You’re Asking the WRONG Question About AI

How are you leveraging AI inside of your business?

Exactly the way that we propose to our clients to do this. So for us, what I found is, let me share a secret with you. AI is just a technology, right? AI is not a silver bullet. AI cannot fix everything. Let me be more specific. If you think about the world of operations is 2 distinctive descriptions. One is the machines, like the engines that work to get your company what it does. Like literally in a plant, there would be machines that, that make the parts or assemble them. And there is a coordination layer between those. So the coordination is how 10 machines On your factory floor work together, or more broadly, how you schedule those machines and get people to them and get materials to them and figure out what demand is and figure out where to put things into the warehouse and figure out how to sell, to sell the results. That is the coordination. So what I have observed is that the machines inside the department, The machines that are responsible for individual bits and pieces of work are actually pretty good. Most often, uh, we have made those machines work well. Exceptions are there, but, uh, for now, let me just defocus on this.

Okay.

What doesn't work very well is the communication connection, the, the, those escalations, those pesky things of, hey, did you look at this? Did you look at that? Is this plan aligned with that plan? That is a coordination layer to me. And that is where we are much less excellent than in working individual machines. Now, back to your question about AI. This is exactly where AI is good at. So AI, if you ask AI, can you run my machine that produces metal parts, just produces 20 parts a minute. Can it produce 40 parts? No, AI cannot do that. The machine is already pretty efficient. AI wouldn't be able to make this machine work faster because the speed is limited by physical laws of nature. It won't be able to make details with higher quality because it's already optimized. What AI can do is coordinate the things better. So, for example, react that hey, your demand has changed. You actually need to produce different part on this machine. Or you know what? You will have your demand is high and you plan to work over the weekend, but you are going to run out of the materials by Monday. So there is no rhyme and reason working this weekend because then you would you would be idle on Monday. These are the things that require a lot of. Like work, a lot of coordination, a lot of work that human beings are not very good at because it requires looking at 20, 30,000 different sources, making sense out of different data points in data that is not the same, and some of the data is wrong. So if you create a piece of program, uh, few lines, a few thousand lines of the code, Java code, you would have 20,000 edge cases and you would spend a year, 2 years, 3 years debugging this. AI is very good at those things where you can describe what good looks like and get a pretty good result fast in coordination in solving those tasks.

Would you say that AI makes the human better?

Absolutely. Being used in the right way. The way how I see it and the way how I've learned this in my last few years of consulting work is that for whatever reason, many people think about this as science fiction came true, silver bullet, artificial intelligence just can do things much better and will replace us. Best case scenario will be protected animals in the zoo. Worst case scenario will be extinct by the Skynet. In reality, artificial intelligence, at least in the foreseeable future, as far as I can tell, is a very good copilot. And copilot, don't take this anything about Microsoft. Microsoft used this word in a very nifty way. But this is a, this is the wingman. This is a goose. If you will. And AI can amplify your abilities. Of course, there needs to be the ability to amplify first. And the problem is if AI is left on its own, it's basically like it's working with the people, with the person who is missing zero intelligence. It amplifies zero intelligence. Guess what? Zero multiplied by 20 is still zero.

Yeah.

If you have a good person with the judgment, with the breadth of view, with the goals, with the emotions, with the desire to get things better, that is what you want to amplify. So AI on its own is a good engine, is a good tool. A person on their own is a person with limited capability, capacity, and hours per day that the person can concentrate on things. Together it makes a much better, much better unit, if you will, a duo. So people, people said it, I think, very eloquently a few months ago, maybe a few years. It's not that an AI would be competing with a person and would eat your lunch. It's really the person with AI. will compete successfully with the person without AI and make the person without AI irrelevant in many places.

I think that so many people were afraid that AI was going to come in and it was going to take everything away and the world was going to be run by robots or AI bots or whatever you want to call it. But the reality is, you said it a few moments ago in not these words, But it amplifies what the human's able to do and what the human's knowledge is and gives them things to make decisions that without AI they wouldn't have had. So I hear you saying it allows people to make better decisions with more information their first time, rather than having to make mistake after mistake after mistake to learn what AI was able to tell them.

