How AI is Transforming Work: A Chat with Rebecca Hinds
The future of work is a hot topic, especially with AI shaking things up in ways we can't ignore. This episode dives deep into this question with Rebecca Hinds—not just any guest, she’s got a PhD in how AI transforms work and leads the Work Transformation Lab at Glean. We discuss how organizations are currently using AI and what that means for the structure of work in the future. Rebecca shares her insights from research showing that it’s not just about individuals adopting AI, but how teams and entire organizations can leverage it to improve workflows and collaboration. We explore the challenges companies face in scaling AI and how a more integrated approach can change the game for everyone involved.
Alongside her research, Rebecca also brings a practical lens to collaboration as the author of Your Best Meeting Ever, a blueprint for transforming meetings from time-wasters into powerful tools for teamwork and innovation.
Takeaways:
- The future of work is heavily influenced by AI, but there's still much uncertainty about its impact on jobs and organizations.
- Organizations struggle with AI adoption as many employees remain skeptical and are fearful of being replaced by technology.
- Successful deployment of AI requires understanding workflows and empowering the right influencers within organizations to drive change.
- The transformation of work is not just about individual tasks but rethinking entire work streams and team dynamics in a collaborative manner.
- AI has the potential to automate busy work, allowing employees to focus on more strategic and creative tasks, leading to a more engaged workforce.
- Creating a psychologically safe environment is crucial for employees to share their experiences and ideas about using AI without fear of negative repercussions.
Links referenced in this episode:
Companies mentioned in this episode:
- Feedforward
- Asana
- Glean
Transcript
One of the biggest questions we have at feedforward, both we and the feedforward team, but also every single one of our members, is what is work going to look like in the future with AI?
Speaker A:What is the future of work?
Speaker A:That has just become such a key question with sadly far too little research.
Speaker A:There's just not enough out there about what we see now in office spaces.
Speaker A:How is AI being deployed now?
Speaker A:What early indicators are there that tell us how it will be deployed in the future and how it will change the nature of work?
Speaker A:One of the people I go to the most to figure that out is Rebecca Hines.
Speaker A:She got her PhD studying how AI transforms work.
Speaker A:She ran the Work Innovation Lab at Asana, the software company, and now she runs the Work Transformation Lab at Glean, one of the leading AI powered B2B SaaS tools.
Speaker A:Good chance your company uses Glean.
Speaker A:This is a deep and wide ranging conversation about Rebecca's research.
Speaker A:I think you'll find it really fascinating.
Speaker A:I personally feel like the single most important question, not just about AI or about enterprise, but in a way for society, is how will AI impact workers, jobs, organizations?
Speaker A:And so that's why I'm excited to talk to Rebecca Hines.
Speaker A:Hi, Rebecca.
Speaker B:Hi, Adam.
Speaker B:I'm really looking forward to the conversation.
Speaker A:Yeah, you and I have talked before about this.
Speaker A:This is what you spend all your time thinking about.
Speaker A:So, Rebecca, you're now at asana.
Speaker A:I think most people know what ASANA is, but maybe explain quickly what ASANA is and then your job within asana.
Speaker B:Sure.
Speaker B:So Asana is a enterprise work management platform that helps teams work better together.
Speaker B:And.
Speaker B:And the Work Innovation Lab is our internal innovation lab within Asana.
Speaker B:So we conduct research on various future of work topics.
Speaker B: pretty much since day one in: Speaker B:And at that time I was coming off a PhD where I became really frustrated by the disconnect between great academic research and then what I was seeing on the ground in practice.
Speaker B:And the real impetus behind the lab was to develop a center where we could take great academic research and package it into a way that was digestible and impactful for executives.
Speaker B:And so the real differentiator is our close relationships with academic experts.
Speaker B:At any point, we have at least six to 10 studies going on with various academic partners across the world.
Speaker B:We involve them end to end in our research process in a way that I think has been impactful in staying abreast at all the changes that are happening, but still being able to have that lens of, okay, what does this mean for organizations right now, today?
Speaker A:Yeah, I was talking to someone the other day saying this might be my own prejudice, but for many areas of the world, if I hear that academics and practitioners are in different places, I'm going to assume that the academics are way behind and the practitioners are way ahead.
Speaker A:But my sense of AI is, and you know, there's obviously brilliant people in every, you know, brilliant practitioners, but my sense of AI is at least some of the conversation happening among academia is far more interesting, far more cutting edge, forward looking than a lot of the conversation I'm hearing among practitioners, which so, so I'm, I'm finding myself at least in this area and at least with specific.
Speaker A:I mean, it's not to say, you know, randomly grab any academic off the street and ask them what they think about AI.
Speaker A:They're just going to say it's a machine for undergrads to cheat and write papers.
Speaker A:They're not going to have a very sophisticated answer.
Speaker A:But the academics who are really looking at future of work stuff, it feels like that conversation is way ahead of the bulk of the corporate conversation.
Speaker A:Does that feel fair?
Speaker B:For me, it's less about who's ahead and who's behind and it's often just fundamentally different vantage points in terms of how these folks are looking at a problem.
