From Automation to Autonomy with Ismail Amla

AI is moving beyond automation and into autonomy and that shift could fundamentally change how businesses operate, how decisions are made and how people work.

In this episode of The Entropy Podcast, Francis Gorman sits down with Ismail Amla to explore the rise of agentic AI, digital workers and AI-native organisations. They discuss why the biggest opportunity is not simply using AI to make existing processes faster, but redesigning businesses around what intelligent systems can now do.

They also examine the harder questions: what happens when AI agents begin making decisions independently, who is accountable when those decisions go wrong, how organisations can avoid agentic drift, and whether relying too heavily on AI risks weakening human judgement and curiosity.

Key takeaways: 

• Why AI is shifting from automation to autonomy
• How AI agents could become part of the workforce 
• Why most AI transformation efforts fall short
• What leaders need to rethink, and why governance and human judgement matter more as systems become increasingly autonomous.

Soundbites: 

“I think it is a revolution. I don’t think it’s just another tool.”

“Whoever this technology is working on behalf of is responsible for every decision that’s made.”

Transcript

Copy Transcript Copied

Francis (00:02.705)
Hi everyone, welcome to the Entropy Podcast. I'm your host Francis Gorman. Before we dive in, if today's conversation challenges you, sparks a new idea, or sharpens how you think about the world, don't keep it to yourself. Subscribe, leave a review, and share this episode with someone who enjoys staying curious. Today I'm joined by Ishmael Amla, who has spent more than thirty five years in business consulting and the IT services industries, including leadership positions at NCR, IBM Global Business Services, and Accenture.

Ismail (00:31.84)
As if it's going to be a general way, and so I mean more time to alternate. Thanks for having me on, Francis.

Francis (00:32.529)
Ismail sits on several boards including the Network Rail and is a fellow at the University of Salford Business School. Ishmael is also the co author of the book From Incremental to Exponential, How Large Companies Can See the Future and Rethink Innovation. Ishmael, it's lovely to have you here with me today.

Francis (00:52.371)
Ismail I I kinda wanted to d jump into something that I we talked about ourselves be before the show and I was thinking about it and and I haven't really touched on it in the podcast today. And it's kind of diversity and identity and and and you spoke to me a bit about, you know, you spent a decade in New York and then you moved back to the UK and it's only as you got older and more wise in your years that you started to think a bit more about identity and diversity and how how that fits in. Can you

Ismail (00:56.43)
Ismail (01:01.516)
Yeah.

Ismail (01:05.694)
And it's only as you got longer and then numbered the works for me to talk about that to be because I think we understand it will be more thing here as a as a series of questions that they want to say.

Francis (01:21.471)
Can you talk to me a little bit about that space? 'Cause I'm really interested to kind of understand it from your perspective and how you see the world and how the world sees you as a as a serious profession in the technology space.

Ismail (01:33.102)
Yeah, and it's it's it's it's it's a great question and as you say, as you get older you think about it more. And the and the turning point for me was when I was being interviewed to be a member of the UK Sports Board, and the UK Sports Board look after elite athletics across the UK, essentially ensuring that we win medals in all of the elite competitions, especially the Olympics. And and I was sort of preparing for it and I realized that actually

And I'm I'm British born. I'm born born in Bolton in in the Northwest. my parents came from India in the sixties. And I realized that the first time that I actually felt British was during the 2012 Olympics, where Mofara won a gold medal. And you'll probably remember he was running around the track with the Union Jack being embraced by the British public and the world as being the British Olympic champion.

And that was the first time and I was forty seven. that I think I was forty seven, around I was in my forties and I realized that I was felt proud to be British. and it you know, the power of sports and inclusion and community and what can that can do to help make you feel included because it didn't matter whether I felt British or not, it only mattered when everybody else thought

I was British and I was included. And Mo Faro was included and he was accepted as the British gold medal winner. yeah, though that that's what we were talking about. a quite a profound moment for me actually, in understanding a little bit about myself and about what it means to be included.

