This podcast episode elucidates the pivotal role of artificial intelligence (AI) in contemporary business strategies, featuring insightful discussions with our esteemed guest, Sanjay Rakshit.

We endeavor to demystify the complexities surrounding AI implementation, emphasizing the necessity for businesses to approach this transformative technology with discernment and critical thinking.

As AI evolves at a staggering pace, we highlight the importance of understanding the distinction between mere tool usage and the strategic application of AI to enhance operational efficacy.

Furthermore, we address the prevalent misconceptions and hype surrounding AI, urging entrepreneurs to seek informed perspectives rather than succumbing to the clamor of unqualified voices.

Our discourse aims to equip business leaders with the knowledge requisite for navigating the intricacies of AI integration, thereby fostering sustainable growth and innovation within their organizations.

Takeaways:

1. Businesses must critically assess the necessity of AI in their operations to avoid pitfalls and ensure effective implementation.
2. The rapid evolution of AI technology necessitates a continuous learning mindset among business leaders and practitioners alike.
3. It is imperative for enterprises to differentiate between mere tool usage and strategic AI integration to achieve sustainable competitive advantage.
4. The influx of AI tools does not replace the need for skilled practitioners; rather, it enhances their capabilities, enabling more efficient outcomes.

The real world entrepreneurship podcast with bhairav.

Patel and alan clark.

Welcome to the Real World Entrepreneurship Podcast where we tackle the practical challenges that startup founders and small business owners face and give our own unvarnished opinions.

My name is Bhairav Patel.

I'm the Managing director Atom Ventures and Atom cto where we partner with small businesses and startup founders to help them define and build their tech vision.

And I'm Alan Clark.

I'm the founder of the Business Growth Partnership and we work with high potential businesses who help them unblock the blockages that are preventing them from reaching their full potential.

Welcome everybody to the Real World Entrepreneurship Podcast.

We're back.

It's been a number of months since we've been recording anything.

But that's because Alan seems to be perpetually on holiday.

He seems to have given up work and just got on a plane and traveled around everywhere.

But now he's back and he's AI curious.

And so today we're joined by Sanjay Rakshit.

So anyone who's listened to my atomcto podcast and actually some of the old podcasts here, we've had Sanjay on as a previous guest.

So say hello, Sanjay.

Hello.

Hi.

Thank you for having me.

Alan, good to see you again.

Been a while.

And so today we're going to be talking about AI.

Obviously we've got the AI expert in the room and we have a non expert in the shape of Mr. Alan Clark who is going to pose a number of questions.

And I think I'm actually the norm, to be honest.

I suspect that most people in the big world out there haven't really got much of an idea of what it means other than there's some tools that they're able to use that they don't really understand, but they give magic results and we don't look too hard because as large of work, that's fine.

But clearly there are some much, much, much bigger issues bubbling around the whole world of AI.

So that's Alan, why we're here.

So today we're here to actually help Sanjay to explain and clarify to people what exactly is going on.

Because I think anyone who's really been following AI over the last, I don't know, what, say, a year, every, it changes so rapidly, it's crazy.

And anyone who's kept on top of, let's say even the last Release with, with OpenAI and chat GPT5 and there was a huge kind of raw uproar about that, you'll see that, you know, nothing is stable, everything is in flux.

And if you try to build a business on previous models of, of chat GPT, you probably don't have a business anymore.

So, so Sanj, let's, let's kick it off right when we're looking at this.

What do what, what are the main things about people right now like Alan need to know about AI as it is because there's so much hype, right?

So much hype and people are talking about we're all going to be, all our jobs will be lost.

You know, Skynet is coming, all of this good stuff.

Who should really be listening to in this world?

What voices are there that are actually the voices of sense and reason?

Great question.

I think first thing is we've got to be very careful that we don't fall for hype.

So just the person with the largest platform is not necessarily the person with the largest knowledge or even the largest information.

So I'm going to answer that slightly differently.

Wind back a few hundred years, engineering took time to bed in because there wasn't a way to talk about it.

