444 | Are We Safe With The Latest Ai? Hear Alexander De Ridder's Take!

Mar 8

[00:00:00] Alexander De Ridder : The internet itself is Skynet. The rate of progress in every field is going to hundredfold, thousandfold, thanks to AI. Our brains are not ready to understand or accept that reality. AI is not a fad. It is here improving every business aspect and there's the direct consequences that are immediately obvious.

Those do not rely on any new inventions.

[00:00:27] Chaz Wolfe: What's up everybody? I'm Chaz Wolf gathering the King's Podcast. Coming back to you here today with another king on the stage. My brother, Alexander Ritter. How we doing?

[00:00:38] Alexander De Ridder : I am doing awesome.

[00:00:40] Chaz Wolfe: You know, I'm, I'm glad that you're here. We've got. Yeah, I think I've only had really one other king come in full blown history with ai. And so you've done so much incredible things in this field and I cannot wait to pick your brain. So tell us what kind of business, uh, are you doing right now, Alexander?

[00:01:00] Alexander De Ridder : well, I'm building, an AI orchestration platform to build AI agents. Uh, we think that, you know, just. Talking to a, to a chat that gives you some text back is like, um, it's like, it's fun, but when it can do actual work, it becomes useful. And so building agents is not easy. Uh, we made it as easy as possible.

It becomes drag and drop, no code. Just bring your data, bring your workflow, bring your APIs. By the end of 2024, you can imagine your company having agents at every level. Helping your team, um, do a hundred times more

[00:01:36] Chaz Wolfe: alexander, what I'm hearing you say is that you're building the I robots. Are you and Will Smith over there making this thing happen?

[00:01:45] Alexander De Ridder : You know, the, this is an interesting topic, so AI and safety is like real important, right? And so if you, if you had like this really smart AI say like Marvel's ultra. And you just like unleash it on your business and say, go to town. Like, go and do work. It's too much power concentrated in like one ai. So what we instead do is, uh, we call this const concept constrained alignment.

Uh, alignment. The term means making sure the AI does what you want it to do that's aligned with your values, aligned with your business, and so forth. Right? Well, constrained alignment is about. Uh, making sure that AI can only do what you allow it to do in every step. So in, in aggregate, it can only execute your workflow that you've already approved.

So you constrain it, um, you constrain it by the workflow and you constrain it by, um, only using the best AI necessary for each individual step. So the joke I make around that is don't mow your lawn with a bazooka. The grass will be gone. But maybe the ground too.

[00:02:56] Chaz Wolfe: Yeah. Yeah, exactly. Okay, well so this is super interesting and I'm sure some people listening here today have thought about how AI in the next year, five years, 10 years, is going to rapidly change their business. So let's, let's talk about first maybe the difference between current automation and what you are talking about.

'cause when you talk about a workflow, I think there's still people here today, probably many people listening who are currently doing a workflow manually. Right? That could be. Put on some sort of automation with a tool that's out there today. But what you're talking about is something next level from even just an automation.

Uh, you know, I'm gonna create a flow and an automatic email and an automatic text message and a reminder's gonna come up. You're talking about kind of, maybe, maybe even like 10 steps above that. Give us just a little bit more juice here.

[00:03:40] Alexander De Ridder : When you learn to program and you go to school to become a programmer or yeah, computer science, whatever, um, your brain rewires itself, you think differently. Um, and it's very interesting because, um, just yesterday, mark Zuckerberg announced that they were dropping a new version of Go Code Lama, which is their.

Um, code generation, fine tuned AI model. Uh, they're open sourcing it, and then he said, and we're making it part of our next version of Lama, Lama three, because as it turns out, training an AI model on how to think like coders also improves its reasoning abilities, and everything else about it

[00:04:34] Chaz Wolfe: Interesting.

[00:04:35] Alexander De Ridder : turns out.

It is not only helping humans at making better decisions and think in a different way, it also is helping AI models, uh, expand their their reasoning capabilities. Now with that said, everything in your life is a process or workflow.

[00:04:57] Chaz Wolfe: Yeah.

[00:04:58] Alexander De Ridder : A cooking recipe is a process or workflow when you're, um, when you're thinking and code brain.