Francis (03:18.049)
It's really interesting and like when you say you're in your forties, does that mean you spent twenty odd years of your professional life feeling like you were on the outside of the of the company and not in the in the center? And you you talked to me about your time in Accenture and having to drive the inclusion and diversity kind of program and that took you out of your comfort zone a little bit. And you know, did that help drive some realizations?

Ismail (03:31.502)
Yeah.

Yeah, I mean you you're absolutely right. I mean I think I have always felt like the outsider, you know, I didn't go to Oxford or Cambridge, I didn't live in London, I wasn't born in London, you know, I'm a very Muslim guy, in case you haven't noticed. So I always fight felt like you know, I needed to prove that I was just like everybody else who was in the community. And I remember when I was at Accenture and my boss boss then Liz

Who was the UK CEO decided that I would be a great diversity and inclusion lead for UK and Ireland. And I felt that was the biggest, that was the worst thing that could ever happen. You know, I was gonna be the brown guy talking about inclusion and diversity. But as it turned out, it was one of the most incredible moments in my career because it forced me to think about the value of people, actually.

It forced me to think about my platform and what I could and couldn't do. And also, it was at that time, and you know, fair due to Liz for seeing this, it was at that time that I felt started to feel comfortable in my own skin. you know, I was fairly old by then, as you can imagine. And what happened is when you feel comfortable in your own skin, Francis, you you know, you're authentic, you perform better actually. So there's a business benefit of this. And so for me, it became really clear about

in my as I went more into more senior leadership roles, how important it was, you know, if you take away the the the tag around diversity or inclusion, but how important it was to allow people to be themselves in work, at home, wherever they were, because only in that could you really get incredible business performance. And it was a lesson I learned over many hard knocks really, but thank you, Liz, if she that if she's listening.

Ismail (05:32.279)
But that was one of the best decisions that anybody ever made for me actually, appointing me into that role.

Francis (05:37.667)
it's great it's great to hear it and I suppose reflection is also important. It it allows you to see where the the lessons were. If if there's some young guys listening to this podcast that have emigrated from India or have spent their whole life in the UK or Ireland, have you any advice for them as they set off on that pre professional ladder as to, you know, be you?

Ismail (05:47.351)
Yeah, we'd put them all like that.

Ismail (05:58.178)
Yeah, I mean I think that's it, that's the advice. Try I mean you'll you we we all go through, you know, all sorts of questions and reflections on ourselves, and at some some stage we land with what we think we are. And I would say try and accelerate that journey because you know what you are. You know what you like, what you love, what your purpose is, what your passion is. At least accept that for yourself.

That's a starting position. And the sooner you do that, the sooner you'll get comfortable about liking yourself and you know, get mix sharing with others. And the amazing thing is other people want to know you, the authentic you, the real you. and it's such a powerful moment for me. And I would urge people and I do, you know, when I talk to some of the our grads coming through, I do make that point that regardless of who you are, where you're from, what background you're from, gender or anything.

Get comfortable with yourself. 'Cause you know, you if you want other people to like you, you better like yourself. And that's what you should start.

Francis (07:05.341)
I think that's great great advice, Ishmael. And you it's it's kind of sobering to hear hear hear a life lived and you know that these lessons seem so obvious, but they're they're really profound when the when they do land. And look I I really appreciate you kinda talking to me. I I did want to just touch on that for a minute before we we talk tech, but let's let let's let's talk a bit of let's talk a bit of tech as well. I wan I wanna touch on AI and not because everyone's talking about it and the airways are saturated with it, but

Ismail (07:10.701)
Okay.

Clearly.

Ismail (07:19.106)
Yeah.

Yeah. And I don't remember the thought of never have to have it.

Francis (07:35.261)
More so in terms of AI as another technology transformation. Do you think the fundamentally what has happened at the moment understates the reality of AI? Is it just another tool or is it far more I suppose fundamental than that?

