Zach Rattner, co-founder of global software development company Yembo, says that when it comes to artificial intelligence, organisations are eager to get started with AI but wary of the pitfalls. “Every group is figuring out what does it mean for them, and the rules of how it is going to end up are still being decided,” he said on this episode of the FM podcast, recorded at UK & Ireland ENGAGE.
Rattner explains beneficial ways to explore those tools to help professionals and leaders work better with them, what interpersonal skills professionals could sharpen as organisations become more driven by technology, and why it’s counterproductive to resist experimenting with technology.
What you’ll learn from this episode:
- Where finance professionals can expand their thinking around the possibilities of AI.
- How to avoid “losing your seat at the table” for AI conversations.
- Platforms that can help professionals upskill and enhance AI literacy.
- The essential interpersonal skills to develop for the digital era.
- Why fear of failure is a roadblock to AI proficiency.
Play the episode below or read the edited transcript:
— To comment on this episode or to suggest an idea for another episode, contact Steph Brown at Stephanie.Brown@aicpa-cima.com.
Transcript
Steph Brown: Welcome to the FM podcast. I’m Steph Brown from UK & Ireland ENGAGE. I’m joined by Zach Rattner, a guest speaker at this year’s conference. Zach is the co-founder and chief technology officer of global company Yembo, a software development company based in San Diego, California, in the US. We’ll be diving into what AI means for the finance and accounting profession, the potential benefits of using emerging technologies at work, and how organisations can unlock those benefits.
Welcome, Zach. Glad to have you on the podcast today.
Zach Rattner: Good afternoon, Steph. Thanks for having me.
Brown: Why should accountants be optimistic about AI?
Rattner: It doesn’t happen every day where a new technical advancement comes around that allows us to rethink the way work gets done, and I think AI opens up that opportunity where we don’t necessarily have to keep doing work the same way that we were doing before. It gives us an opportunity to take a look at what is working well that we want to do more of and what is not working that we maybe want to change.
I think the interesting part about AI in the finance sector and more generally across industries is that every group is figuring out what does it mean for them, and the rules of how it is going to end up are still being decided right now. I don’t think there’s any one entity out there that has everything all figured out, and that’s really encouraging and empowering to me because it means that we can have a seat at this table. When I look at the deployments that we’ve done to bring AI into different companies through my work at Yembo, I’ve seen a lot of examples on what needs to be done, and it changes the way that work happens.
I think folks generally think about AI maybe one-dimensionally, where they think about some aspects of their day and how nice it would be if they didn’t have to do that. While that’s certainly an aspect of it, I think the real powerful and important part is: What are we going to be doing, not what are we not going to be doing? If you focus on the areas that we’re going to be spending more time on, I think it will allow us to be more interdisciplinary, more strategic, more thoughtful, more prepared for large conversations, big deals. And, I think, overall, that makes the economy grow and makes everyone better off as a result.
Brown: What are some significant breakthroughs, say, the finance profession could achieve by implementing AI, in your perspective?
Rattner: There is a lot of work that goes into auditing, compliance, reporting, anomaly detection, and if you look at how these tasks are traditionally done, it’s constrained by labour. There are just so many hours in a day, and a human being can’t look at every single bank transaction over the last year or so when there are tens of thousands or millions of transactions. If you look at what AI has been doing, specifically with anomaly detection, there’s been some recent research out of MIT, and it’s been really remarkable what AI has been able to do.
But the interesting part is that AI doesn’t replace a person. You can think of it like having a personal computer and a spreadsheet versus having pen and paper, where it’s a powerful tool. I can’t imagine doing work without my computer. But, if you have just a laptop and no person, you’re still not getting that much work done.
Brown: In your session, you discussed how AI allows value to be tied to output but not effort. Can you expand on that?
Rattner: Sure. One of the incredible things that AI creates is it takes certain workflows which, traditionally, have been very labour-intensive and therefore very expensive, and it can completely commoditise it and make it be done in seconds, or for pennies. I believe that’s really interesting from an economic point of view because it doesn’t make the output any less valuable.