Correct. And this is not just a dream. This is the evidence from the market that I'm observing right now. There are many different people with different ideas, and I see attempts to just reload this, reload the whole human jobs on AI. For example, in HR, I've seen this time and time again. The recruiters would, oh, the leaders above recruiters will try to remove recruiter from the equation altogether and have the candidates come to see AI and then AI advances them to the next stage and then AI maybe even makes a decision of whom to hire. It doesn't work. So far, the only impact that I have seen is that people are scrambling after a few months of trying this, scrambling to find good recruiters back. It doesn't work this way. Just the evidence is that the success of people hired this way is much lower than the success of the people that are hired by recruiters. And I would argue that the success of people hired by recruiters with the help of AI Is going to be higher. So it was people needed to experiment and go to the lower point to understand that it's not, it doesn't work this way. And I would, my experience, let me expand it to a broader field. I gave you an example where just I, I've seen that from my friends in this area. And I got some, some feeds of non-public information of how exactly it was unwinding. That's why I am judging with confidence that, that That's what I'm seeing. But if I generalize across the functions and maybe to some extent across the industries, what I've seen so far of successes is that when people are amplified by AI, that's a success. When somebody tries to replace people by AI, that's in majority of cases, if not all, it's a failure. So my prediction is that we won't see massive job losses because of AI. We'll see massive reformatting of the jobs. People tend to be not too good and actually don't like the jobs that AI is very best at. That is coordination, coordination kind of activities. So we will see people much more successful. We'll see the places where jobs will disappear is not because AI takes over. It's because the leaders are trying to flood, to basically buy the way out of their ineffective structures, ineffective operational networks. Buy the way out of that was just adding more people. And those people would disappear. Not because jobs disappear, but because these people weren't needed to begin with in those jobs.

Well, I think a lot of times companies look at a problem, or at least they used to, they look at a problem and even the workforce would say, we need more people, we need more people, we need more people, but the systemic problems were never fixed. And to go to your example about the recruiters, AI can't screen for culture. Because all they have is a piece of paper. And so they can screen for basic skills and they can screen for relevant history, but you're gonna miss out on really great cultural fitting people, which you need to grow your company. Yeah. If you rely solely on AI, which doesn't have that emotion, it doesn't have that feeling, and it certainly doesn't understand from a piece of paper what culture is going to thrive inside of a specific company or organization.

Furthermore, AI is pretty good at faking those pieces of paper. So when people start playing this game, they, on the recruiting side, the candidates happily continue playing this game on their side. And the outcome is that you get a bunch of irrelevant folks who don't know what to do with your business. You start failing. What works really well, and a friend of mine has been running this kind of business for a while, for about a couple of years by now, is say the checks, the background checks. That's where AI excels because that's where you ideally want to check so many points that if you assign a human being, it's going to be A, expensive, B, long time, and C, it's just not fun. It's the type of job that you need to type 20,000 different points in Excel. You don't want to do this.

Yes.

Good people don't want to do this, so they won't do this, but AI can and will. So what I've seen is that we had a conversation with the CEO of that company a few months ago, and he was excitedly telling me that, hey, do you know I found like 30% of the people who were applying for this role Or fake. And I found that my AI found that with evidence. That is where a good recruiter who can judge you by, by the cues, by the way how you speak, by what texts you and what doesn't. With AI who can actually check the real background references and all that stuff much more thorough than a person would like to. Together, that tandem gives you a great result.

And to your point, with AI creating these fake resumes and building all of these things up, it's so easy to apply to jobs and end up in this massive pool that's just being screened by yet another level of AI. I was looking on LinkedIn the other day and I'm A post had come up and it said over 1,000 applicants. And I thought, well, who in their right mind can go through 1,000 applicants in a way where they're really going to remember the difference between A and B? So, you know, I think that there's certainly some good case uses to help get to a manageable pool.

You want to shrink that pool to probably 10-ish candidates whom you want to review, whom you want to actually get a sense what these people are. And after that, you would probably have an understanding, hey, this is, this is somebody who, who is really best right now. And that is somebody who, who has best potential, but needs, needs input, development, investment. And your question, the human judgment would be, which one is better for me? And this is a conflict of priorities. This is where the AI would not be good at doing this or solving this. That's where you need a person. That's where you need a person, a recruiter who formulates that, who finds, discovers that, and communicates with the leader responsible owning this role. Hey, these are 2 cases that are good. Which one is better for you?

Mm-hmm.

And then we, we need to have judgment on that side. Uh, that would be really something that AI cannot do.