Speaker B:So academics tend to be way more theoretical, whereas practitioners tend to be way more practical.
Speaker B:And I think even when we look at, we do a lot of research on collaboration and teamwork and a lot of the academic research that's decades old on teamwork can help us significantly as we transition to more distributed work, as we transition to new ways of work.
Speaker B:And there's just not that dialogue that tends to happen between the two communities in earnest where I think that's the opportunity here with AI.
Speaker B:And outside of AI, how can we make sure we understand the great theory?
Speaker B:And also I do believe the practical AI applications are quite sophisticated in the academic realm.
Speaker B:But how do we have that dialogue so that we're learning from each other and we're truly creating solutions together that will work and will be impactful in real organizations.
Speaker A:So let's get into it.
Speaker A:So obviously the goal of any company should be to just use AI, replace all your workers, have, you know, there should be like a CEO and a big AI machine.
Speaker A:Right?
Speaker A:That's the best way to handle AI, am I correct?
Speaker B:It's a really big important question and it's tough.
Speaker B:And the number one question I get from executives, and I'm sure You get the same question is how do we get employees to adopt the technology?
Speaker B:I think that's the struggle right now.
Speaker B:Organizations have seen pockets, they've seen pilots of AI work in their organizations, but they're not seeing wall to wall adoption and they're not seeing AI at scale within their organizations.
Speaker B:Only about a third of employees say their organizations have actually scaled AI in multiple different business units.
Speaker B:And what our research consistently points to is the fact that one size fits all models do not work when it comes to encouraging that adoption.
Speaker B:We've done a lot of research on who we call the skeptics within the organization.
Speaker B:These are people who are skeptical of AI.
Speaker B:They don't think it's that transformative.
Speaker B:They're fearful of being replaced, displaced.
Speaker B:And in particular, how you treat and approach that group of skeptics needs to differ vastly from how you treat what we sometimes call the enthusiasts within the, within the organization.
Speaker A:Who's using AI?
Speaker A:How are they using AI?
Speaker B:So in general, who is using AI?
Speaker B:It's more concentrated at the top levels of the organization.
Speaker B:Executives are significantly more likely to not only use the technology, but also be excited and enthusiastic about it.
Speaker B:They're also significantly more likely to say their organizations have policies and trainings and all the governments and support around it.
Speaker B:But senior executives are about 66% more likely to be early adopters of AI.
Speaker A:Oh, that's interesting.
Speaker A:I can see like senior, like senior vice president, executive vice president.
Speaker A:I meet a lot of people like that who are super engaged with AI.
Speaker A:The C suite I find is a mixed bag, like the very senior executives.
Speaker A:And sometimes I think in part it's just age that, you know, you generally see more early adoption of technology at younger ages.
Speaker A:So when you say executives, do you mean the very senior, like C suite executives or just broadly kind of knowledge workers, managers?
Speaker B:When we look at executives, typically it's the director plus level.
Speaker B:So it wouldn't be exclusively that that C suite.
Speaker B:I'm not sure if we've actually delineated the very C suite from the Director plus, but that, that would be interesting.
Speaker B:What were as well seeing are these massive network effects within organizations such that we can pretty reliably predict who's going to adopt Asana AI.
Speaker B:We have our AI solutions where you can build AI powered workflows and we know that if certain people adopt that technology, the people around them are significantly more likely to adopt the technology.
Speaker A:Yeah, you shared this with me earlier and I love this fact.
Speaker A:Walk me through that.
Speaker A:Like if I'm working in an office and one Is it my boss or just anyone in my team uses it, I'm more likely to use it.
Speaker B:So in general it's anyone on your team.
Speaker B:But we've looked at three main dimensions.
Speaker B:So the first is if you have a teammate who's adopted the technology, you're about 30% more likely to adopt AI.
Speaker B:If you have an executive, so either your boss or someone in your reporting chain adopts the technology, you're 39% more likely to adopt.
Speaker B:But the number one biggest driver of adoption is actually if a cross functional teammate or collaborator adopts the technology.
Speaker B:And I think there are a couple reasons for that.
Speaker B:One is it's evidence that the technology is scaled beyond just one team in an organization and it tends to be a more workable, impactful, scalable workflow that is being interacted around.
Speaker B:And so that cross functional piece is very important.
Speaker B:And I think when we talk about the state of AI right now, it's so laser focused on individual productivity where we and the academic community in particular see enormous potential to get out of that individual productivity focus.
Speaker B:And how do we use this technology to change how teams work, change how organizations are structured?
Speaker B:And I think that's the opportunity that we're not yet seeing in most organizations.
Speaker B:Organizations.
Speaker A:Okay, so, so the biggest indicator is like finance and HR or marketing and R D or whatever.
Speaker A:Like two more than one team is using it.
Speaker B:And yeah, and, and, and what we see is significantly different than other technologies is the relationship between HR and it.
Speaker B:We see HR and IT are significantly more likely to be interacting around this technology than, than ever before.