Now everybody can talk.

I can.

You know, anyone who has five spare minutes and would love to be bored, they listen to me or Sam Altman's Elon Musk.

They say things which are quite provocative or could be far more interesting.

So there is a platform for everyone to get their voice across.

That wasn't the case 500 years ago, even 10 years, 15 years ago.

So things took time to bed in and they were.

The information about them were propagated through experts who had tried it, had failed, had learned from it, tried again.

There was an evolution.

So we had people we could listen to who had tried it, failed and they would come at a more measured approach.

It was not just anybody talking about anything.

Civil engineering, computer science, engineering, electronics, engineering.

AI has been around for a while, but it kind of took off in the past few years when the social media and the ability to just voice your opinion took off at the same time.

So there's not been much of a chance for practitioners.

Nobody's an expert.

This is brand new.

We're still learning.

It's been around for a long time.

Even now we are learning, right?

But we have a platform for anyone to say anything about it.

And what we have are all opinions.

What we see is that majority of what's being said is either regurgitated from somewhere or it's an opinion and most of the time not informed because those who are actually trying to figure out what's going on, they are busy trying to making mistakes and learning from it.

And then collating things together.

But because things are moving at a breakneck speed, the moment and anybody can say anything, pretty much there is a tendency to fall for the largest or the loudest voice.

And so that's where hype is Hype, right, correct.

And that's not just us, it's also VCs, private equities, everyone.

So suddenly, if there's a business, I mean, go Back to when ChatGPT came out, the six months after that, we had so many GPT companies, startups that came up and how much money was pumped in.

I think if I remember right, it was about 2 billion was lost on the market.

I think 10.

Yeah, you're right, because everybody went, oh, the GPD must be there.

Again, that was a hype.

Even VCs fall for it because there's a, there's a race of, for not losing out.

So if they fall for it, what chance have we got?

Or what chance of anybody who's not had a chance to.

So I think the key thing here is to ensure that nobody to understand is that everybody is learning.

Even the loudest voice on the largest platform is learning.

Most people are not deeply involved in technology.

It was the launch of ChatGPT that really brought this into our mindset and actually gave us something that for the first time we were able to use that was called AI.

But how far back did the genesis and creation moment of AI go?

How long has this been a development for before it became something tangible for regular people?

AI has been around, it's been knocking about for a long time.

I mean, Alan Tyrann go back even that far, but AI has been going for a long, long time.

Because artificial intelligence, what it is or what is it, it is getting a machine to think like a human.

And what does a human brain do?

It's replicating that.

What we're trying to do is we analyze, we perceive.

So we see something, then we analyze it, or we interpret it, we detect a pattern, we come at a decision and then we act on it through our body function as mechanical.

But up until the point when we decide what to do, it's our brain doing.

So what we're saying is that can we get a machine to do that?

Can a machine see something, an image?

Can it detect some objects in it?

Perceive?

Can it then figure out what that object is, not just a set of boundaries?

And then based on that, can it decide to do something?

Can it extract some insights and then decide, so that is what AI is?

And to do that?

Research has been Going on for more than 50, 60 years, probably more.

We can go into that in a separate podcast about the history and the timescales.

But I think that what's important to understand is that it's been there for a long time.

The reason it didn't gain much prominence is because of the lack of resources.

We didn't have the compute power to crunch through all that data and learn, because that's what you learn from.

Data is the key ingredient for AI.

We learn from.

Otherwise it's an expert system.

So then came the AI winter, because research dried up because we weren't really seeing the benefits that we thought AI would give us because of the lack of compute power we didn't have.

There are many chips, GPUs, you know, we didn't know even know at the time that GPUs could be used.

Yeah.

So we had.

So because of the lack of resources, we didn't get the.

The results that we perceive.

So a lot of money, research money was considered wasted.

But then through the AI winter, there were a handful of universities around the globe, UK and Edinburgh, leading the way, had still some money, and they continued research.

And then resources were made available.

People discovered general computing or General Purpose Computing.