You look at a business process and you look at everything that goes into it, and you're already thinking in your mind, well, you know, if this, then that, then this, then that. Okay. The difference between old automation tools and AI agents is that when we try to automate certain processes, it's not as simple as.

Defining a hard-coded rule.

[00:05:36] Chaz Wolfe: Right.

[00:05:37] Alexander De Ridder : Life is fuzzy. Life is random. Life isn't expected. AI models now for the first time are able to execute workflows with a level of intelligence, that allows 'em to discern these fuzzy encounters and still know how to make progress. They know when to ask for help. They know when to use a tool.

They know when a spam email looks like a spam email without that particular word in your filter. And so that, that, and a few other aspects around it are fundamentally changing what can be automated in the workplace.

[00:06:24] Chaz Wolfe: Yeah. One thing that you said a few minutes ago that I wanna, uh, kind of allow you to chance to elaborate on, you gave us this picture of, okay, so it's not just. Build in the word spam and, and then the AI or automated tool is scanning for the word spam and then it boots it out. Like we're talking about intelligence where I can read it and I can infer or I can

[00:06:44] Alexander De Ridder : exactly.

[00:06:45] Chaz Wolfe: Right. Okay. So that's cool. But you also mentioned safety.

[00:06:52] Alexander De Ridder : Mm-Hmm.

[00:06:53] Chaz Wolfe: this is a big topic for you, so I'm sure we'll maybe get into that deeper, but just to kind of keep the listener kinda on their toes here a little bit, how do you keep that scenario from going all the way to the iRobot, you know, movie where it's like everything's taking over by ai.

[00:07:07] Alexander De Ridder : I won't be the one stopping our robot from happening. Um, but constrained alignment. We'll make sure that the ais that you are using, that you choose to use, can't, can't go beyond their, um, allowed workflow. They can't go beyond their constrained model, what they're capable of. They can't see outside of their box

they can not do other things than just pass in their inputs and outputs to the, to the next step that you have, that you have approved. So at any point in time, a human is owning the workflow and owning the decisions with full transparency of what happens inside of its brain at every step. Um, now the reason I say that we're not the ones who are going to stop iRobot from happening is because

the way I see things is, is not typically how they're conveyed in movies. Um, if you have an antron, you just have to kind of like kill its server or kill the robot and you're done. The way I see the web evolving is a collective consciousness with intelligence orchestration. The internet itself is skynet.

Every computer, every device. Connected to it is part of that Skynet. Here's how that works. One website may have the skill to do uh, writing. Another one may have the skill to do editing. Another may have the skill to do posting another may have the skill to do math. Another may have an a skill to do project management and so forth and so on.

All you need is a sufficiently advanced ai. That can know when to call on your website to solve that part of the problem and just check if it was done. And then you have an autonomous AI that can use the entire internet. Now the internet is a collective consciousness and in a sense it is alive. Okay, so it is a live how well the web, uh, grows and changes over time.

New websites come, um. We have, um, a process in the body. Um, I probably will mispronounce it, but gy, where dead cells are removed that are no longer useful. Well, sites that are outdated also get off the internet. New sites come, individual websites improve as well. Over time, they get stronger and better, and their capabilities expand.

Web3 is all about. Uh, making websites, domains, companies, individuals, websites more accessible to be consumed by AI models, which will become the largest consumers of the internet. So let that really sink in, right? So you will sell more of your service to AI agents than you will to human surfing your website.

And so, as such. Information will be consumed by AI agents who will read the website on behalf of the users who will purchase services from your site on behalf of their users, on behalf of businesses. This is Web3, the Executable Web, and it is a collective consciousness. It is alive in the sense that it constantly expands, constantly improves, and constantly optimizes itself and all we need to unlock that.

Is, um, an AI that knows how to orchestrate that intelligence.

[00:10:54] Chaz Wolfe: Yeah. I love it. Okay, let's, we've, we've been like up here in the clouds probably for the listener a little bit. Let's bring this down. Okay. I am a garage door business in middle America.

[00:11:07] Alexander De Ridder : Okay.

