AI x Future of Work

Atif Rafiq is the cofounder and CEO of Ritual, a software app to speed up innovation and move ideas to action. He has held C-suite roles at McDonald's, Volvo, and MGM Resorts. He was previously the cofounder/CEO of Covigna, which is a content management software provider. He's the author of the book Decision Sprint, which was featured on the Wall Street Journal's bestseller list.

(00:08) Defining the Future of Work
(02:06) Skills for an AI-First Workplace
(05:46) Evolution of Education Systems
(08:24) Impact of AI on Skilled Professionals
(10:30) Remote Work Effectiveness
(12:47) Equipping Teams with AI Knowledge
(15:46) Career Pivots towards AI
(17:49) Practical Steps for Problem Solving
(20:12) Value of AI Certifications
(21:27) Introduction to Ritual
(23:18) Surprising Discoveries with Ritual
(26:17) The Future Workplace
(29:09) Rapid Fire Round

Atif's favorite book: Thinking, Fast and Slow (Author: Daniel Kahneman)

--------
Where to find Prateek Joshi:

Newsletter: https://prateekjoshi.substack.com 
Website: https://prateekj.com 
LinkedIn: https://www.linkedin.com/in/prateek-joshi-91047b19 
Twitter: https://twitter.com/prateekvjoshi 

Transcript

Prateek Joshi (00:01.816)
Atif, thank you so much for joining me today.

Atif (00:04.918)
Prateek, it's a real pleasure to join you.

Prateek Joshi (00:08.667)
Let's start with the definition of the term future of work. I'm sure we've all heard it and it's being, it's used in many different contexts, but for this conversation, can you define what it means?

Atif (00:26.478)
Well, it's a great question, Prateek. I really center it on knowledge work, because in the end, a lot of the value created in companies comes from making sense of the things we know and doing high quality work to create new knowledge within the organization so that we can unlock the things that managers actually get paid to do, which is to make decisions and then be accountable for the results as a result of.

you know, taking those decisions. So for me, when we talk about the future of work, the reason why it's important is because if you, if you take a step back, you know, a corporation in the current form has only existed for about a hundred years. You have people like Alfred Sloan, which were the major sort of conceptual thinkers and the leading companies, you know, a hundred years ago started off as like a General Motors.

And then you had a Peter Drucker come along in the 1950s and talk about management theory and what is the role of a manager. And that has essentially been pretty static up until, you know, exceptional companies like Amazon came along, but really it's still pretty much how the fortune 500 operates is these models built in the 1950s. And the reason why we talk about the future of work is because when we talk about systems, whether, you know,

and now software-driven systems, primarily by AI, that really hasn't been factored into the workplace and how knowledge work take happens. And that I think is really the dawn of a new era.

Prateek Joshi (02:06.707)
And as we look forward, all of the news, the movement around AI, and we are looking at a future where we will live in some shape or form an AI first workplace. Now that means you could be building AI, you could be using AI, or you could be just part of your workflows. So if you look at

any professional, not just engineers, but sales, marketing, product, all different like knowledge professionals. What are the skills they should learn to stay relevant in this, in this new AI first workplace?

Atif (02:45.442)
Well, to me, it's really two layers. It depends on the kind of the level of value you want to provide. And of course, I'm going to start with the highest level of value, because I want your audience to be the most relevant and most highly paid. So but really, that comes down to very concrete things like defining the problem to be solved, defining the questions to explore, developing as best you can.

getting to the bottom of these questions through highly reasoned answers. There's not always data available. And then the ability to draw conclusions that allow for confident decision making or actions to be taken because we never have perfect certainty around decision making, but we can have higher levels of confidence. And, you know, that is actually the, I think the feature of knowledge of a lot of the highly

valued knowledge work. And so to make it more practical, you know, every company has some problem solving frontier where they're like, oh, these are strategic pillars. If we get these things right, we're going to grow. We're going to be relevant. We're going to get market share, those kinds of things. And then they break those down into initiatives or objectives. And, you know, things like OKRs are very common in companies, but the problem is that the O is all the way on the top left.