In the session, we were talking about narrating audiobooks using AI. This is something that had traditionally been done with humans in studios and it’s very labour-intensive, and takes a lot of time. It took us about five weeks to create a US English audiobook. But when you are using emerging technology like AI voice cloning and, with the narrator’s consent, of course, being able to clone the voice and create and synthesise new materials, I think that gets really interesting because a person can be way more productive than was ever possible before. If you think about what we showed in that little video clip where we had about ten different languages, if each one of those is going to be five weeks of work like the first book, you’re looking at one year worth of labour, and that was something that was done in about two weeks.
This is the promise that AI brings. That you can take these workflows that are historically constrained by labour, and now people who would have ordinarily not been able to do anything are able to do it. I think that’s going to create many more products, many more services. It’s going to change the nature of work, I think, for the better because it’ll allow people to better cater to demand, and that’s what, I think, is really exciting about AI.
Brown: There is some resistance to AI from companies and some workforces. You mentioned that AI will magnify us, it will not replace us.
Rattner: I think it’s valuable that AI has had its renaissance moment the last couple of years. What that means is that all these different industries — from removals where I come from, property insurance, finance, hospitality, banking — all these different industries are grappling with the same questions at the same time. The AI researchers created powerful technology, but they don’t know all of the implications. If you look at previous technical revolutions like the personal computer or mobile phones, it wasn’t just the vendor who created the laptop or the phone that benefited. They certainly did well for themselves, but, also, the hospitality companies that took advantage of it, the point-of-sale terminals in retail environments. There were so many different businesses that were built on the fact that everyone can have their own computer; everyone can carry it in their pocket.
With AI, I think we’re seeing the same thing again, where the rules of the nature of work in the 21st century now are up for debate and we can borrow concepts from other industries. We can take things that are working well, and we can make sure that the issues that we have in our workforce are being addressed. And, I think, that’s really empowering because otherwise you’re stuck with more of the same. I feel like it’s much easier to define how you want the nature of work to be when you have a forcing function that’s creating change, instead of just chopping up pieces of the same pie and trying to reshuffle what’s already there.
Brown: What would you say to someone who is either resistant or just not used to using those technological tools? What would you say to them to allow them to work better with it?
Rattner: Sure. When I hear hesitation, I do want to be sympathetic. I often agree with the concern that’s being voiced. Now, I disagree with the implications, but I think the concern around, “Would you embrace a tool that you believe is going to create obsolescence in your industry?” I could see why there’s resistance to that. I just don’t agree with the conclusion of that.
I think if you look, not just at me, but at other studies that have been done: the accounting giant PwC did a recent study and they saw that around 90% of CEOs who had adopted AI technologies agreed that it was going to change their day-to-day work post-pilot. And, I believe, those are the workflows that are going to allow us to do more and do better things is, the folks who are trying it, it’s not boiling the ocean, rewriting everything, and doing everything with AI. But if you can find one workflow, just to have one pilot that’s ongoing, 90% of the time there’s some outcome that’s beneficial that comes from that.
I would encourage you to, not necessarily do everything on day one, but find something to get started if there’s some area where you can try it. I think the worst thing to do is to sit by idly while others in the industry figure it out because then you had the opportunity for a seat at the table, but then you’re giving it up. I think it’s valuable if everybody who is going to be affected by this, which I believe is everyone, has a voice and has some experience under their belt and can speak credibly to the risks that they want to mitigate, the powers that they want to magnify. It’s a great equalizer because a lot of the people that are seeing the benefit of AI are not only in senior leadership positions, the day-to-day entry-level jobs have it, all across the board. Those are the folks that often have really good insights because they know what it takes to do that day-to-day work.
Brown: For people who are just entering the job market or perhaps re-entering — older workers are returning to different industries — where should they go to improve their AI proficiency?