Ismail (07:44.407)
Wonder to take a lot. Is it just another picture as if

Yeah, I think it is it's a revolution. I don't think it's just another tool. I think it changed everything, including society, enterprise, commerce, macroeconomics, global trade. I I honestly believe it changed everything. To the point where I don't actually even know whether we know where all of this ends up. And I think a lot of the leading lights in the world are saying the same thing.

I'm you know you can be really excited about it or you could be really scared about it. I choose to be excited and optimistic about what it brings to the world. but it is it you know, if you don't think this is gonna be a revolution for individuals and business and families and societies, even religion, then I don't think you've been paying attention.

Francis (08:45.021)
I I tend to agree there and I you know, I was reading the business post, the Irish business post, not the not the UK business post, last week and there was a headline that thirty five thousand tech jobs have been lost so far this year. You know, and you have to kind of extrapolate that over what's happening with artificial intelligence. And then I was talking to Schnee Hollandani, the CEO of Horizon Three, and he he made a very good point. He said if you if you're not curious about what you do and if you just go to work

Ismail (08:50.219)
Yeah.

That's

Ismail (09:01.421)
Expanding on top of the narrative.

Ismail (09:09.601)
Hmm. Okay.

Yeah,

Francis (09:14.633)
to, you know, go to work and you know, it's mundane and it's repetitive, good chance AI's coming for your for your job, you know? I I'd I I I'd like to get your view on AI, workplace. Where do you think it's gonna go?

Ismail (09:18.421)
that's good.

That's right.

Ismail (09:30.507)
Yeah, so I don't think actually that AI will replace people. I think AI will replace people who don't have AI tools. So I think that's point number one. I just read somewhere of IKEA introducing AI and removing hundreds of jobs, but not exiting those people, but

making those people interior designers with a bottom line result of I think it was 1.6 billion in additional sales or something. It was some crazy number like that. And I think that's the first of many examples we'll see where organizations are thinking strategically about what it means and what value it could mean and what you could do for your business. This is not I mean, I think the short term view is that this is a cosplay. We automate stuff, we get rid of people

I think the smart organizations are thinking about how business and society and governments change as a result of this technology and how people can do more more augmented activity, get better service. and that's where I think a lot of the opportunities will lie. We're we're working with one government, for example, who are wanting to create a native AI government, and they have, I don't know, forty to fifty thousand civil servants.

And an a number of those roles will disappear to be something different, not exit to be unemployed. But if you can if you think about what that means in in terms of getting service from a government which is frictionless, which is proactive, you know, you're not filling in a form to go and get a passport, but the passport comes to you at the time that you need it, when you need it, all the ID and certification and security.

things are done automatically in the background, all decisions are made, there's transparency around the decisions so we know how they're being made and so on. I think there's there is opportunity that I think will benefit mankind as long as we've got guardrails around what happens when things go wrong. And we've seen already seen things go wrong, right?

Francis (11:50.548)
No, we have indeed. And I think the IKEA example is the the pivotal, you know, this is where you can go. And you know, it I suppose it's it's almost aspirational, but they've they've proven it to be reality where they took the customer service staff, retrained them in in that front face and service, that human touch, and yeah, I think one point six billion of of a profit as a result in nobody lost their jobs. So, you know, they kind of doubled their workforce in a way without hiring more people, you know, which is I wanna see their token spend

Ismail (12:03.021)
Yeah.

Yeah.

Ismail (12:14.092)
Crazy. Amazing. Amazing.

Francis (12:16.671)
Hopefully they're talking spending this in eight hundred and ninety million or something, you know. That they're not burning through. Where did that where did that come from? How how is that formulated or like that seems very futuristic?

Ismail (12:19.838)
Yeah, yeah. Yeah, who knows? Who knows?

Ismail (12:31.745)
Mm.

Yeah, you know, it's two years in the making actually. And we we were just talking about the use of AI at that time. It had just been formed and a gentic technology had just come onto f being. And then the the government exam question was two things. One is sovereignty. How can we protect ourselves against

us using other government's technology if we need to do that. And secondly, what everybody is saying is to be able to use AI efficiently, then we need to do a lot of work with our data. And the approach we want to take is we want the technology to come to our data rather than us giving the data in a data lake or a data mortal, whatever it might be. And a Gentic at that time offered the possibility of that might happen.