Speaker B:And I think that's a reflection of the fact that there needs to be a very human centric view as well as a very technical view where you're having the eye towards scaling and security, but you're also really understanding, okay, how does this look on the ground in terms of how people are adopting it?
Speaker B:From a human centric lens, this is always a question.
Speaker A:Correlation and causation is one scenario.
Speaker A:Could be there's some firms where there's like top down senior leadership that's implementing AI and that's what you're catching.
Speaker A:So that would make sense that if they're doing that, that we would see, oh, there's one teammate and then another teammate and another division.
Speaker A:I also can easily imagine a more bottoms up.
Speaker A:There just happens to be an enthusiast.
Speaker A:They figured out a few things, they told a friend, that friend told someone in another department.
Speaker A:Do you have a sense of causation?
Speaker A:Like since our members of people listening to this are they really want to get more than 30% of their workforce using AI.
Speaker A:So they want to know like, what are the levers I can pull to make that happen?
Speaker B:Yeah, so we do do a bunch of controls where we look at, okay, all else being equal in two organizations, how do we see AI spread?
Speaker B:There's always, especially in the survey data, it's nearly impossible to separate.
Speaker B:But we do have a good sense where all else being equal, if you have as much as we can control, if you have an individual within an organization adopt the technology versus not if they're in similar network positions, how does that then impact other people?
Speaker B:And what we do see is there, in addition to these sort of top down mandates, we see these people that we're calling AI influencers within organization have a disproportionate impact on whether other people around them adopt the technology.
Speaker B:So in particular this group of individuals that we call bridgers.
Speaker B:So these are essentially people who before I have, span different departments.
Speaker B:So they're HR folk who has been heavily embedded in the marketing organization, for example.
Speaker B:Those people are significantly more likely to influence others to adopt the technology.
Speaker B:And that's something we've known.
Speaker B:It's not unique to AI where we know that in general, if there top down change initiatives, those tend to reach only about 30% of employees.
Speaker B:If you go this influencer strategy and really leverage those influencers within your organization, it tends to be twice as effective in terms of inspiring change among employees within your organization.
Speaker B:And that's what we're seeing with AI too, by and large.
Speaker A:Okay, so if I'm like whatever, executive vice president in charge of AI deployment or whatever, I'm someone in a large Fortune 100 company with a goal of spreading it.
Speaker A:You know, spending a lot of time writing policy documents or company wide L and D efforts may or may not be beneficial.
Speaker A:But what I'm hearing is finding those people who are already using it, who already get it, and somehow empowering celebrating them.
Speaker A:Is that a good intuition?
Speaker B:I think it's incredibly important.
Speaker B:And what's fascinating is these influencers aren't necessarily the people who you think they would be there.
Speaker B:They tend not to be super technical.
Speaker B:They tend to be these bridgers.
Speaker B:They also tend to be the domain experts.
Speaker B:So maybe the HR person who has written policies for years or decades and has an intimate understanding of how the HR operation works, they can speak, they're not spewing technical jargon.
Speaker B:Right.
Speaker B:They can speak the language of their peers.
Speaker B:And so it's really important to, to understand who these people Are, chances are they're not who you think they are.
Speaker A:And they're not necessarily like a big boss.
Speaker B:No, they're often not at all.
Speaker B:They're not at all.
Speaker B:In the senior levels of the organization, it's much more these bridgers, these domain experts.
Speaker B:And then we also see later on in the adoption journey, people who are really good at process, process engineering tend to be, tend to hold these influencer positions as well.
Speaker A:Do you know Matt Bean at UC Santa Barbara?
Speaker B:Yes.
Speaker A:Yeah.
Speaker A:You know, he's one of our experts and I do love his work on kind of, you know, often he'll find in an organization people who don't even themselves know that they are major disruptor influencers.
Speaker A:They're just figuring out and not necessarily in AI.
Speaker A:I mean he's found these amazing like very low level factory workers who just figure out workflows that add millions of dollars of value to a company and don't even realize they did it.
Speaker A:They just saw a problem solved, it told someone about it.
Speaker B:Yes.
Speaker B:And often they don't have that platform to share and vocalize that they've developed these workflows.
Speaker B:And so it's not just finding your influencers.
Speaker B:We see an important next step as empowering them, giving them platforms, giving them early access to these tools and setting them up for success.
Speaker A:So let's talk about that.
Speaker A:How do we do that?
Speaker A:How do we get these people to know that, that we like what they're doing?
Speaker B:So the first step is to find them and identify them.
Speaker B:And we do it a lot of the time using network science.
Speaker B:But it's easy to do a similar exercise using surveys.
Speaker B:Right.
Speaker B:Figuring out where people are on the AI spectrum, where they're using AI, what are their use cases?
Speaker B:Are those use cases significantly individually focused versus are they more team oriented?
Speaker B:And then are they more cross functionally oriented?
Speaker A:And you want more team and cross functionally as opposed to just I'm using it to write memos.
Speaker B:Exactly.
Speaker B:You need the mix.
Speaker B:But in terms of if you're specifically looking at how do we get better, more holistic adoption of AI, you want the team.
