify for.

And all the work that we do, like I'm more in the AI side, Karen's in the grant side, is totally like there's zero cost for you and it's purely success -based, right?

So if Karen is not able to get you any money, there is no cost to you. So this is a free way for you to get an assessment on what you'd qualify for in your business, okay?

Closing Message: A Number Going Up Isn’t Proof of Progress

So again, coming back to the main point, a number going up isn't proof that you're measuring the right thing, okay?

And one more thing, even if you're looking to start a business in the future, we can still tell you now how to structure a company, how to do any of that, so that you're well set up for the future.

So don't be shy, okay? I'll give you three more seconds if you want to screenshot it, and then I will move on, okay? So, okay, there's one person screenshotting. Okay, here you go.

okay so yeah this is my last slide so again to bring it back around a number going up isn't proof that you're measuring the right thing okay that's not proof you're measuring the right thing it's not proof of progress in your business for any of you that aren't businesses I hope you've seen some of

the metrics that we've tried to use with AI adoption that have not made a difference in the company most companies and so in in the fortune 500 are asking the Silicon Valley guys hey we've we're spending X dollars on tokens we're not really seeing concurrent return we're not seeing an adequate return that's because the prioritization has not been done properly okay so find the one

number in your business that's going to move the needle is it always not one number it's not always one number okay but typically it can be reduced to one one key number and you can, sorry, okay. And then you can organize your business around that.

Conclusion and Q&A

So thank you very much. And yeah, I hope you enjoyed the presentation. If you have a question or two, I can take it.