And the result is all the way on the bottom right. What's in the middle is a lot of things to figure out. And now you introduce AI into that equation where when you have to get into the details of messy ambiguity of solving problems, forks on the road, different options, AI is going to be, it might be an accelerator that, in fact, I think it's gonna be a bar raiser, meaning you will...

your contribution will be understood in terms of its value to kind of disambiguating these problems that the company is trying to solve in order to get results. So to stay relevant, I think it's really around the critical thinking skills and defining the problem to be solved, making sure that the right considerations are looked into, and then synthesizing that to what the companies should actually do. That is the highest value.

Atif (05:08.214)
It's very different than other uses of AI where you say, okay, oh, now there's AI for accounting. Sure, well, accounting is kind of Byzantine complex. There's a lot of rules. So yes, sure, you're going to prompt an AI to speed up and answer to a question, but that's like, that's much more tactical because it's knowledge that's already existed before. It's complex, but it already exists.

I'm talking about creating new knowledge for companies. That's actually what they're really, really gonna value. And I want people to be able to help skilled towards that.

Prateek Joshi (05:46.087)
Right. And how should our education systems evolve to prepare the next generation of professionals to live and work in this AI-centric future?

Atif (06:01.878)
Well, if we take it that problem solving is really this core skill set, then we have, you know, obviously we have a lot of problems in the world, so that there's no shortage of things to apply people to. But being able to have people, you know, define these problems to begin with, you know, you're solving, let's say, a sustainability problem or something with marine life or something like that. Of course,

how you define the problem is very important. So you're not just solving a part of the problem, but you're solving the whole problem. Let me give you an example. Let's take a commercial example. Let's say your McDonald's, a place I used to work, and the CEO says, hey, we should launch a subscription for coffee because that sounds really cool. And I saw it in another business and wouldn't that make sense? What happens in a company typically is

Half the people love it, half of them hate it, and they rush to judgment. They don't really understand, you know, and they haven't spent time exploring it. But, you know, if you look at that and you ask why, well, why would we even want to do that? It's because you'll get a higher level problem to solve, which is like, oh, well, we need to grow, and that could be an area to grow. And we don't do as well in coffee as we do in food. Okay, well, that makes sense. Well, why?

You know, why is this still a good idea to grow? Well, because coffee is growing and we're third in the market, not first. So now you begin to see, okay, well, are there other ways to sell more coffee? Well, you're not going to change the beans and you're not going to invent new flavors because there's a lot of that out there. So maybe we, as opposed to changing the product, we change the business model. Okay, so now you actually can define the problem a lot better.

and validate it's a problem that's worth solving. But coming back to education, I think the problem-solving skills, I think are going to be really very, very central to it. But also I think on top of that, doing that in the context of collaboration and teams, because collective intelligence will be very important. You and I, Pratik, working together on a problem is gonna go five times faster and better than just, let's say, one of us.

Prateek Joshi (08:24.795)
Right. And do you think if you look at the spectrum of skills, there are high skilled professionals, there are low skilled professionals, it's a spectrum society needs, needs all of that. But do you think AI will widen the gap between the high skilled professionals and the low skilled professionals?

Atif (08:25.174)
Thank you.

Atif (08:47.662)
You know, I think at this point in time, you know, it does appear that might be the case. I mean, we could have an outcome where the kinds of skills I'm preaching here and upskilling make people more valuable. But not only that, Pratik, you can create some level of attribution between the contribution and the result, which today is murky. If we talked about the coffee subscription program at McDonald's, you know, five years from now, we don't know, okay, we'll critique.

he had the breakthrough, the right question, or he really helped us reason through the really hard part. But in the future, all of that's gonna be living on some software workflow, which we're gonna use AI for. And as a result of that, we'll know that the contribution was really important. And so the result of that might be that the corporate knowledge worker gets paid like an NBA player.