Speaker B:We actually published a report a couple weeks ago led by my amazing colleague and we essentially looked at three different models of building AI workflows.
Speaker B:One is what we call sort of like track and field.
Speaker B:Right.
Speaker B:You're very individual in building these workflows.
Speaker B:Those workflows predictably don't tend to be adopted within the organization widely.
Speaker B:The second model is what we compare to more of football.
Speaker B:Right.
Speaker B:You have a quarterback who's really driving most of the activity.
Speaker B:And you have people involved in contributing, but it's not equal contribution.
Speaker B:And then the third model is, is what we call co creation, similar in many ways to basketball where you have people actively contributing.
Speaker B:They're roughly contributing in equal parts.
Speaker B:That model tends to be significantly more impactful in driving adoption and spreading those workflows across the organization.
Speaker B:And so it very much should be oriented towards those cross functional use cases.
Speaker B:And we also know that when people are using AI for cross functional use cases, they're significantly more likely to keep using the technology and be enthusiastic about the technology because it again tends to help streamline these really messy cross functional processes that are so often broken in organizations.
Speaker A:So to understand it's so, I mean there can also just be like contests.
Speaker A:And you know, Ethan Malik jokes that for the average Fortune 100 company, like if, if they just have sacks of $10,000 bills and just every day they hand it to whoever did the most AI that day or you know, everyone gets a gold bar if they, you know, you would save money in the long run.
Speaker A:But it's, it's so, it's finding those people who are just using it in creative ways.
Speaker A:I do want to note because this is another thing I learned from Ethan.
Speaker A:I think he might tell me I'm wrong.
Speaker A:It's great to celebrate successful like, oh, we figured out this tool that helps us with this annoying process.
Speaker A:But it's also great to celebrate failure that in an organization learning like on the frontier of a brand new technology, celebrating, wow, Rebecca really worked so hard on building this tool.
Speaker A:It's not buildable yet, but we love all the things she learned, you know, that kind of thing.
Speaker A:So walk me through like, can we like what that might look like in practice.
Speaker B:So the first bucket is really, what are those incentives?
Speaker B:What are those rewards to.
Speaker B:Once you've identified these people, how do you, how do you celebrate them?
Speaker B:And it's sharing the workflows, it's having slack channels where you're sharing what you've built.
Speaker B:You're celebrating those new AI workflows, you're having contests.
Speaker B:There's a big sort of movement it seems around prompt a thons.
Speaker B:You're building space, whether that's in the workday or in a virtual space that showcases these accomplishments to celebrate and convey the message that this is expected, that this is now a part of our muscle where we're experimenting around, around AI Equally important, if not more.
Speaker B:And Amy Edmondson in particular has inspired so much of this thinking.
Speaker A:Amy Edmondson At Harvard.
Speaker A:At Harvard.
Speaker A:Who has the psychological safety literature that she's famous for.
Speaker B:Yeah, exactly.
Speaker B:So that psychological safety is incredibly important and it's similar to what we know to be impactful for other aspects of work where, you know, post mortems where you're celebrating failures and people aren't afraid to not disclose what's truly worked and what's truly not worked.
Speaker B:Creating space for and it has to be well intentioned failures.
Speaker B:Right.
Speaker B:It has to be been thought.
Speaker A:Right.
Speaker A:It's not just any failure.
Speaker A:Right.
Speaker B:It's not carelessness.
Speaker B:And that's especially important with AI where it's not irresponsible failure, but you're sharing learning learnings because it helps other people not make those, those mistakes.
Speaker B:And it again creates this environment where it's okay to fail if you've been intentional about the process.
Speaker A:I don't know if this falls under psychological safety, but the main message I think that is out there, or certainly a plurality message is AI is a machine to get rid of people.
Speaker A:Is there good survey data on how people feel about their own future and AI?
Speaker B:Yes, I think we, this has been a question we've explored for, for at least a couple years now.
Speaker B:And what we see is right now about 33% of employees are afraid of being displaced by AI.
Speaker B:Obviously it's, it's different in different industries in different regions of the world.
Speaker B:But in general, a third of employees is, they think that 31% of their jobs can be replaced by AI.
Speaker B:So they're also recognizing that perhaps it's not the entire job, but portions of it.
Speaker B:That's the likely scenario here.
Speaker B:Right.
Speaker A:AI is only like that's.
Speaker A:I'm just noticing all the survey data is a third.
Speaker B:It seems like what's interesting and these are why the cuts are really important.
Speaker B:But what's interesting is the more people use the technology and the more they embed it in their day to day, the more that number increases in terms of just how much AI can replace.
Speaker B:And that makes sense, right?
Speaker B:The more you're familiar with the technology, the more you realize how powerful it is, the more you say, oh shoot, you know, it's not just 30% of my job that can be replaced.
Speaker B:It's actually probably closer to 60 or 70%.
Speaker B:So that's interesting.
Speaker B:But what I think is, is also fascinating is we see these fears that go beyond replacement and displacement.
Speaker B:We see more than a quarter of are concerned that other people will think they're lazy for using AI at work.
Speaker B:We see people fear they'll be perceived as a fraud for using AI at work.