But at the time, you could only have workflows of between one and ten agents. And so we were talking about, you know, how do you do? And then suddenly in three months, you can have infinite amounts of agents, you know, because the technology is moving so fast. And so they started experimenting with what can agents do that can be transparent, that can be audited, that can have controls around it so it doesn't hallucinate. and we we created the thing policy as code.

which allowed you to have all of those things and the government got more and more comfortable a l around implementing an infrastructure and they are going all out to be a native AI government as in they're designing their government services with outcomes in mind, not how they do things today. And there's a great example they had they had a process of if you were a new business coming into the country and you wanted to open a new a new business

Ismail (14:25.782)
Then the licensing time took something like I can't remember exactly, but let's say 12 weeks or something like that. And that went in a native AI process from 12 weeks to five minutes. Because you know, all the processes could be done background and agent decision making, all that things could happen. And so the benefits of the government in terms of attracting businesses into that community, of course, are profound. and so that so they begin all they're really beginning to see.

quite a bit of benefit f from a society perspective.

Francis (15:02.229)
No, that is that's really insightful. I I think I want an offline conversation on that one to get really into the in into the weeds, but but it does kind of st you you said something there. Designing with outcomes in mind, 'cause I I I kinda hear a lot is where can we add AI? You know, you know, where can we put AI? d that takes a lot of time. Can we stick some AI in there, you know? But this outcome driven, so is there an entire shift in mindset required

Ismail (15:06.578)
Yeah, you think we're just going to scream.

Okay.

Ismail (15:18.816)
Yeah. Like all the nothing about like

Ismail (15:28.34)
It's raining from the fire. That's what you're going to do with that.

Francis (15:29.855)
And that's probably why we've seen for the last years a lot of pilots, a lot of failure, not a lot of outcomes. Do we need to do we need to pivot our mindsets to look differently at the possibilities of AI rather than just trying to hockey AI into something that already exists?

Ismail (15:34.988)
Yeah, yeah, I I think that's exactly the shift is going on. And a lot of people have thought it started to talk about it. But there's a couple of other, I think, founding things, and Saturn Adela talks about this in this reverse paradox essay he wrote a little while ago, where he says if

actually you're using the LLMs everybody else is using, then you've got no competitive advantage. And by the way, what you're doing is you're training the LLM with your data. So point number one is you need to create a moat. Be very careful about what you are giving out to the public in terms of LLMs in terms of depth learn. So there's a moat around what your propriety data is going to be. And then the the other foundational thing if you like is

This idea that

you know what what are you gonna bet on has gotta change the economics dramatically, not fine-tune to get twenty percent, thirty percent savings. If you have those two things in mind and then you know the the idea and we just had run a workshop with one of the major European airline companies around what would what would an outcome look like around airline disruption if you were to manage it

in an AI native manner. And you know, the and the outcomes suddenly start to be passenger experience oriented, managing the rot roster of the crew, managing where the planes might be, getting people rebooked on other things. Those are all the outcomes. And then you build agents and graphs and loops based on those outcomes rather than how do we do passenger experience today? And how where could we add agents to do that?

Ismail (17:32.543)
And that is absolutely Francis where the shift is going on. And in fact, there's some data came out which said I think again, rough order of magnitude, twenty percent of people who've implemented AI are getting benefits. And the twenty p p percent of people who are getting benefits are treating AI as a different way of doing business rather than augmented technology on existing business.

Francis (17:54.262)
I that's really key though. It's and it's that outcome driven view. Start start at the end and then figure out what technology combined gives us the outcome. But we've we've kind of moved away from talking about chatbots and generative AI to agentic AI, you know, systems that can, you know, access databases, initiate transactions and make decisions. Are we effectively creating a new kind of digital employee in in in that sense?