[00:11:08] Chaz Wolfe: How does this apply to me in 2024? 'cause you said by the end of 2024. There's these agents that you're even building that are gonna be able to help me as this exampled garage door business owner. You know, I'm doing repairs, I'm doing installations. We're answering phone calls like you get the business. But how are these agents helping me? The listener right now?

[00:11:29] Alexander De Ridder : I, I will just tell a story. O one day you'll sit in your sofa and you will say. Uh, assistant you're gonna talk to, whatever it is, apple, Android, whatever, right? Um, assistant is my car charged and it will, how will know that? Well, it will know that you own, for example, an electric car model, let's just say Tesla.

So it will know that Tesla has an A PIA website you can talk to. Um, it will authenticate you with its credentials. You've trusted this assistant to, to have access to your apps on your phone or whatnot, and it's going to talk on your behalf to Tesla and the, the Tesla is going to answer back, yes, your, your car is charged.

And you're like, great. When my show ends, pull up my car outside in the driveway, I want to leave. Now it's going to talk to your tv. It's going to talk to the TV app you're watching. It's gonna figure out how much time you have left on your behalf. This is, you're just sitting in your sofa. You don't, this has all happened behind the scenes.

Alright? Now it knows how much time is left. Then it's going to send that signal back to your car and say, um, you know, self-driving and stuff pull up outta the garage Summon mode. Okay. Get the car ready. Ch uh, Chaz wants to go places, right? So the signal said, but there's a problem. Tesla answers back to your AI assistant and says, the garage door is closed obstacle.

Right. At this point, the, the assistant would want to solve that problem, but how, how would you solve that problem? Well, you would. Solve it by opening the garage door. So the assistant tries to communicate to the garage door opener and says, open the garage door. Um, when it is successful, it is then going to back, go back to Tesla and say, now you can try again.

Alright, you're still sitting in your sofa, you're still enjoying your show. if you, uh, make garage door openers right, you'll for sure want to make. Make it so that your garage door opener can talk to all the different assistants that people might possibly have. The Alexas, the Googles, the apples, the Microsofts, the whatever ends up being the thing they use.

So you want to build your, your garage door opener in a way that can plug into this collective consciousness so that AI assistance can orchestrate that. Beyond that, you are, um, you are likely going to have. Uh, customers who want to know about your amazing product or service. So now customers, um, wouldn't go around browsing 10 different websites, clicking around links, trying to gather the information.

You're, again, you're sitting in your sofa and you say, Hey, um, my Tesla, uh, didn't pull up because my garage door didn't open by itself. Assistant, can you find me some company that, um, could open my garage door for me? Um, and, and the assistant will go and do that research for you. How will that AI assistant know that you are the solution to that problem?

Well, you could say, well, Google search, but search itself will change, uh, because, um, here's what the future looks like. Well, I've got brand or service one and two. Uh, which one would you like? And, and here's what they do. Who would you like to talk to? Well, I'd like to talk to Chaz's Company. Alright, great.

Um, why? And then you start a conversation with that company. Well, this is what we call a brand agent. The brand agent becomes your representative across the dozens of different AI assistant people are running. So you build it one time and you want to deploy it everywhere. Um, now you can see why this is the future of marketing, why this is the future of sales, why this is the future of engagement.

So why should you care? Well, well, for one, you need to make sure your product and service can be on this new future so that it can be useful. Otherwise, people won't buy you. Number two, you want to make sure your discovery of your product and service is also going with this new paradigm, this Web3 paradigm.

Um, and then beyond that, well, if you run any kind of back office, you probably have tasks like content creation, maintenance, project management, accounting, finance. Maybe you need to optimize your, uh, manufacturing, uh, line. Uh, you know your business better than anyone and you know what the obstacles are.

Um, now any of those processes, any of those departments can also internally benefit from helpers. AI agents that know your business, know your process, know your data, and can offer a helping hand. It is not human versus machine, but humans multiplied. By machine or agents. These are a few of the ways that, um, you should care.

Now, granted, if you're a blue collar worker and you go out and fix people's homes or like plumbing or something, uh, you may think, well, this sounds all very sci-fi, and all out there, I just go to people's houses and I, I do my work. That's, that's wonderful and great for a certain segment of the market.