Speaker B:And those are not that dissimilar in terms of the magnitude to some of these replacement and displacement fears.
Speaker B:And so it really is important for organizational leaders to recognize it's not just the job displacement and replacement.
Speaker B:It's also fear of the.
Speaker B:The perceptions of how people are using the technologies.
Speaker B:And we see that show up in fascinating ways in terms of people will interact with the same AI technology very differently if their boss is overseeing it or there's more humans in the loop versus not.
Speaker B:This is as much psychological as it is technical.
Speaker B:And I think that's the, that's the message that needs to be, or that's the thinking that needs to be.
Speaker A:Is it good if your boss is.
Speaker A:Because I could see that going either way.
Speaker B:That's why I think the psychological safety environment is crucial.
Speaker B:If you're in a psychologically safe environment, then probably it is good if your boss, for certain workflows, is in the loop and is understanding how you're using it and sharing feedback.
Speaker B:But if it's a micromanaging situation where you're afraid of making mistakes and you're afraid of AI making mistakes, then no, it's not going to be helpful or productive.
Speaker A:Roughly a third are afraid it'll replace them.
Speaker A:Roughly a third are using it.
Speaker B:Using it.
Speaker B:We see higher.
Speaker B:It depends on what timeframe you're.
Speaker B:You're looking at whether you're using it monthly, weekly.
Speaker B:But in general, we see more than half of knowledge workers in the US are using generative AI weekly right now.
Speaker B:So we have seen.
Speaker A:Oh, I'm okay.
Speaker A:Because earlier you talked about a third of the employees are actively.
Speaker A:But that includes non knowledge workers.
Speaker B:Is that the third was how many employees say their organizations have actually scaled AI.
Speaker A:I see, I see.
Speaker A:Okay, that's helpful.
Speaker A:Sorry.
Speaker B:So that's, that's where we see, you know, as last year it was about adoption.
Speaker B:This year we see every organization, in terms of every industry struggle with scaling the technology.
Speaker B:That's where I'm seeing, you know, lots of questions, lots of concerns.
Speaker B:You've rolled out this technology, you've invested in it, only one team is using it.
Speaker B:You can't move past the pilot.
Speaker B:That's where we're seeing sort of this pilot purgatory situation where most organizations have not seen the impact they were hoping for in terms of true organization transformation.
Speaker A:That resonates with my anecdotal experience as well.
Speaker A:But that means.
Speaker A:So if, if a third say their company is deploying it and more than half are using it that means some significant number are using it independent of whatever their company is is saying.
Speaker B:We we see that and again the scaling is multiple different organizational groups.
Speaker B:And so you could very much have situations which we see is you roll it out to the IT function and it stays in the IT function and it doesn't go beyond that.
Speaker B:That would be an example where employ that the technology is not scaled across multiple different functions within the organization.
Speaker B:We see a lot of this use of AI without disclosing the use of AI.
Speaker B:And I think that tracks with what a lot of other surveys and studies have found as well.
Speaker A:Right, right.
Speaker A:So that, I mean and that is something we often say is as a company you don't get to decide if your employees are using it, that probably many are using it.
Speaker A:You do get to decide decide how they think about using it, how they think you think they should think about using it.
Speaker A:But that, that's.
Speaker A:But you don't get to decide if they're using it.
Speaker A:You also get to decide if they tell you about it.
Speaker A:Right.
Speaker A:That gets to the psychological safety.
Speaker A:If I feel pretty confident that telling my boss I'm using it is going to get me fired or you know, we have some long policy document that scares the heck out of me because it's filled with all the things that, that I could mess up, then the last thing I'm going to do is tell anyone I'm using it.
Speaker A:Right.
Speaker A:Okay.
Speaker A:So I do want to pivot.
Speaker A:I mean this is all fabulous and I would love to put as much of this survey information on the web because this is really helpful.
Speaker A:Obviously any individual company's mileage may vary.
Speaker A:You know, it might be a little different depending on what's going on with them.
Speaker A:But, but it's really helpful to have this framing.
Speaker A:Let's talk about kind of the workplace of the future.
Speaker A:Obviously you don't know.
Speaker A:I don't know.
Speaker A:We don't know.
Speaker A:Right.
Speaker A:Nobody knows.
Speaker A:But we talked about the kind of individual contributor lens on AI that it's going to write my memos, it's going to go through my email, it's going to whatever and eventually maybe it'll do so much I don't need a job anymore.
Speaker A:But you were talking about rethinking the organization overall.
Speaker A:And this is something Daniel, rock on.
Speaker A:Our one of our experts who you know at Warren and also talks really thoughtfully about is is the biggest bang for your buck is not looking at individuals and can I get rid of 30% of their job?
Speaker A:50%, 80%, 100%.
Speaker A:But rather looking at entire work streams.
Speaker A:Help me understand how you think about that.
Speaker B:Sure.
Speaker B:So I, a lot of my thinking is grounded in the academic research that has long shown you can have this, this tragedy of the commons, it's sometimes called situation where you can easily imagine a world where individuals are using AI to strictly enhance their own personal productivity.