Ismail (18:06.388)
Not only

Ismail (18:15.679)
Yeah. Hundred percent. I mean, you know, we're we're now talking about an IT individual in an IT organization really being the HR lead for agents. So they train agents, they introduce agents, they coach agents if the agents are hallucinating and behaving badly, they might replace agents by brand new agents who's coming on board.

I and in some instances organizations have already started to put together the HR department leadership and the IT department leadership together. I mean that's massive change, right? and we're beginning to see that as well. So I do think that agents will become digital you know, people if you like, and all of us will have to get used to to working in teams where some of our team members are agents.

that we engage with. That also then drives another discussion around how do you get me as a human ready to work in a team full of technology agents. I'm used to working with Francis and Aidan and Tom and Bill and we have a chance we have a we have a banter on a Monday morning around the football or whatever it might be. And that won't happen anymore. And but that was a really important part of my work day.

And so there's a there's a there's a extension of the change management and the people readiness aspect that's happening right now as to how do you get people and teams and individuals ready and what are the you know what are the what are what are the processes you need to take the individuals through. already in that there's two brand new roles that have been created. One is the human systems architect, so this is the

a person who into ensures that the person working in hybrid teams is ready, trained and gets whatever they need to work in that team. And the forward deployed engineer that you will have heard of that everybody is talking about, which is the engineer who sits with the end user and builds the AI rather than say I'll I'll be back in three weeks, haven't built something. So even the i early days of a genetic, which is where I think we are, we're beginning to see brand new roles pop up that will be needed.

Ismail (20:33.695)
To be able to make this new world a reality.

Francis (20:37.525)
I'm just thinking of the psychological shift it takes to go from having a conversation with Ishmael and you know, a bit of crack, you know, what did you do at the weekend? you did that, did you that's to that is outside of my guardrails, please you know. And the the the the the thing the thing that I'm coming back to though isn't necessarily the psychological aspect, it's the decision making aspect. So obviously we all have roles and remites and a conscience to know

Ismail (20:42.623)
Yeah, yeah.

Ismail (20:50.372)
Hello.

Francis (21:07.687)
I can blur this line slightly and make this decision and I'll be forgiven. But an agent may not necessarily have that intuition that is human. Is decision making gonna become I won't I won't even say blurred, but extremely complex?

Ismail (21:07.903)
Yeah.

Ismail (21:14.315)
Extremely complex. And and you know, this I'll give you an example through this idea about agentic drift, right? Which is the agent is behaving as designed, but not what it should be doing. And so we'll came we've come across a few of these examples. And in this instance, you've got a customer contact agent, and the customer contact agent has been given a target.

To improve the NPS scores of engagements with people who are calling in with complaints. And so somebody calls in with a complaint and the agent said, Look, I'm sorry, you know, we won't do it again. And please mark me on an NPS as you leave. So he gets a mark of six. Next next person calls in and said, This happened if you know, and the agent said, Look, I'm really sorry, I'm gonna give you a bit of a refund. Twenty percent off what you paid.

Please give me an NPS mark and he gets the mark of seven. And the agent thing, that's interesting. Next person calls in and he says, Okay, I'm gonna give you a full refund. So he gives a full refund, he gets an NPS of ten. The agent is in his element, right? So he's given full refunds because he's hitting ten NPS scores. So this this agentic dress is behaving exactly, but you he's gonna bang rot the company in behaving in this way. And don't forget.

This is all happening in quick, fast time. There's no person involved to check whether this agent is behaving at this stage. So it gets really complex because how do you implement purpose into this agentic behavior, not just objective? And so it is, and you know, people talk about humans in the loop being adding meaning to the loop, and this I think is going to be really important. How do you architecture meaning into a decision that you want the

the the the agent to take. And also how do you do it in a manner where by the time the human helps with the decision making, the damage is not already been done. Like this case. So yeah, it's gonna firstly it's gonna get super complex. And I think there's lots of approaches underway to try and manage that, but it's not a it's not a space which we've fixed. It it's work and work underway, I'd say.