And so if we're looking at, you know, a $50 million four-year contract for a knowledge worker, you know, then that will, of course, widen the gap. And that's obviously a very different issue on than how everybody gets taken care of.

Prateek Joshi (09:48.189)
Right.

Prateek Joshi (10:03.631)
Let's talk about remote work. And it's been a bit of a hot button topic. Many people on both sides of the camp, but where do you stand on remote work? Meaning, is it effective? Is it not effective? And if so, in what situations does it work well versus it just fails to function?

Atif (10:30.562)
Well, I think the cases where it works well are creating space to think, which can, you know, I think it can do a better job of that if people are, have discretion on their time and like when they actually do these tasks and when they actually do and think and put that all together. So there's some type of, you know, some of that coming from the flexibility factor, which I think we want to retain. But yeah, I think.

We need, we still have some things to solve for when it comes to remote work, which is basically making people feel, everyone talks about sense of belonging, but I'm not going to talk about that because I'm not talking about the soft sense of belonging, which of course I believe in, but I think companies are thinking that through and putting in different ceremonies and opportunities for that sense of belonging, the soft stuff. I'm talking about, you know,

the connective tissue you need when you're part of a team where this brain trust needs to be very effective in addressing something that's very hard, you know, where the answers are not obvious, kind of your core teamwork. And I think in the absence of some clear workflow of like, you know, what are we working on? Where are we against it? You know, we're not at execution, we're more in the exploration.

Here's what we're exploring. Here's how we're going to explore it. Here's when we're gonna like formulate our opinions. Here's when we're gonna align on it. See, all of these things are the reasons why the meetings happen. If you can't put that in the context of a workflow and you do remote work, then a lot of things break down in terms of trust, human factors, and then I think the quality of the work and the problem solving.

suffers because it becomes about, you know, in some of these trust and people dynamics. So I'm actually for hybrid work. I just think it needs to be paired with very clear workflow on how teams actually collaborate.

Prateek Joshi (12:47.731)
Right. And when you look at very large companies, like Procter & Gamble or Dow Chemical or Bridgestone, very large companies, and they've been around for a long time, they've seen many waves, and now they're thinking about, okay, we should equip every team member across the entire company with some knowledge about how to utilize this tool. So if a company is thinking about

Atif (12:50.574)
Okay.

Prateek Joshi (13:17.203)
of starting such a program. Like what would you tell them? How would you guide them?

Atif (13:23.298)
Well, I think the land and expand is the best way to go. So you say, okay, well, you know, let's take work, let's do this at a working team level. So there's an initiative, there's a group of people grouped around this common challenge, it's called the working team. And let's do a test and learn where we can compare, let's do an A-B test where a certain initiative has some.

some of this method or engine behind it. And we can talk about that, whether that's workflow or software or more AI driven approaches to doing their work. And then let's, those same people may work on two other initiatives where it's just the normal way that it has been for decades. And I think this would actually give a lot of information around the benefits. You know, the benefits you're...

you're seeking are, you know, higher velocity, higher quality, you know, outputs like higher quality problem solving, stronger recommendations, more confident ability to make decisions and fewer meetings. So those are the three things, you know, speed, quality, and lighter lift in terms of meetings and is the engine we're putting behind, you know, providing that. And I think that provides a lot of clear evidence.

And then what happens is because it's a lived experience by team members, it just grows very quickly. So that's how I would recommend it. And I think if companies are not doing that, it's really a shame because there's really no downside. I mean, the companies that are going to figure out like, oh, how do we put in place, you know, systems more AI driven, more software driven to actually drive our collaboration, those are the teams that are just going to grow faster. And when

Prateek Joshi (15:18.619)
Let's talk about career pivots, meaning there are many competent professionals. They might be at a stage where they've been successful in their own chosen area, and now they wanna do something in AI. Everyone's talking about it, it seems like future, but obviously their career trajectory has taken them elsewhere. So how can seasoned professionals pivot their careers towards AI-centric?

controls.