Speaker B:They're delegating tasks to their colleagues without thinking about the bigger picture.
Speaker B:They're overwhelming them.
Speaker B:And we see it where you can have situations where the system becomes more inefficient or slower, even if an individual within that system is, is adopting and using the technology to enhance their individual productivity.
Speaker B:So that's sort of the mental model that I've come into this and we've seen it confirmed in the telemetry data as well as the survey data.
Speaker B:What I'm really excited about, and we're seeing pieces of it is this opportunity to rethink how teams work and how organizations work.
Speaker B:This is not, you know, a next year transformation but, but I do think we're seeing enough evidence and the technology is pointing in this direction where you can imagine a world where we become much more effective in terms of how we structure teams within our organization right now.
Speaker B:And this is a lot of my academic research during the PhD focused on how AI was changing or intention with organizational charts.
Speaker B:And if you think about an org chart, the main purpose of that has been to reduce complexity.
Speaker B:Right.
Speaker B:It tells you who makes decisions, it tells you who you need to escalate work to.
Speaker A:If we have, it's about information flow.
Speaker B:Like how does like information flow, you know, who to, you know, tap on the shoulders.
Speaker B:And if AI is starting to take on some of that work, we're starting to delegate the decisions to AI.
Speaker B:We have more access to information.
Speaker B:We're able to search and not only search, but also discover that information easier, more easily.
Speaker B:It calls into question, okay, does the org chart need to look this way in the future?
Speaker B:And my advisor in the PhD was Melissa Valentine who's done a lot of research on flash teams and org design and she has a new book coming out on flash teams, which are essentially these hyper dynamic teams that can be formed incredibly quickly to solve ad hoc problems problems.
Speaker A:And I'm looking this up, I'm going to order it right away.
Speaker A:Flash teams leading the future of AI enhanced on demand work.
Speaker B:Yeah, it's, it's coming out next year I believe with Michael Bernstein, who's been a great, great.
Speaker A:You must have told me about it before because I, it says I pre ordered it all right.
Speaker A:Okay.
Speaker B:Yeah, it's, it'll.
Speaker B:It'll be a.
Speaker B:A great book.
Speaker B:And you can, you can see how powerful that is, right?
Speaker B:Why, when we start a project, do we only look to our individual team to solve that project?
Speaker B:Why don't we look across our entire organization or even beyond that to customers, partners, to figure out, okay, given priorities, given the problem at hand, given personalities, given prior work experiences, what is the exact right composition of people who can solve this specific problem right now?
Speaker B:And that rarely is confined to the team that the org chart has structured.
Speaker B:And so I think that's incredibly exciting to have a more holistic sense of what's going on in our organization.
Speaker B:We have very little visibility in general to how work happens in our organizations.
Speaker B:Who's collaborating, who should be collaborating?
Speaker B:What are the priorities?
Speaker B:I can help with a lot of that if we're designing it at this team level, organizational level approach.
Speaker B:And then I do think we're going to see changes to the org chart.
Speaker B:I don't know.
Speaker B:No one knows what that will look like.
Speaker B:But there's quite a bit of evidence pointing to the fact that flatter org charts are likely in specific industries.
Speaker B:And I think that's my working prediction that probably you're going to have flatter, more specialized folks within organizations who still have the breadth.
Speaker B:But I think we're going to see flatter org charts moving forward.
Speaker B:Forward.
Speaker A:I'd love to make this tangible.
Speaker A:It just helps me to think.
Speaker A:I once wrote a column when I was at the New York Times Magazine about the Hollywood model when I worked on a movie, the Big Short.
Speaker A:And it just kind of amazed me how one morning there was this massive team of 250 people or whatever it was, and the carpenters were doing what carpenters do and the electricians doing what electricians do, and the hair and makeup people and the set designers and the actors and the directors and.
Speaker A:And they just, you know, some of them had worked together on projects, but this particular team had never been brought together.
Speaker A:And all of a sudden, on a Tuesday or whatever it was, at 6am they're all there.
Speaker A:Everyone's doing what they're supposed to be doing.
Speaker A:Six weeks later, they disbanded.
Speaker A:And then some other team, you know, some of them, they all went to some other team.
Speaker A:All right, so here's the model I have in my head.
Speaker A:And I want to see if this is a good or, you know, kind of toy example.
Speaker A:I think of like, all right, so someone has a tech support issue.
Speaker A:Let's say we make some kind of widget and someone has a tech support issue.
Speaker A:My widget doesn't work.
Speaker A:They call up tech support.
Speaker A:Tech support solves their problem.
Speaker A:Tech support reports up through it.
Speaker A:It has its own agenda, its own, you know, requirements.
Speaker A:Then a few days later, a few months later, a few weeks later, some salesperson calls that same client.
Speaker A:They probably don't know much about that tech support call call.
Speaker A:At the same time, there's an R and D team working on whatever R and D for the next product.
Speaker A:You know, there's a marketing team, et cetera.