GPUs came into prominence.

There were faster processors, storage became cheap, data centers came along.

So we had resources and then again gain traction.

And suddenly we realized we were able to see the benefits of it all the research that happened.

So AI has been there for a while, but because in the past few years we've had an abundance of resources and cheap resources, the costs have gone down, it took off.

Sounds very much like so many technologies or sciences that in the very early days there were people stumbling about trying to work out how to do something.

And it takes a long time before they work their way through.

And it kind of catches fire.

And then we usually go through a bit of a period of what we might call.

And then later we were just trying to watch out what this is really all about and where are the parabols.

Is that what's happening in AI just now?

So what's happened in AI just now?

Is that a bit of that, but also that.

Because as I said earlier, because now everything gets played out in the open, which means that.

That universities will have to start having courses.

So there was a time go back on what we call emerging technologies in any field, manufacturing, anything.

There were only a few universities in the world who would offer those courses.

They were called center of Excellences.

So that's where they learned they Made mistakes, research came out and then other universities then adopted that studies.

But today, because everything gets played out, there is a race, it's a constant race to be the first or at least to be seen to be in that game.

Which means that the quality, the due diligence, etc.

Or the rigor that's required to offer such courses are lost.

There's still a handful of universities in the world who are centers of excellence, but everybody are offering AI, data science, data analysis courses.

So who do you go to?

Because you need the body of knowledge and research behind it to teach students.

Where is that coming from?

So there are quite a few considerations that need to be, to be looked at.

But the underlying fact here is that now there is a race to be seen and that is coming at the expense of knowledge.

What we have today are opinions largely, but very few factual information.

So going back to the quality piece, right, Because I think that's very important for people who are considering adopting AI into their business or into generally in their lives, right?

Because if you think of all the trends that have happened over the years, typical one is something like gdpr, right?

When GDPR first came out, suddenly within a week of the GDPR announcements, you had people who are experts in gdpr, right?

Even though no one knew how it was going to be applied, no one knew how, how the laws were going to actually be interpreted, all of that kind of stuff.

And you had these big conferences and, and everyone was a data data expert and they still aren't, you know, they still still have that trend, but it's died off as much.

Same thing when, let's say Chat GPT came out, you suddenly had all these AI consultants popping up everyone's cv suddenly, people who you'd never even known did a master's in AI or anything with machine language or machine learning, Sorry, they will suddenly.

Well, yeah, yeah, I've been building AI AI models or I've been actually building AI applications in the last two years.

It's like, well, this is on, you know, never heard of this before.

So it's been very difficult for the average person to figure out.

How do you actually know when you're talking to someone who knows about, who knows what they're talking about, right?

Or actually talking to someone whose opinion really should be considered or whether you're hiring someone who actually does have an expertise in a, or hasn't just read, you know, some random newsletter or gone to chat GPT used it a couple of times and, and learned things.

So how do you differentiate that chaff in the week, because in the early days of technology, it wasn't really like that when, you know, I mean, back in the day when we first started developing.

Right.

It was a lot.

You actually had to practice the skill.

It was a hard thing to do.

You had to do everything because it wasn't as easy with all the tooling that you have now.

And now the tooling has become much easier to set up.

Everyone is now an AI expert.

Right.

So how do you do that?

Differentiation?

There are no AI experts.

Well, there's no point.

Yeah.

That is the thing that terrifies me, and this is anybody can jump onto ChatGPT and say, give me enough information to make me look like an expert and then convailing them a little bit more than the person they're talking to.

You.

Know, I think I'd be tempted to segregate, categorize different people.

So the good thing here is that I would treat AI tools as a power tool.

So let's look at individuals, like consumers.

For them, AI is a tool.

How can that be used?

So it's me with diy, I'm useless at it, but doesn't mean I can't use.

I'm fascinated by power tools.

I may not be able to use it.

Somebody who is good at diy, give them a power tool.

My God.

Great.

A power drill or a jigsaws.

What do you call it?

A jigsaw.