Francis (23:37.674)
I remember when I did my first threat model on Agentique, there was three things that that came up. It was speed, scale, and opacity. And you know, speed is this can happen in real time at real scale and it's black boxed to a lot of cases. So how do you how do you measure when everything looks like it's the wheels are turning in the right direction? You know, how do you measure it? You know, so that that gave me as a security professional But

Ismail (23:43.4)
We need to have the

Ismail (23:53.226)
how do you imagine whenever we look like the current and it might directly how do you measure it. Yeah, yeah, exactly. Yeah, right. Exactly. I can only imagine.

Francis (24:07.391)
The the question that I wrote down at the time I remember and I and I remember putting a line under it was who is accountable?

And I don't think I've ever got to the end of that answer.

Ismail (24:20.498)
Well, so there's a real world example, you probably know, right? Where Air Canada. Did you hear about this Gen I Gen AI thing? So so somebody put in a Gen AI bought, you know, my my mum's died or somebody's died in the family. I need to make a a booking and you've got you know, will I get a refund? And the response was yes, you'll get a refund. Just go ahead and do the trip and then come back for a refund. Person does the trip, comes back for the refund.

And then it kind of said, actually according to our policy, you don't get a refund. And and the guys by the Gen AI bot said I'd get a refund. Yeah, but we're not responsible for the technology. Anyway, long story short, the the the court found that actually the company was responsible for the technology and the decisions made by the technology. And I think that is what everybody has to work for, which is whoever this technology is working on behalf of is responsible for every decision that's made. That's a scary thought.

Francis (25:18.591)
That's the answer. So I'm gonna I'm gonna write I'm gonna write the answer now when I finish this call and go th this is it from Ishmael Z. That's it, that's the that's it done. And the Air Canada case is my my my go to guide on it. Yeah, I d I think it's fascinating, so it's yeah, it it is something that I'm gonna keep mulling over in in the back of my mind.

Ismail (25:20.03)
Yeah, basically.

Ismail (25:38.11)
Yeah, absolutely.

Francis (25:40.518)
One one other thing, Ishmal, when when I look at this, you obviously work with massive companies all over the world. And massive companies that have existed for any more than twenty years all have massive amounts of technical debt, huge amounts of bad data, layers of complex processes and procedures, etc. How are they going to compete as the agile, fresh, technology centric company emerges

Ismail (25:48.794)
You know what I'm saying here and talk about that's about taking them.

And make it

Francis (26:10.281)
And the legacy constraint holds them back. Is there is there gonna be a need to split out organizations?

Ismail (26:16.936)
Well actually, you know, that's historically that's what's happened and that a little bit of what our book was about back in the day, and that book is way out of date, so don't buy it. It it's it's five years old, which means it's totally useless now. But we did talk about that at that time. but I think what's happening now, which is which gives hope, is that if you can imagine anything that exists in physical form or even digital form, for example, a policy statement, a

bit of code, some old technology written in Fortran, anything that exists, can be reverse engineered by agents to become machine readable. And then can be forward engineered to become modern. And again, all this technology is like nine months old, right? So this is brand new and suddenly it gives opportunities for organizations who who've got all this technology there and thinking how am I going to use any of this agentic.

To be able to actually start with modernizing what you've got, because suddenly you can do it. You don't you don't have to rely on people, most of them who have retired you know, or don't are not around anymore at all if they're that old. But the agent agents can, you know, we we can we can reverse engineer and extract and discover and do analysis and organization charts and old code, even machine executable code, all sorts of stuff and get it machine readable policies.

And then based on what we've got and any reference architecture we could forward engineer. So I think actually I what we're recommending as we talk to organizations which you know all by the way, large population Francis have got more technology debt than anybody you can imagine, right? And they've been holding on to it because they don't know what to do with it, have suddenly started embarking on this modernization track, which said actually for the first time I can in a risk mitigated way.

start to look at what's running in production and have a means to get it into a modern state.

Francis (28:23.059)
Yeah, I I th I think it's it's it's probably a headache for most CEOs and board members at the moment, just looking down and going, You know, we need to do this AI thing, this this magic box, but how how how do we get there? W what what do you think will distinguish the companies that genuinely win in the AI era from the companies that just simply spend a lot of money on AI?