They don't need to know how the pudding is made. They just need to be able to go to the store and buy the pudding. So for that, this is not gonna happen by us. We are not going to be the ones doing this. But the Googles, the Microsofts, the open ais, the apples, they're going to create op stores with pre-built agents.

Pre-built functionalities. Think of it like an internet. Of agents and if you are in such a position where you're not so technical, you don't think that you'll ever create your own agents or your own garage or opening, uh, agent that talks to and so forth, well then just look forward to leverage the technology that's already out there and use it to your benefit.

You don't even have to create it to benefit from it.

[00:18:26] Chaz Wolfe: Yeah. That's incredible. The, the story that you gave there is pretty, pretty powerful from all angles too. How in your best guess. When does this scenario of me sitting on the couch or you sitting on the couch having this whole thing happen, and me as the garage door, uh, company owner needing to be plugged into this, this collective consciousness?

When, when does this happen?

[00:18:46] Alexander De Ridder : it is happening now. But has not yet hit critical mass in the market where everyone buys their next phone because that's the killer feature that they've come to expect. Um, the, we call this like inflection points in business, right? We have reached, um, we've, we have hit a few inflection points already.

The first inflection point is, uh, September last year. Well, where all the tech giants committed to this future, billions of dollars and, and, and their entire, they bet their entire future on this. So since then, um, in the words of Yoda begun, the clone wars have right

[00:19:36] Chaz Wolfe: Yeah.

[00:19:37] Alexander De Ridder : from that point forward, it is not something that is if, but when.

[00:19:44] Chaz Wolfe: Yeah.

[00:19:45] Alexander De Ridder : Uh, since then, uh, the chat GPT store launch was a major inflection point last week. Um, they made it possible to talk to your GPTs as if you're talking to a colleague by typing at, and then the, the agent name, which means that in a single conversation you can have conversations with at researcher, at writer.

At editor, at publisher, and it can take care of your entire work in a single conversation. This is the start of a multi-agent future. Beyond that, um, Microsoft is making their co-pilot part of Windows 12 as a core, central part of that. we feel like we're back in the internet explorer bundling days.

[00:20:39] Chaz Wolfe: right. Yep.

[00:20:40] Alexander De Ridder : Amazon has come out with two dozen devices to win your, your, your home assistant. Apple has committed a billion dollars in this field to catch up and, and, and win. Uh, meta has announced and is rolling out their AI system to technology across their billion plus users. Um, so this is on, um, windows copilot.

And the business is very important as an inflection point. Microsoft Teams is used by a large portion of the enterprise, okay? These agents can be deployed via copilots as members of your team. They're there in your company, in your chat, on your secure infrastructure. It's not. It's not if it's when, and in some sense it's already begun.

It just hasn't gained the massive traction where everyone's talking about it yet. But that's why you listen to Chaz's podcast because you want to learn these things before everybody else.

[00:22:32] Chaz Wolfe: That's right. I appreciate the plug there. It only is created by great guests by you though. So let's, let's compare these timelines because you just made a reference to, you know, early Microsoft days. You know, I can remember being a kid with a OL chat and dial up and, you know, even before that obviously was the.com boom, uh, and bust.

And so. We're 20, 24. Right now, we're, we're in essence, let's just call it 24 years into the internet. I know it's probably more like, you know, 40, but for the, for the mass, right? 25, 30 years. What, what do we look like here? What do we, what can we learn, I guess from that maybe 25 year period to where we gotten here? Because we know the internet grew a whole lot faster than the previous technologies 'cause of the, the design. And so we inherently know that this is going to like that hockey stick to mass adoption that you're talking about is going to happen faster. can we learn, I guess is what I'm trying to say from what we know of the early Microsoft days that you just kinda mentioned we can kinda like keep in mind as this thing starts to happen.

[00:23:40] Alexander De Ridder : Humans have a perception of time and an experience of time that is linear. The sun goes down, the sun comes up. Full moon, half moon, full moon, half moon. And we look at time as it is a linear thing the way we anticipate problems. If you were playing basketball with your friends and they always pass the ball like weak sauce, like right, you come accustomed to it, right?