See, that's how bad I am.

That's a consumer.

It's important to separate the various Personas here.

That's a consumer.

Then there is the enterprises.

This is where a consumer will not lose money.

My concern has always been, and that's what drives me towards democratizing AI, is that it's one thing, a consumer using tools and trying them out, that's like buying, going to bmq, buying all the power tools you can get.

The outlay is not that much.

And the result of having the wrong tool is not as bad, it's not as significant.

But for enterprises, it's different when enterprises embark on a particular approach of implementing AI.

So let's take.

There are four levels of AI.

The one is the infrastructure, then there is the model provider, then there is the tools provider, and then the application writers.

Right?

So you look at the.

An enterprise, when they embark on a particular strategy or an approach to implementing AI.

Let's say they're AI application developers.

If they don't make an informed decision to begin with, at the very beginning to lay the foundation, that's a very Expensive.

And I mean, when I say expensive, I mean a few tens of million dollars of worth of cost or hundreds.

Sometimes that would be down the sink, really, because they chose the wrong approach.

So that's the area I focus on where I look at and say, okay, how can an enterprise go in it with their eyes wide open?

So if you've got to decide at the very beginning, do you want to be a tool user?

So as an enterprise say, okay, you know what?

I build train wheels.

I use that all the time.

Bhairav knows that.

Train wheels, right.

I want to use AI to help me build train wheels.

So I'm not an AI company, I'm a trainer.

But I'm using it as a tool.

So it's like a building company uses power tools.

They don't produce power tools so that they can build more houses.

Great, no problem.

But when a building company goes, you know what, I am also going to dictate what kind of power tools needs to be used.

Now, that is a different Persona because then you're not a tool user, you are a building company who is using the tools that they use as a competitive edge to other builders.

That enterprise cannot just say, I've got LangChain, I've got Lang Graph and I've got Crewai.

I can just use those tools.

That's a poc.

They've got to take a more considered approach of expertise in that domain of AI.

Say, how can I apply leverage AI in my area of business to achieve a competitive edge so that their customers are looked after not just from the service they offer today, but the service that will evolve as the domain of AI evolves.

So the customers can say, right, don't worry, I've got Alan's business.

That's where I am on.

They are an AI practition business, which means that as this field grows and evolves, they're not just going to jump on the latest and the greatest.

They will take a measured approach and they will build a system that scales and evolves with evolution of this, of this discipline.

I think that's the fundamental difference.

What's a tool user and what's an AI company?

What did journey before with any other technology?

Are we relearning or are we visiting a path that we've been done?

Or is AI only world?

I'll give an example.

When I worked in NCR, we were an OS2 business.

And then we went to Windows XP and then we had to move to Windows 10.

I remember Windows 7, sorry.

And then Windows 10.

And just because there was a newer version of Windows available.

We just didn't go lock, stock and barrel and say, right, let's spend 10 million quid and change all our operating system to XP.

We drove a deal with Microsoft to extend the service support for us, even though Windows XP life long term support had come to an end.

Why?

Because as a business, we said, right, it's one thing being on the latest and the greatest.

R and D teams can do that, they can try things out.

But we have a, we had a basic responsibility to our customers.

So if we just went and chased the latest and the greatest, how do we know it's going to work?

AI.

That kind of rigor has to be set in the AI.

I'm not saying for a minute that we don't try the latest, but that's the word.

We try, we trial, we pilot.

But we can't just be that company enterprise that just chases the latest and the greatest because somebody says so on a platform or someone who said on a podcast.

That's why it's coming.

That's why we need practitioners.

We need people who have learned this discipline, who are learning this discipline, who are in touch with the researchers working with universities to understand the pitfalls or what's unknown.

So we can go into it with our eyes wide open.

So how do I identify those people?

How do I know that I've actually got someone in front of me?

It can really help me solve problems and not just be one chapter in the book further on than I am.

It's difficult.

There's no easy answer to that.

There's no easy answer to that at all.

Because primarily because of what the drivers are.