Ismail (28:31.081)
We can do the idea that's magic box.

Ismail (28:41.741)
in that morning I AI. Yeah, I think the first thing we've seen is the commitment at the board level and a very clear view on the board as to why and how they're gonna use AI. So that posture I think is is really important. And actually it's not that different from how do you do any change, which is senior level commitment, but actually it needs to be explicit. And then s

you know, concentrating on a number a small number of bets which change the game for that organization rather than let's do five pilots to see how it works. Because all pilots will be successful. Maybe one of them will go into production and none of that none of that works. so that that's a couple of things. And then there's a couple of cultural things. One is that this world is designed for speed.

Right. As in you've got to build stuff 'cause things are popping down the road all the time. You've got to build stuff, put it into production, see what you learn from it and put it back into production. and so you need a culture to support that. And, you know, in some of these large organizations that you and I w have worked in and working, you know, the culture is actually not built to do that. So that's I think a big change that the board needs to support.

And then finally I would say governance has got to be very clear, as in who has accountability, who has the the funding rights and return expectations, and who is responsible for ensuring the culture is aligned with the posture that the board has set. I think and there's lots of technical things that need to be done, but I think from a business leadership perspective, those are probably the big things.

Francis (30:32.253)
That's great insight, Ishmael. I I know you said your book is out of date, but I I'm gonna ask you about the book and not specifically about the content of the book, but about the process of of writing a book and the depths that brought you to and the lessons learned that came out of those depths. And my question is kind of future. If if I take a look at the future and go, we're handing off a lot of our mental burden to AI.

Ismail (30:38.173)
Yes, I'm not a walk.

Ismail (30:49.095)
Yeah. Hmm.

Ismail (30:58.291)
Mm-hmm.

Francis (30:59.605)
There's a lot of value in putting in the hard work to get the the rough diamond and polish it up to become the book per se. Do you worry that companies may get this wrong to an extent where our capacity to innovate as a human may be diminished by offloading our our mental state to something else to do the hard work and not actually falling down and going, that doesn't make sense. I made a mistake there. I need to figure this one, I need to tweak that.

Ismail (31:01.329)
And the number of marks and the profit and what is not the hook for it. Yeah.

Ismail (31:11.593)
Yeah, that's what you think I think.

Ismail (31:21.745)
Yeah, actually I can't follow it.

Ismail (31:26.249)
Right.

Francis (31:29.011)
And then understanding how you got from the start of the journey to the end, which is the book in this case.

Ismail (31:34.299)
Yeah, yeah, yeah. That's a great question. And you know, everybody always brings up GPS in map reading. And I can tell you, I can give a map to f my four kids who are in their twenties and I think they'll all struggle to get come to terms with taking me from here to Manchester. because of GPS, right? Nobody needs it anymore. And we've de-skilled ourselves to the point where we can't do some basic stuff like that. And and I think that of course absolutely applies into the world. and and I think

into the world of AI. And I think if I think about the book and it's a great analogy, you know what where that started was I I at that time had just left IBM Global Business Service that I was running North America consult. And I ha had a really profound belief that actually innovation is best suited to large organizations because they had access

To resources, they had channels, they had customers already existing, they could afford to make mistakes, etc. etc. I had a very close friend who spent the last most of all of his career in the startup world, and he had exactly the opposite view and said, actually, all the big giants are gonna die, and the small guys, because of the agility and so on, will survive. And so we decided to write about that. And so we were this process of bantering and and

t challenging each other's ideas, of researching, of reading, interpreting, you know, we'd read the same thing and the two of us would interpret it in different ways and we'd say, okay, you know, does this prove your point or does it prove my point? And you know, in the world of creating, and of course AI also offers books, all of that would be missing. 'Cause 'cause the AI would do all of that for us. And I've and I have to say the the

you know, that that does leave a loss for people who want to go through that process and and it's hugely beneficial, of course, because you learn so much more than you set out to learn as you go through this. and yeah, I do I do worry about that. I do worry about us being de skilled in particular areas. and I think again back to the the point about you know I see opportunity, I see as getting reskilled in other areas that might be more interesting.