Speaker A:And what I was imagining with AI is if, if there's an AI model, if AI is listening to that tech support call and you know, and you're using multiple agents and one agent, maybe it's an AI tech support, maybe it's a human tech support, but there's one agent that's listening for what is the underlying issue here?
Speaker A:You know, is this a sales opportunity?
Speaker A:Is this, is this a customer who has the WR product or doesn't have enough of the product or isn't properly trained?
Speaker A:And then another AI is listening to that tech support call and also every other tech support call for how can we redesign our products to make them better?
Speaker A:And then another agent is listening for keywords and thinking, oh, this is how our customers talk about our product.
Speaker A:We should market with those words instead of the words we use.
Speaker A:And then you can sort of imagine we're amassing huge amounts of this information.
Speaker A:But then AI is really good at going through huge amounts of information.
Speaker A:So people in different departments are able to approach an issue in a very different way.
Speaker A:They can go into an R and D process with a whole host of information they wouldn't have had otherwise.
Speaker A:And I start, you know, my, my brain kind of stops at that point, but I, it starts to, you start to think, like, wait, why do we have like an IT department that's a silo?
Speaker A:Why do we have a marketing department?
Speaker A:Why do we broadly kind of have profit revenue departments and kind of cost centers and, and, you know, you know, there's some great work by Zainab Tan on how you can actually structure retail operations so that your workers are not just a cost, they're actually a source of profit and revenue and it.
Speaker A:And it transforms.
Speaker A:So am I on the right track of thinking this?
Speaker A:And then I, I really do have struggle to go to the next step.
Speaker A:Like how?
Speaker A:Like, if someone then said to me, all right, Adam, you're now in charge of X Y by Z Corporation, redesign it.
Speaker A:I don't actually know how I would redesign it.
Speaker B:Yes.
Speaker B:And I don't think anyone knows knows fully, but you're absolutely on the right track.
Speaker B:And I.
Speaker B:So I have a book on meetings coming out next year and one of the massive transformations that we're going to see in the meeting landscape is this shift from what I call search in terms of transcript search to discovery where there are already lots of platforms building infrastructure so that you can imagine a world where you have a video conferencing platform across your organization.
Speaker B:AI is extracting those insights from the transcript and serving them up to the relevant folks within the organization.
Speaker B:So if you have a sales call you're not even on and they mention a product that you're selling or your customer talks to a different person in the organization, proactively surfacing those insights sites to then take action, we're going to see a huge shift.
Speaker B:I've seen large, large organizations start to move towards memos as opposed to slide decks.
Speaker B:And we've seen this.
Speaker B:Amazon was the leading thinker in this.
Speaker B:But why we're seeing organizations move that way now is because they recognize that AI and LLMs are much better at processing memos than they are slide decks.
Speaker B:And so they're taking that proactive step and saying, okay, we know this new world of discovery and more intelligent search is going to be key.
Speaker B:We're going to set ourselves up for more success in that world.
Speaker B:And I think that's going to be the biggest short term change we see is meetings are going to be fundamentally redesigned and also the insights that come out of them are going to be much more useful than we've seen in the past.
Speaker A:Can we just have no meetings?
Speaker A:Is that something we could do or just fewer meetings?
Speaker A:Or meetings are more like to build camaraderie and have fun and.
Speaker B:Exactly, exactly.
Speaker B:And I sort of lay out the yardstick in terms of how we should be evaluating meetings in the book.
Speaker B:But we are in too many meetings and they're used as duct tape.
Speaker B:Right.
Speaker B:Every time we have a problem, we schedule a meeting and it's the biggest cost in our organization.
Speaker B:It's the biggest, most, least optimized product or collaboration way of working that we have in our organization.
Speaker B:So massive opportunity for redesign and restructure.
Speaker B:Yesterday I was on a call with three different bots, three different vendor bots, along with humans.
Speaker B:And it's also going to be overwhelming if we just are sending bots to our meetings.
Speaker B:That's also a big, big problem.
Speaker A:Right.
Speaker A:So.
Speaker A:Right.
Speaker A:Because that's the other side.
Speaker A:We could see meeting proliferation and you know, and like the New Yorker cartoon version of I just give a one sentence prompt so that my avatar can have an hour long meeting with a bunch of other avatars so that another AI agent can condense it all into a one sentence response.
Speaker A:And yeah, I'm excited for this meeting.
Speaker A:It is amazing.
Speaker A:I also love, I think everyone's been to meetings about how we have too many meetings and then somehow that generates more meetings.
Speaker A:Yeah, I remember talking to someone who said, you know, at our company, like, if I wanted to launch a project that costs a hundred thousand dollars, I would, there'd be a whole bunch of approvals, but I can just send a bunch of invites and have a meeting that is way over a hundred thousand.
Speaker A:You know, I can just start a day or a weekly, you know, Tuesday at 6, 5, you know, marketing check in meeting.
Speaker A:And if you look at all the people involved and the time involved, that's like a $500,000 project and nobody even thinks about it.
Speaker A:So let's get back to the kind of core question that I hinted at in the beginning, which is what this means for the future of jobs.
Speaker A:Like where my mind goes like it.