Let's take recruiters for a minute.

They look for keywords.

If we look at businesses, they, they tend to listen to podcasts with maximum.

So that's there.

There's peer pressure to say, hey, did you listen to Alan?

No.

Oh my God.

You are not with it.

Oh my God.

I better go listen to Alan.

He makes a lot of sense.

Great producer.

Sanjay.

No.

Oh my God.

Why not?

And then he realized he doesn't make sense.

But I can't say that because, you know, he's got another 50 people.

I think there are.

There, there is that as well.

So question is boldness.

Question is there's no harm in being wrong.

But one has to have applied critical thinking.

I cannot emphasize that enough.

It's the critical thinking, it's not the herd mentality.

We have to move away from it.

You know, we cannot just go and say, I Want to listen to X because they have all these platforms.

We have to apply our own critical thinking.

We have that even we don't need to know the domain, but we can still apply critical thinking.

You can see, okay, where is that person coming from?

Oh, let's do a bit of digging and say, oh, okay, you know, they need funding.

So you say, because nobody gives us the complete picture because they don't have it.

We are all learning.

We don't have the complete picture.

So we should not take an opinion as gospel just because another million people listen to it.

I've interviewed people for AI roles in the past, right.

And obviously knowing Sanjay helps a lot when, you know, when you're interviewing people for these types of roles and it's, you've got to be as critical as you would be with any other role, right?

So for anyone in front of you, you've got to understand what have you done?

How complex is that thing that you've done?

Where has that thing been used?

Has it succeeded?

All of these different, you know, and people will say it's like anyone.

The equivalent of this is like mobile apps.

I mean the real equivalent of this is blockchain.

Because when blockchain came out, suddenly we were all going to be on the blockchain, right?

The world was going to be on the blockchain.

We were all going to have instant payments, we would kill the dollar, would disappear, all of that kind of stuff.

That hype was huge and there's still people talking about it, right?

But it hasn't happened, right?

This revolution hasn't happened.

But if you remember back in the day when everyone.

I still get 20 emails a day from people who can tell me they can build mobile applications for, for us.

And the way you would interrogate whether they're going to be worthwhile or not is again going back.

What have they done?

Who are they?

Who's using them at how scalable is it has.

What did it cost to produce all of these things, right?

You would be.

It's the same critical.

The reason that I think we, we kind of let go of that critical thinking with AI is because it's AI and then suddenly it's like, wow, it's amazing.

It life changing could be game changing.

So we kind of let, we are more willing to hear it's that kind of listening bias or behavioral.

But whatever the bias is, right?

Of we want it to be right.

So we're going to listen to whoever's in front of us because we think, wow, this is where we want it to get to.

Because we all love Star Trek.

Right.

Or we all love sci fi movies.

Yeah.

And I think one of the things that helps me understand because like I said, none of us know everything.

Right?

So I'm learning every day.

Every day.

It's terrifying.

As to how quickly this field moves, I always ask, I start by saying, why did you need AI to solve this problem?

And that helps me also understand how that person thinks, but also to learn some nuances because the person can come back and go, actually, this is why AI was requested.

Oh my God, that's something new.

Great, thank you.

Or very quickly realize whether it was just, I want to apply AI to everything.

So that's a good starting point.

The first thing, whenever I look at any problem that I solve or any use cases that we look at, the first question I ask.

I think in the past, Bhairav and I, we've evaluated many startups for our friends to invest into.

Like we had one in a few in London.

But first question I would ask the founder is, why did you buy AI?

Why did you think this use case required AI?

And many a times he realized that they did it so that it was an AI company, it wasn't an AI use case at all, and they didn't get the investment they need.

I think that's a good starting point.

The other thing that I want to just chip in here is to say, don't you have to be very critical of a lot of the studies that come out?

So just today there's two things, three things I read.

One was that, you know, in the Lancet has come out that doctors have become less good at diagnosing certain types of cancers through their use of AI.

It's a Lancet study that's done.