Atif (15:50.73)
Well, I think everyone can start by looking at the work that they do today and say, well, what pieces of this can we speed up where AI may be able to produce the first draft? And I think putting that into the kind of the setup of the team, I think is very doable. And I think it's very helpful. And I think what it unlocks, it kind of unlocks everything else because it's like, okay,

Well, let's establish a routine where, you know, this is, this is a core thing that we do over and over again. And for this step, you know, yeah, we'll produce the first draft. Now what's going to happen is we're going to kind of move, we're going to have a higher starting points or, or it's, it's like a marathon where you start at mile eight, right? Everyone would love that to show in the distance. And so where are those opportunities? I would really encourage people to do that. Because I think.

In terms of jobs and careers, like they're always subject to reinvention. So if you're an AI driven, you know, marketeer that works, an AI driven account and an AI driven financial operations person that works, right? So the kind of the race is on, so to speak, and the opportunity is there for people to be kind of the tip of the spear in their current field. And that's what I would encourage people to do.

Prateek Joshi (17:19.439)
Right. And earlier you talked about the things you need to do, like learning to solve problems, breaking down the objectives and figuring out what needs to be done. So if people want to be equipped to do that, on a day-to-day basis, on a practical level, what can they start doing today so that three, six, nine months down the line, they're able to do a lot of these things really well?

Atif (17:49.102)
I mean, let's say you're part of a team and you're at the idea stage or you have a raw objective. The company is giving you the objective for the quarter, for the year, that kind of thing. I think, and the next thing you want to do is have a kickoff meeting with the team and begin to define some milestones, maybe brainstorm some things. I think today, I mean, you can use the tools out there, the JetGBT and the others to basically

start with a more precise definition of the problem. So you can say, I'm working on a way to increase adoption for, you know, we've built a great product, but we don't have a great adoption. How can I define this problem in a way that makes it clear what we're trying to achieve? And the GBT will basically give you several variations of the problem statement that you can choose from. And then you can ask it,

Well, based on this problem statement, what are some key considerations that the team should be spending time exploring? Okay, great. Well, given those subject matters, what are some key questions we might want to look into? And the level of precision on this, of course, is going to be maybe 70 percent, right? Because it doesn't have the context of your company right now, and it doesn't know all the things that you know. But essentially, it's 70 percent good. And so automatically what you have...

is the ability to kind of lay out the canvas of where the team should be spending their work. So I think those three things of, you know, using software to better define the problem, understand the questions that need to be explored, and basically help you figure out where you're going to spend the time doing detective work to get to the bottom of these questions. That's where I'm spending a lot of my time through.

my new software company Ritual, but I think people can do this obviously some of that directly with the tools.

Prateek Joshi (19:53.896)
And when you look at the other people who think about getting a certification, like an AI certification, what's your view on that? Are these AI certifications valuable or are people just wasting their time?

Atif (20:12.311)
You know, I think I haven't looked deeply into things like the prompt engineering certifications and things like that. I think it can't hurt, but I think it needs an application, you see, because like, I think just the idea of raw interaction with a tool is just so unstructured. So what is it really? You know, that it could be.

you know, what are you actually applying it to? So I believe probably more in certification. If someone said there was a certification for product management or project management, which are clear roles with clear responsibilities and now you have an AI certification, that to me is kind of interesting. But I think the application to a field is really, it's that Venn diagram between the field and the,

AI sort of driven steps, you know, that you're taking advantage of.

Prateek Joshi (21:15.195)
Maybe it's a good stopping point to quickly talk about ritual. You mentioned earlier you're building the company. For listeners who don't know, can you quickly describe what the company does?