[00:24:09] Chaz Wolfe: Right,

[00:24:11] Alexander De Ridder : If that same friend all of a sudden start passing the ball to you like unexpectedly super strong and fast and direct, you're not gonna catch it. You're not gonna, yeah, you're not gonna catch it. Right? So the. The acceleration rate of progress is really hard for us to wrap our brains around.

[00:24:31] Chaz Wolfe: Yeah. Understood.

[00:24:32] Alexander De Ridder : We are right, so it's like living in linear time, but then progress passes us that basketball in a very direct way and we're like not ready to catch it in one year.

Humans produce knowledge or content. That equates to the sum of all written knowledge and the sum of all human history, including the year before this.

[00:24:57] Chaz Wolfe: Yeah. Wow. That's incredible.

[00:24:59] Alexander De Ridder : The, the rate of progress in every field is going to a hundred fold thousand fold. Thanks to ai, our brains are not ready to understand or accept that reality, not even mine.

Every single time I make a prediction about the market, things happen way faster than I could possibly imagine. It becoming a reality. For software startup founders like myself, agencies, it is a very difficult time to be in business because how quick everything is moving and how quick you need to adapt.

And to be honest, it can feel exhausting.

[00:25:41] Chaz Wolfe: Yeah.

[00:25:42] Alexander De Ridder : Um, it is, it is very hard when you have to pivot and update so quickly

in terms of understanding or predicting where the future is headed. And what lessons we can learn from the past is that the companies who are most agile, lean, and adaptable are the ones who are going to make it. There is no point fighting this future. You need to embrace it wholeheartedly, and if you are fat, you need to trim and get lean.

If you are not using these technologies, you need to realize that four fighters years from now, you may become irrelevant, not because you're not doing a good service, but because your competitors are doing a good service at one 10th of the cost.

[00:26:29] Chaz Wolfe: Yeah.

[00:26:30] Alexander De Ridder : And so the biggest lesson is that yes, people do adapt, and I believe humans will do amazing things in this future and will have amazing opportunities.

But the transition period is hard, and it is the ones that are adaptable and lean and embrace this, that will thrive throughout that period.

[00:26:55] Chaz Wolfe: Yeah, that's it. It's, it's good to say that and we understand. That. Okay. Yes. The, the, these practical things make sense. Give me a picture of what it looks like to toss me that ball with a little bit of heat on it and it goes right past my hands and whacks me in the face. For your example, they were passing the ball.

What does that look like in the market? Because you're right, I've seen that happen even with, with I blockchain, right? Like it was like really exciting and it kinda whipped everybody in the face and fell off the map. And now it's kind of come back a little bit. It's like. What is this like ebb and flow of like getting excited and then we get whipped in the face and so we kind of back off and then we get excited about it again.

[00:27:35] Alexander De Ridder : Yeah, so. the more proper understanding of AI is more like you would think about electricity than you would think about crypto.

[00:27:43] Chaz Wolfe: Okay.

[00:27:44] Alexander De Ridder : Um, once electricity was introduced, it took a while before we all had cell phones in our pocket, but it never really like went away. Right. And when electricity was first introduced, um, people were dreaming about a future with flying cars and so forth.

Right. So you have the two aspects, right? On the one hand, AI is not a fad. It is here it is. It is improving every business aspect. And there's the direct consequences that are immediately obvious, and those are the things I talked about. Those do not rely on any new inventions. Okay, so let me make that crystal clear.

Everything I said so far on this show is a future that is possible. With current technology

[00:28:36] Chaz Wolfe: Yeah.

[00:28:36] Alexander De Ridder : at the same rate, you have a group of people who are thinking about AI that is so smart that it will outperform humans and make humans irrelevant. That will merge with machines that will cure every disease will live 500 years old.

Okay? That is the same thing. People had in the year 1900, they thought by now we'd be living on the moon. And I don't know, you know, flying cars everywhere. There's the aspirational aspect and the unknown of how this technology progresses. And here we could run into a, um, the Gartner chart, the hype, the Gartner hype cycle, right?

The peak, uh, of inflated expectations. We might be there now. Uh, with, with this, and then you go through the valley of disillusionment when you realize it is maybe not infinitely getting better exponentially in every direction. In other words, you're not reaching the singularity tomorrow, I'm sorry, Reddit.