I haven't been through the whole thing, but, you know, reading the, reading the kind of the premise, it's saying, well, actually, people who are using AI have actually seemingly dulled down their own ability to diagnose the cancer because they're too reliant on the AI.

Whereas, and there's another article that's saying, you know, AI has become much better at diagnosing certain types of cancers.

But then you have.

So you've got to be very critical when you're looking at these, these studies because again, it's all about people wanting it to believe it to be true.

A great.

Another one that came out today was around developers.

So their big hype isn't helping develop AI.

AI can create code and AI is going to get rid of developers.

You'll have the unicorn billionaire, 1, 1, 1, 1, person unicorn startup.

Right.

They did another study based on all of that hype that came out and what they found was that initially developers would say it's making my life 30% or 20% quick, I'm producing code 20% quicker.

But as they've used that and over time that they're realizing is that actually I'm spending more time fixing the code that was delivered than spending.

So yes, I got code quicker but now I'm spending more time fixing this code.

So.

And it might help me do small things, but it doesn't, it's not able to see the whole end to end picture.

You can't feed, you know, massive databases or massive code sets into, into language models.

So the original hype was never going to need developers.

You know, Salesforce Sell, you know, fired all of the developers, right.

Then they quietly hired them all back because they realized that they actually can't do the job just with AI.

But they hired back as contractors, changing the balance sheet.

Right.

So the, there has to be critical thinking in all of the, all of the kind of articles that are put out there, all of the hype that's put out there.

Yeah, I think I'm going to touch upon that because that's a very.

One has to understand that in my view, and we've been in several keynotes on that, AI is not a replacement of a practitioner.

I want to be very clear on that.

If there's one thing I stand by and I say, you know, I think it's that AI is not a replacement.

AI is a power tool to work alongside.

It's a turbocharge the practitioner.

That's important to understand.

If I want to be, let's say, an excellent coder or a programmer or developer, AI is not going to make me an excellent.

It's going to help me turbocharge me to achieve my outcomes, but it's not going to make me a great engineer unless I go and study it and I practice it.

I apply some rigor to it.

It does not replace.

So give an example is that before AI tools came into play, we were doing things, we were putting in hours to do stuff.

A lot of work was getting done to try and get to the outcome, but a lot of work was getting done to get there.

So it took us longer to achieve our outcomes.

Our business is all about outcomes, not about number of hours spent.

What AI tools have done in the past years, because of them using, being used as a power tool has helped us achieve our outcomes by doing a lot less.

We cannot mix these two things.

We've got to understand a practitioner is still required to achieve the outcomes.

What we are doing is turbocharging them so they can AI can work alongside to help them achieve the outcome by doing a lot less.

Allow me to explain this.

And this is buildings.

Right when we we are all old enough to go back time when it took two years to build a house.

Today, how many of us would wait two years to get a house of our dreams?

I don't think anyone would.

Why is that?

Because the power tools came in and that allowed builders to build houses faster.

Now here's the thing.

Did that take less people to build?

Of course it did.

So one can look at it and say 50% lost their jobs because power tools came in.

But the question I ask is how many more houses are getting built compared to 50 years ago or 30 years ago?

Alan, there's a little interesting dynamic here.

As I look at the three of us, I appreciate the people listening in.

Don't have that.

None of us is exactly a spring chicken.

We've all been around a while.

But you have to even though you die right here, we don't even know anymore.

My point was that one thing that worries me about AI is that enemy can feel that they're a genius very quickly.

And the reality of experience and hard yards and learning your craft in the context of AI without AI.

One thing that I mean my living on for the last 10 years is knowing what are the right questions to ask and knowing what is the stimulus I need to put into the AI that I want.

I want to get way related to concerning with your case ideas.

The that actually to get the best out of AI, you actually need to understand what this is.

Without that you're hoping AI can come up with a magic answer to a problem you don't understand.

Yeah, I think take developers for instance.

Bhairav, your example.

You cannot replace excellent engineers with AI.