Rattner: There are plenty of courses that are out there. I’ve also seen [that] the industry is moving very fast. Things that were state-of-the-art six months ago are sometimes obsolete a few months later. I would encourage learning by doing. You can only learn so much from reading a book or taking a course. If that helps, then certainly feel free to go and do it, but I would encourage not to stop there.
It’s relatively easy to find some workflow where you can actually use AI. If you’re doing project management or training, there are tools like Synthesia that allow you to generate AI avatars, where rather than going to a studio and recording a video of yourself talking, you can clone yourself and then generate training materials. There’s ChatGPT where you have data analysis, where you can have it analyse spreadsheets, build reports, all these different workflows. I would just encourage find something starting in an area you know really well and then testing it out and seeing what works.
I’ve been practising in this space for about a decade now, and that’s what I do. I don’t think I’ve figured everything out. In order to see if something’s going to work or not, I have to test, I have to experiment. I think that’s where the real innovations come, is where somebody who is not the inventor of the technology uses it and has an interesting innovative spin on it, and maybe applies it in a direction that the inventors didn’t realise. I think that’s what creates the creative economy and keeps things interesting and keeps the momentum going forward.
Brown: That’s great. Thank you, Zach. What skills, outside of technical skills, do you think will be essential for professionals to have and develop as industries and sectors become more driven by or reliant on emerging technologies?
Rattner: I believe that AI is going to make it harder to hide behind a desk. If you look at the kind of things that it does very well, crunching large amounts of data, building reports. If you look at a day in the life of a worker, I believe that means that the interpersonal aspects are going to become a lot more important. I know when I was starting my career over a decade ago, I would maybe only have a meeting with senior leadership in my department once a month, once every two months. A lot of the other time, I was doing work that was important but somewhat mechanical, routine.
I think what’s going to happen going forward is the ability to communicate cross-disciplinarily, the ability to motivate someone maybe with a different background, or maybe a fence-sitter, or maybe a naysayer. The ability to inspire someone. The ability to collaborate with people and to tackle a tough problem where a lot of the blocking and tackling is maybe done for you. I don’t see AI replacing any of those kinds of things. I think that’s going to create more opportunities for people to sharpen those skills, and I think that’s where society really needs people to be focusing on: Not just looking at what do we build, but how do we build it? Who do we include at the table? I would love to have in my day more time to focus on those things.
I feel like strategic work doesn’t really monopolise your schedule so much, it’s usually the operational work that boils over — maybe it’s a customer complaint, maybe it’s a project that’s behind schedule, [and] you have to do some extra hours. If AI can help keep a lid on that, so people are freed up to do more of that strategic work, I think that’s an economy that I would certainly welcome in my day and I believe a lot of other people [would] as well.
Brown: Thank you very much for this conversation today, Zach. Is there anything that we haven’t explored in this conversation that you think is important to mention?
Rattner: Yeah. I think the one piece I would add is that AI in its very nature — I don’t want to get too technical — but it’s all based on probabilities. That means that rather than being black and white, things are generally 80% sure that something has occurred. In our vision AI on Yembo, we’ll be like “72% sure we think this thing is a lamp,” and you have to figure out what the implications are for that.
What it means for us is that this work is, by nature, iterative, and that means that if you try something that doesn’t work out, don’t consider that a failure. It’s just one pass through the system, and you’ve learned something. You’ve learned maybe an example of over-reliance on AI, or maybe an example of what it doesn’t do great. All great AI companies that I know of are not one-hit wonders. They’re really good at listening to feedback, making corrections when something is wrong, and they’ll try to shrink the time that it takes to get through these iterations.
I would encourage you to go ahead and try something and, if it doesn’t work out, be analytical about it, note down what didn’t work, what was the assumption that was wrong. Then, go and have another pass at it. If you do that enough times, you’ll end up with something that often is super useful. But, you can’t have that shame of something not working out in the beginning because that’s just not how this industry works.
Brown: Thanks, Zach. That’s a great note to end on. Very good advice for anyone tuning in to this episode. Thank you so much again for coming on the FM podcast.
Rattner: Thanks so much for having me, Steph.