And then you get to the plateau of productivity where everything kind of settles, right? When electricity first came around. If you would take it to certain tribes or villages, people would accuse you of magician. Right? It feels like magic. AI right now feels like magic, and we're contacting Amazonian tribes who've never seen the outside world.

When I show what AI can do to, uh, my grandparents, they, they can't believe it, right? Some will say, well, um, this is something spiritual, like, don't trust this, right? Because. It's so foreign. It's like magic, right? We, we have to distinguish where we are and what is the immediate, uh, first and second, uh, degree consequences.

And third and fourth, based on currently where we are, we have only begun to tap the applications of already discovered electricity in the form of ai. And then. That right there can be in itself a ball that is passed to you and you can't catch. 'cause if you don't catch that, everything I'm telling you this is going to happen in the next year, 2, 3, 4, but it is already here.

It is inevitable this is happening. However, the other part, the sci-fi part, if you have huge expectations for that, you may also go through some disillusionment. The G PT five may not be such a leap as you hoped it was gonna be, and we may discover that it is too expensive and too slow to run at such intelligence.

Or we may have to wait another five, six years for a major breakthrough in architecture or for hardware costs to comb down even more before we can make that next leap. There's many things that can happen that could lead to disappointment. Uh, if you are that group of people that expects sci-fi tomorrow, however.

The existing, and I want to underscore this, the existing technology, the existing AI models are so powerful and we have only scratched the surface. And with that, I mean a surface of Mount Everest, we've only scratched very tip of that mountain in terms of what is possible in terms of it, uh, being applied to our lives and changing.

How everything works in the market, how everything works in business, and we need to be agile to be ready to adapt to that.

[00:32:11] Chaz Wolfe: let's, let's keep using this, um, example of catching this pass here and let's talk about practicals around catching the pass. Right? And we just talked about what happens if it smokes past my hands and hits me in the face. That kind of like reverberates, but what. What can I do practically right now as an entrepreneur listening today?

Maybe I've just started my business in the first couple years. Maybe I'm a big enterprise and we're doing hundreds of millions of dollars. What can I do today that helps me get a hold of that pass when the ball comes to me? I.

[00:32:41] Alexander De Ridder : Right. So at the very minimum, you need to start exploring how you, yourself, and your organization can start executing an AI transformation. Even if a human is still better at a certain thing, if you have not explored the limits of how AI can help you in that thing, you're behind. Um, because again, today is one thing, tomorrow something new can happen.

And if you're not, haven't already done the basics, you're not going to go to the next step and you'll be behind. So at the very basic level, um, you need to do that and there is no reason why you cannot. Um, using AI in many software services is integrated. Canva, bing, uh, you don't even necessarily have to pay for it for basic versions, so there is no excuse.

and just fun fact, uh, my kid was asked, had trouble with an assignment. He didn't know what the teacher meant when they were asking a question. And he said, my teacher's not. I emailed the teacher not answering in real time, but I wanna finish my assignment. Now. Every other question about the book and topic, I can look online, I can ask for help, but I don't know what the teacher's thinking.

So I don't know how to ask for help online. So we took a picture of the assignment, we gave it to ai. We asked it to help interpret what the teacher might likely mean by it, and gave a really coherent answer. Um, my kid of course, was amazed that AI could do that. And I said, look, you now have tools where you do not have to be stuck, right?

Uh, it's like having the world's condensed knowledge in your pocket. If as a business you don't, al you're not already doing the basics, right? Start there. If you are more, uh, ambitious. Um, we created Smith os.com Smith with a y. We created it with a future in mind, where, where you'll have. Three people operating a hundred agents, right?

You'll be able to do work of a hundred people with just three and multiplying humans, right? Agent engineering, multi-age, agent system engineering. But let's just call agent engineering is a core skill that I believe everyone should acquire if you want to remain relevant. If you know how to build, operate, and maintain agents to do work at scale in finance, accounting, marketing, manufacturing, you take your specialty and knowledge of the of how to do the work, and then you apply that to build, operate, and maintain agents.