But what you can do is turbocharged them so they can achieve their outcome quick and a lot better because they can focus on what they're good at is being excellent engineers in building and designing systems.

What was happening before is without those power tools of AI they were spending a lot of time on doing and getting there.

Now they don't have to worry about it.

And the terrifying part of that is there is a lot of people that maybe don't realize how silly or stupid they are and think I can get this.

I can do amazing things really quickly.

But if you don't understand what you're doing, how will you ever know you've got a good answer.

Exactly.

So the Salesforce example is that you probably find that once they realize that AI is not going to produce great engineers, but will make engineers great, good engineers great, they hired back a subset of those engineers who they knew were really good.

So if they got rid of hundred example, they probably got back 10 and they achieved more.

Remember, it's all about outcomes.

They achieved a bigger outcome that they would have with a hundred because AI powered those ten.

I think my view is that with developers or any stick with developers, that's what we are come from.

They will be turbocharged.

Good developers will always remain good engineers and architects will always remain.

They'll even better be better and they'll be greater because what AI would do is be their power tool, would be their ally.

So I think the fear should not be AI will replace me fear.

The excitement is how great I can be with these tools, how great these tools can make me and focus on the craft.

I think there is a wider question there, Alan, as to, you know, how well are people when it comes to how good are people when it comes to critical thinking, right.

Understanding what the output is and then interrogating that output to see, you know, does this make sense or is it actually real?

Or that's that.

But that's a whole kind of a different area in question.

Right.

I think that's where that's another podcast.

But one thing I have seen when it comes to, let's say we stick with the development theme, is that you've got startup founders now who are using things like tools like Cursor or Lovable to build platforms.

And what they're finding is that, yeah, they can whip together some simple things, but then when they try to change it or they try to to adjust what they've built, it's taking them, takes them five minutes to build something, takes them, you know, three days to adjust it.

So in a weird way, what you're seeing is that there's a bit more people understanding the value of an expert who understands how to get it done and then it done properly.

Right.

One founder, I, you know, she had spent a month developing something which was, I mean, it was all right, it worked for the principles of what she needed to do, but it was pretty terrible when you kind of look at it.

But it was enough to get out of the edge.

So she saved money on hiring a, or trying to hire a CTO or get a developer in because she did that and it took her a couple of months, but if that thing gets traction, then actually she needs to bring someone on board who knows what they're doing.

Right.

You can't just because it goes back to what we've always said is where should you spending your time, resources and effort.

Right.

So I think, but it's, I think with any kind of technical, any kind of, you know, progress that we make as humans, it's always going to be there are people who will be critically thinking and then thinking about how can we take what we've got and make it even better.

And there's other people who are just going to sit back and say, yeah, fine.

There's a. I think that is an example.

What is true though is that before a startup needed 20 people, the outlay to go and go from zero to one.

Now, with AI powered tools, if you have the right practitioners, right expertise, you don't need 20, you could get away with two because the experts or the practitioners are freed up from the non value or low value tasks and they can focus on what actually gets them to the outcome quicker.

So what we could have achieved before, what we could achieve with 20 people in a month, we can achieve that with three people with some right AI tools in a week.

Now that's not a bad thing because that also means that another 20 of us could go away and start up our own companies.

Is again back to the building thing.

More houses are being built now because it takes less time to build.

So all the people who lost their jobs, there were so many other jobs for them to go to.

So actually we had more builders, we had more houses.

So that's what I see.

I see AI and AI tools, turbocharging practitioners, so that we've got more opportunities, we've got better products, we've got better enterprises, better application.

Everything is just going to be better because we can focus on what matters.

Let me challenge you on that one.

Everything is going to be better.

What shook my head there, Amelia, and you were chatting, was we can do more and more, faster and faster.

So sometimes it just takes time to do something good.

You know, fast food arguably isn't superior to slow food Culture takes time to develop and grow.

Quick answers are seldom as good and well considered answers.

Do we think fast food has a market?

As a market, but it's not the answer, it's a little bit of it.

Okay, so let's look at it in a different way.