Atif (21:27.918)
Sure, so Ritual is a software product. It's an app, desktop, and mobile app that is designed for teams who are looking to build and run explorations. And so we say, well, what is the point of an exploration? Well, often, you are at the idea stage or the objective stage, and you're responsible for producing some recommendations. So let's say your organization says, hey,

you know, this is something we want you to look into. Come back in a couple of weeks with your thoughts and recommendations. Well, usually you're, you know, there's a lot of ambiguity involved and it's not so easy what the recommendation should be. So when you're building an exploration, you go through, you know, specific steps, what we call workflows, and we help you craft a problem statement, source questions, AI will produce the first draft, but the humans...

review these questions and make them better or add missing ones. And then we help you structure the workflow of answering these questions and with some AI-driven suggestions. But of course, humans are involved in that. And then you use the content coming out of that in the form of like a narrative or an FAQ to develop recommendations. So that end-to-end process of building and running explorations so that you can go from

ambiguity of a raw idea to some recommendations that can unlock some buy-in or commitment or execution or some planning, you know, that is really what we do with AI plus workflow.

Prateek Joshi (23:06.105)
Amazing. And working with all of your customers, what has been a surprising thing that you've discovered through this process?

Atif (23:18.43)
Yeah, I think some of the surprising things are, you know, it's hard to... Defining the problem is actually, you know, is difficult for people sometimes. And so when you help them with that, it's actually really powerful. It makes... Activates people. I guess the main surprise to me is that, you know, humans... You can take the brightest humans in the world. I remember one of the companies I mentioned...

one of our product managers, very bright woman, MIT, right? And I asked her to write a narrative, like I worked at Amazon previously, and it's a common thing there. And she was stuck. You know, she was looking at a blank word document. She could not, this idea that they were very passionate about, they could not actually write three good paragraphs about it, you know? There's a lot of cold start problems with humans where

Basically, if they just look at a blank canvas, they can't even get all their brilliance out there. Now, so what has surprised me with the ritual and software is that if you give some people something to react to, now all of a sudden, all their juices start flowing. So if you say, hey, this is a draft list of questions or here's a draft problem statement, they'll critique it, they'll improve it, they'll make it a lot better. And then if you do that at every step, you, you know, you break it down into the bits.

you can get a good, let's say, three-page document at the end, which is solid, like bulletproof. But the same person with a cold start makes no progress. And the same person with a little bit of help of breaking it down into bits, they actually produce something really cool.

Prateek Joshi (25:09.063)
That is actually a brilliant observation. And it happens a lot in writing of any form because people, they're way more likely to do, actually more work if there's like a first draft ready, they're ready to critique, make changes, edit it, add words. In fact, they'll add more words than they would have on a blank sheet, but they're just like worried or afraid of touching a blank sheet of paper. It's very interesting. And I think that's where AI can.

do a lot of good work is just get that first draft out. It could be bad, but it's a starting point and most people will just jump on it and run with it. So that's actually a very interesting observation. Cool, all right. So looking forward today, when we think about the concept of work, that we have a construct in mind. We look at work and we have a certain thing that pop into our heads. So if you look at it five years from now.

What does that workplace look like? Or rather, what do you wish that workplace looks like in five years?

Atif (26:17.57)
Well, my wish for people is, you know, I start with employee experience where like, how can we improve the experience of being a knowledge worker, let's say. And I think, you know, no doubt there needs to be challenge in that. So it's not going to be, you know, Oh, your day is easy. You're working on easy things. But I think we need to combine challenging work with psychological safety.

and flexibility. To me, that's like a really powerful kind of assortment of things or mix of things. So just breaking each one of those things down, I do see, you know, five years from now where the work is more, I think, intelligently defined. Like, we know where we're stuck and what's the problem that we need to get, you know, some group of people to look into.