You will not be replaced if you are a business owner. Uh, this is what I myself do. I will pick 1, 2, 3, 4 people depending on the size of my organization and train them to be agent engineers and have them actively start automating every part of my business. If you do not do this now, you'll find yourself at the end of 2024 against competitors who have done it, and you'll just be amazed that all of a sudden.

They can drop their prices tenfold, and you'll be wondering, how did I not catch that ball? How do I compete at that time? It is too late. How long can your business survive with a competitor who's 10 times more efficient than you? Not for very long before you lose your customers, before you lose your brand.

So if you are forward looking, I would say pick. 1, 2, 3, 4 people in your organization. Teach them agent engineering and start automating every part of your business for efficiency. And, um, beyond that, if you're say, I really do not have the capacity to do this, that's okay too. just like the App Store, we're going to see agent stores.

This is the end of Upwork and Fiverr. Okay. Uh, instead of, um, instead of having physical Chas advise you, um, you can talk to Agent Chas who has all the knowledge of Chas as a lower cost alternative that still can do 80% of most things. People will ask you while you're at the beach and the agent is working on your behalf.

if you're not into creating this yourself, at least be a consumer. If you have an organization, go and upgrade to GPT For teams. It's an extra five to 10 bucks or whatnot per month. But what you can do is you can have agents that are exclusive to your business and you can have, you can have each department start communicating like that and create departments as such.

So be a consumer if you're not gonna be a creator. Not everybody needs to be a creator. We all have lots of apps on our phone that are useful, that benefit us that we haven't made ourself, so that would be my advice. Step one is everybody, including my children, need to know how to use AI responsibly and effectively.

Number two, if you really want to swing for the bleachers, go and learn. Agent Engineering Smith West makes it very easy. But you can also learn how to code it yourself if you're brave enough, and it's more work, takes more time. But you could do that. And then finally, if you're not going to build, build anything custom for your own data APIs process, if you're not interested in that, at least start taking advantage of these emerging app stores or uh, extension stores that will allow you to amplify your productivity.

Make sure you're on the front lines, at least of the consumption benefit side. Of this revolution.

[00:38:47] Chaz Wolfe: Yeah, you did a, all your answers have been so succinct. The, the piece there with Smith os I, I want to just give you an opportunity, a platform here to kind of give us a little bit more there. Obviously that's your company and this is what you do. So yes, he gave you plenty of like maybe free or easy options, but if you're like me and you're trying to, you know.

Be ahead of the curve and not behind the curve. Uh, what, what can I find specifically if I go to Smiths right now and I'm interested in engineering and, uh, implementing some of these things, what can I actually expect from you and your company? I.

[00:39:20] Alexander De Ridder : Yeah. So at this time we're in early access and we're, we are picking and choosing a hundred companies or small teams to work with. Some are very large. Um, I can't name their names. Um, and some are very small and they're, um, they all have these thing in common where they want to be, uh, thought leaders in their industry and they want to tackle AI transformation head on.

Um, before general access in our early access program, um, we limit to a hundred. So I have time to work with them personally, understand their problem, train their team, and provide very intense support for them. I. Um, this helps us in many ways as a company because we learn firsthand what their challenges are so we can make general documentation better.

It also allows me to, uh, create some incredible applications of agents, and we are, we would need another podcast to delve into what people are doing with this. But it can blow your mind. It will, it would blow your mind what people are doing with this and, um, beyond early access. Uh, we will then open it up to the general public after some time, and that will be very accessible.

Um, start very, very affordably and just scale up with, based on how much work your agents actually do. That's the metric by which you'll pay. So if you are just trying to learn it, it basically costs almost nothing if, if not free, if you're then deploying it to actually start doing work in your organization.

Um, the, the comparison with hiring more workers that do the exact same thing will be so favorable that it's a no brainer. And then the cost will drop up to 2000% in terms of the cost per task as you scale as well. Um, then you will orchestrate different AI models and tasks. Um, well good news there too.

Uh, chat, GPT just last week cut their prices of input tokens by 50%. The third time in one year, they make a double digit massive cut in their prices. AI models are the next spelling and grammar checkers. Alright? They're trending to free and so take advantage of that. Uh, cost is not the prohibitive factor.