So fast food has a market, but it's not the only market.

It has not Killed gourmet.

So the question is horses for courses.

Because we are able to cook faster, produce fast food, there is an option.

Someone wants to go away quickly, get something, you can go in 20 minutes, you've got your food, you're out.

Is that the only thing?

No.

But similarly, is gurmed always the answer?

No.

No.

That's the thing.

What I'm saying is that AI will give us the opportunities, it'll open up.

I love it when you give me these kind of analogies.

There's a simple non technical person.

I can understand that.

And I know we're coming towards the end of our time on this podcast.

I think we might be having more on this that I find utterly fascinating.

I've written down in my notes.

The first thing I wrote was you're not reassuring me.

And the next one I wrote is, you are reassuring me.

So what I wasn't being reassured on is don't worry, everything is okay.

Because it's not okay right now because it's all in development.

Because things are changing.

We're learning as we're going, we're learning very, very quickly.

But to learn we've got to make loads of mistakes and we're making loads of mistakes all at the same time.

The bit I find reassuring in this, as somebody who will always be with the best of the world on the outside looking in rather than not my world, this is just history repeating itself.

This is just so many advances where loads of mistakes were made before the world.

What I can do with this, I'm immediately thinking of a couple of things.

I can always remember when businesses started adopting computers.

Instead of rooms full of people with pens and bits of paper and typewriters and like in those days lots of businesses bought computers and said look at this amazing building, full of real spinning about laughing like aren't me fabulous.

And then you realize they haven't bloody helped in any way at all.

We can solve the duff here.

It's not helping.

The other bit I remember is that famous line, nobody ever get fired for buying an IBM and I remember that of, you know what?

If you can go out and find and I'm sure IBM showing some ducks along the way.

But if you could go out and find the right people with the right tool that can actually help you, this is a brilliant opportunity.

But be really careful of Sneako children.

It will promise the earth and it will be you that pays the price on installing all those witty things.

And then realizing this doesn't help.

Alan.

I think I'll add to that I don't think, and like I said, I could be wrong, but I don't.

We are making the same advances and the same learnings at the same pace as it was happening before.

The difference is today everything gets played out in the open.

And that's why the perception is we've got to move faster today.

Yes, we have to, because it's out in the open.

If you're not first, you're last.

Because the mistakes and the learnings were happening behind, away from the public glare.

One didn't get to know.

So when something came out, it would work.

That's exactly what's happening today.

What?

Today we have complete visibility to all the failures.

How much of money has been lost in research or other domains?

Billions.

We don't know that it doesn't touch it because we didn't see it happen.

Today we are seeing companies sink because they made the wrong investment or the wrong bet, which we didn't see before.

So what we see scares us.

What we don't, we don't care.

That's human nature.

So just to wrap up there, because we need to close off this session again, just my thoughts on that is that I remember 25 years ago, 30 years ago, we were always talking about digital transformation and how we need to digitalize companies.

And we still run digital transformation projects, right?

I mean, every year we're running digital transformation projects.

So this thing is never ending.

But it's as is this podcast.

It will never end.

So we're going to have another session of, of this at some point.

So I guess what we're going to move on to next is a little bit more practical around what is AI, different types of AI and looking at how you can think about it, implementing it within your business.

So for now, thanks, Andrew.

Thank you.

Thanks a million.

That was absolutely wonderfully insightful.

I feel I actually come out this feeling more reassured that it's not a magic mix, that actually is history repeating itself with a new technology.

Thank you for having me over.

Any opportunity to.

I learned so much from you, Alan, as well.

Every time I'm a pair of questions.

So insightful.

I'm going to go and learn some more.

Cheers guys.

Thanks a lot.

Goodbye.

Many thanks for taking the time to listen to the podcast today.

If you like what you hear, please leave a review or subscribe.

You can find us on Apple Podcast, Google, Spotify, all the main podcasting sites.

And if you'd like to learn more from us, you can contact Alan.

He's alanbgp.co.uk and you can find me at info atomcto.com.