And I don't know if it's five years away, but I think that will be more software driven through signaling. So it won't be like, oh, we have a meeting. It's a project status. Yeah. Yellow, green, red type thing self reported. Oh, blah, blah. And I think it'll be much more real time around. Hey, this initiative is stuck. So, you know, I think the work will play.

of people will be more defined on where we actually are with the initiative. And I think it'll be more structured what it is that we want people to go figure out. So I think that's one. I think the second thing is that we will create more psychological safety because we know that there'll be more transparency on where we are. It won't be like one person thinks...

We've decided the next person thinks we're at the beginning, the next person is ready to, you know, ship the prototype. You know, I mean, it's just, that's the thing that actually prevents people from actually having the space to do their best work. So I think knowing where we are and creating psychological safety will allow us to bring out the best in people. And then of course, within some framework of flexibility where, yeah, it's achievable this week or in the next two days.

Atif (28:42.85)
but I'm working on other things and I'm a personal life and so we allow you to kind of, you know, work within that constraint in your own way. So to me, that's what good looks like, whether it happens in five years, you know, we'll have to see. Yeah.

Prateek Joshi (28:58.712)
All right, with that, we're at the rapid fire round. I'll ask a series of questions and would love to hear your answers in 15 seconds or less. Are you ready?

Atif (29:09.234)
Sounds good.

Prateek Joshi (29:10.291)
Alright, question number one. What's your favorite book?

Atif (29:15.398)
My favorite book is probably Thinking Fast, Thinking Slow, Daniel Kahn.

Prateek Joshi (29:21.587)
Amazing, phenomenal writer. All right, next question. What has been an important but overlooked AI trend in the last 12 months?

Atif (29:34.158)
but overlooked, I think is basically the, you know, the impact on education and learning in high school, middle school.

Prateek Joshi (29:47.255)
What's the one thing about the future of work that most people don't get?

Atif (29:57.422)
I think it's not at all about command and control. I think it's gonna be about contribution. And we're looking at a very different world of measuring contribution. It won't be soft, it'll be pretty data driven.

Prateek Joshi (30:14.003)
All right, next question. What separates great AI products from the good ones?

Atif (30:21.322)
Well, in a short word, interface, but I think simplifying complexity. So something where all the complexity is hidden and it feels light. Creating lightness in these products is very important.

Prateek Joshi (30:37.487)
All right, that's a nice word to capture the essence. All right, next question. What have you changed your mind on recently?

Atif (30:47.35)
I've changed my mind on company building. I think, you know, bootstrapping, I think, is really kind of cool. It's not a weakness, it's a strength, actually. I think there's a lot of strength in bootstrapping.

Prateek Joshi (31:02.043)
Right, next question. What's your wildest AI prediction for the next 12 months?

Atif (31:10.094)
Oh, you're only giving me 12 months. I think 12 months would be probably some panic. Um, there's going to be some issues, some leaks, some proprietary company data. You know, I think that's going to happen always with the pioneers. They will try something in their enterprise, a high stack and something will go wrong, but that's, that's part of how the world works and maybe over indexing on cyber security as opposed to the upside, you know, so I think the risk first upside.

There probably is something that's going to happen there.

Prateek Joshi (31:41.939)
All right, final question. What's your number one advice to founders who are starting out today?

Atif (31:50.542)
I mean, it's cliche to talk about customer focus, but I would say is, you know, start with the end in mind in terms of like, what in the end is someone saying is the benefit and really just accept. Acceptance is very important as a founder. The benefit, maybe it's not happening and that's okay. Company building is not about two pivots, maybe about like 300 small pivots and you'll figure it out.

Prateek Joshi (32:20.723)
amazing. This has been a fantastic discussion. Loved your viewpoints on how work is evolving and the role that AI has to play in this. So thank you so much for coming on to the show and sharing your insights.

Atif (32:37.358)
It's been my pleasure, Pratik. I really appreciate you having me.

Prateek Joshi (32:41.507)
Alright, uh-