If you're interested to learn how to become an agent engineer, and for for sure if you have an organization or business, which, which would be hugely benefit from. Uh, having such optimizations be an early adopter before your competitors are. And I, me and my team will personally help you make sure you're successful in that training and you'll have a headstart over everybody else if that's not you.

Wait for general access and then you can play with it and explore in your own time. And then you can see, um, can see how that will benefit you over, um, over a longer period of time. But at the very least. If you're not already using AI in your business, get familiar with the models. If you are a little bit more advanced, be a consumer.

If you're not going to be a creator, be the best consumer you can be of these technologies.

[00:42:40] Chaz Wolfe: That's great. That's great. Um, and you've already given the website, but where can they find you? Or even if they just wanna find you online, uh, where can they find you

[00:42:48] Alexander De Ridder : Yeah, well, I, uh, I most post to LinkedIn and, um, they have this feature, like it said. I also repost to Twitter the same thing, and it actually was not working and I thought it was working this whole time and it was not posting there. So I'm making a conscious effort. Yeah, I'm making a conscious effort now to like post some, repost some of my content to Twitter as well.

Um, or X and, um, yeah, mostly LinkedIn and Twitter. Alexander De Ritter. My handle is a d Ritter, A-D-R-I-D-D-E-R, and um, my website smith os.com. Smith with a y. Um, and um, yeah, it's a platform for easily building agents. Fun fact, we have writers, uh, that learned how to be some of our best agent engineers.

Even if you have no development background, you can make some incredible things. So

[00:43:51] Chaz Wolfe: yeah.

[00:43:53] Alexander De Ridder : don't need a degree in computer science to build agents just like you do not need to know Assembly to make a website

[00:43:59] Chaz Wolfe: That's right. That's right. Yeah. There's other tools in place now where that, that becomes obsolete. Alexander, you are doing some incredible work. I think, uh, you know, if anybody listens, um, like truly to the message that you've given him, them today, that their, their mind's gonna be spinning, excited.

They should want to go do something about it, because here in five or. Maybe even less years, but for sure 10 and 20. Like when we talk, you know, I mean, geez, 2024, when it's 2044, I, I can only think back on this conversation be like, I remember when I talked with Alexander, you know, like, you know, the, the predictions of the future and ai, you know, I.

[00:44:35] Alexander De Ridder : if I can just, uh, just there, there was a time before you had your first computer. And then the time you first had internet, and then there's a time before and after smartphones, the iPhone 2007, and there was a time before and after, um, AI and chat GPT, right? There will be a time before and after AI agents entered the work workspace as well and how society will change.

So if we think about the time, the, the, the time we live in, think about in the last 30 years or so. Think about how many times we discovered fire or the wheel and how the world has changed. It's before and after, right? To this day, we say ad uh, an domini after Christ. To this day, we say there was a before and then there was an after and the world was different.

There's a before AI agents and there's an after AI agents. And we are privileged to live in such a time when we have such inflection points happening right before our eyes.

[00:45:43] Chaz Wolfe: Yeah. Wow. Uh, if that doesn't get you fired up and wants you to pay attention to what's going on in the world, I don't know what does Alexander, you are an incredible, uh, human and a mind, and you're just doing incredible work. Thank you for being here. Blessings to. Everything that you touch here in 2024, I wanna be part of it.

I wanna make sure my companies are on the leading edge of, of AI and I know many of our listeners will as well. So thank you for allowing us to do that. Thanks for being here, brother.

[00:46:09] Alexander De Ridder : Thank you.

Host Chaz Wolfe welcomes Alexander De Ridder, founder of Smith OS, to discuss the ever-evolving world of AI and its implications for businesses. They delve into the concept of AI agent engineering, the need for businesses to adapt and utilize AI technologies, and the incredible developments in AI that have already begun to shape various industries. Alexander shares how various AI models are being integrated into software services, the increasing importance of training agent engineers to automate business processes, and the various levels at which businesses can engage with and benefit from AI.

Alexander De Ridder:

Website: https://smythos.com/

linkedIn: https://www.linkedin.com/in/adridder/

Chaz's favorite morning drink to fuel him for his day

10% off Code: GATHERINGKINGS10