Ismail (33:53.275)
and IKEA being the example, right? People being reskilled in interior design. So that's my hope that people get the opportunity but to to reskill into areas that are more aligned with what they want to do. But there is definitely a de skilling thing, which maybe maybe we'll get used to just like we got used to not being able to read a map 'cause we use so on GPS and you know, maybe we'll get used to it. I don't know.

Francis (34:19.219)
You know, I I I think about this one a lot and it I've talked about it a lot on on on the podcast for the very reason 'cause I was talking to Raheem Hirji, he wrote a book in in in kind of the AI, Super Skills for the Age of AI, and he was talking about the algorithms and he talked about Google Maps. So we'll go back to the map analogy and he said when I type in workplace to Google Maps, you know, it gives me the fastest route. But I never ask, is that the fastest route for me?

Ismail (34:22.855)
I thought the icon.

Francis (34:47.773)
Are the best route for the city, let's say, because of traffic conditions and they don't want to route people a certain way. So the curiosity of understanding why the algorithm decided to make that route and what did I lose as a result of it? Did I miss out on a nice drive through the countryside to explore different scenery for my senses? Did I did I miss out on on something else that I could have saw somewhere else? So

Ismail (34:48.649)
Because of traffic and don't want to have people. I mean people understand it why the algorithm has group. I'm not that looking at that, but I think it's up with a nice point for

Francis (35:16.115)
When I think about this I think about curiosity and and I and I I try to encourage people on my team not to not to lean into AI for everything, but to stay curious and to question and to understand why a certain thing is telling you to do something and I think I think it's a debate we need to have as a community.

Ismail (35:17.353)
I agree, I agree, I agree. I think and I think to that point, transparency, auditability, how decisions get made become really in important.

And us insisting on that as a community of users, I think becomes really important to be able to understand exactly to your point around, you know, why and how and and staying curio curious. But I think it is a it is a discussion we need to have ongoing as a community. I totally agree with that.

Francis (36:07.571)
Ishmael, before we we finish up, you know, you you've spent your entire career in in technology. You've lived in the States, you've lived in the UK, you you've you've sat on a number of boards, you have a large viewpoint of of what's happening in the world. Is is there anything that you think we should talk about before we finish up, just as your years lived and your and your experience today, it is kind of egging you to go, I need to say this out loud right now.

Ismail (36:12.785)
Mm.

Ismail (36:28.498)
Yeah, yeah.

Ismail (36:32.934)
I think I think the one thing that I do worry about and I don't know where it ends, is this discussion around sovereignty that has come up over the last couple of months. Of course, triggered if you like by the Cloud Act that was passed in the US, which essentially says, you know, they can switch off services to organizations outside of the US and and they can instruct US

organizations to switch off services to organizations outside the US. And what that means for data sovereignty, technology sovereignty, software supply chain sovereignty, sovereign AI, i.e., what models do we use, I think is really profound. But also I think that a societal impact as sovereign nations think about what that means for them as to

How much they want to partner, trust, and be reliant on the rest of the world, as much as it had been in the past, right? Globalization was a great thing, is it a great thing right now based on everything going on. And and I I you know, I think about that quite a bit just recently, and I don't know what the answers are, but I do think that technology is gonna have more of a profound impact than we think it might.

Because the add on things such as sovereignty, such as the EU AI Act coming into force I think this week, even maybe, is gonna drive governments and businesses to behave differently than they have in the past. And I don't think we fully yet see what that means for us as individuals and citizens.

Francis (38:20.701)
There might be another book in there, you know? Ishmael, it's been an absolute pleasure to have you on. Really insightful, great conversation, thank you very much.

Ismail (38:22.088)
You never know.

Ismail (38:28.146)
Thank you.

Thank you for having me, Francis, really appreciate it.

Francis (38:33.47)
My pleasure.