Speaker A:I, I know how to say the words, don't think about the individual contributor, think about the system.
Speaker A:But then I don't know how to think past that.
Speaker A:And I feel like the best I can do at this point is to say, when you think about the system like a whole work stream, how do we go from identifying customer needs all the way to launching a product and selling it and servicing it and, and if we think about that in an AI enabled way, I feel very comfortable saying it's going to look really different from now.
Speaker A:I have some insight into the kind of people I think will thrive and the kind of people who won't thrive.
Speaker A:But where I get stuck is like, do you end up with more workers, fewer workers, higher paid workers, lower paid workers, you know what, what happens?
Speaker B:It's tough and no one knows.
Speaker B:You can find evidence in every direction.
Speaker B:Really.
Speaker B:What I do know is work is going to change for sure.
Speaker B:And I think the first area of work that's going to change is what we call busy work within organizations.
Speaker B:These are the meetings.
Speaker B:Our research consistently points to about 53% of workers.
Speaker B:Time is spent on busy work.
Speaker B:And so figuring out where that busy work is living, whether it's living in scheduling meetings or sending status updates and automating those processes.
Speaker B:One of my favorite for first applications of AI was automating status updates that, you know, in some of our Customers, they spend Friday afternoons, every Friday afternoons writing status updates on projects.
Speaker B:Right?
Speaker B:You can, you can automate that so easily with the right, with the right AI technology.
Speaker B:And then it's about mapping out what are those workflows within your organization and which parts of them can be and should be delegated to AI.
Speaker B:I think one of the big reasons we're seeing, seeing these companies struggle past the pilot is they're slapping AI onto broken processes and that doesn't work.
Speaker B:AI is not going to magically make your, you know, sales, marketing handoff better if the underlying structure isn't broken.
Speaker B:And I think that's, that's fundamentally, it's so much of the, at the core of so much of the AI failures right now is organizations need to understand the process and they need to be thoughtful at, okay, which parts of the process should humans still be accountable for and oversee versus which ones can be delegated to AI?
Speaker B:If you're in a customer service workflow, humans should still take on the super sensitive issues or issues where there's high emotional intensity, but there are certainly other issues where AI can be a huge part of that process, even in an agentic capacity.
Speaker B:And so I think this depends on understanding your business, understanding the workflows, and then being strategic about, okay, which parts of the workflow do I want to inject AI?
Speaker B:And then in which capacity is it more agentic or is it more, you know, humans are definitely in the loop at every approval stop.
Speaker B:How do we make sure this workflow is going to be most effective?
Speaker A:I think understanding your business is a pretty big request.
Speaker A:I don't understand how you get away with such nonsense.
Speaker A:But yeah, I mean, I remember working, working with a company where they had all sorts of problems and they thought it was a marketing problem.
Speaker A:And then it turned out like sales was selling a product that the engineers were not interested in or capable of delivering.
Speaker A:And so every contract started with unselling what was sold.
Speaker A:And these were multimillion dollar big engagements.
Speaker A:And AI is not going to fix that.
Speaker A:I mean, if anything, AI could just hasten it.
Speaker A:But it is really challenging because, because it's hard to know today what the contours of this is.
Speaker A:There's a ton of stuff that can be done now, but if we start imagining what will be possible two years from now, five years from now, eight years from now, I mean, we're really out over our skis.
Speaker A:We have no idea.
Speaker A:I could see how it would be scary to say, let's just get rid of departments or let's Come up with some brand new model of work collaboration.
Speaker A:If you know just this every week, you know, there's some new models, some new tools, some new approach, some new thought that is transformative.
Speaker A:So far it's all been in the wow, AI is moving faster than we expected direction.
Speaker A:But at some point I assume, and I don't know if that's six months from now or 60 years from now, it's going to go the other way.
Speaker B:Yeah.
Speaker B:And that's why I think leaning into these influencers that we talked about at the beginning is so important.
Speaker B:Because if you're thinking about less in terms of the structure of the org chart, but if you're thinking about, okay, I need to know exactly how my business works, you need to know how those people at the front lines at the lowest levels of the organization are working because they're going to be best suited in so many cases to identify, okay, does this make sense?
Speaker B:Is it a workable new workflow?
Speaker B:Where can it break down?
Speaker B:And leaning into their expertise, which often doesn't happen when we're rolling out these transformative technologies.
Speaker A:Luckily that's something AI is really good at.
Speaker A:I mean that's another major use case is you.
Speaker A:I remember when I became CEO of a small company and my friend who was a CEO of a slightly bigger company was like, one thing that's crazy making is on your team.
Speaker A:There's someone who knows whatever problem you're trying to solve better than you do.
Speaker A:But it's hard to know who that person is and, and to assess their ideas.
Speaker A:But that idea of using AI to like, you know, there's, there's people in shipping and logistics who have lots of ideas about how to cut costs and speed up shipping and logistics.
Speaker A:There's people in like customer service who are like aware of all sorts of persistent issues.
Speaker A:And it's just a deep challenge in a traditional organization to find that information, bring it to light, assess it, make change based on it, there's all sorts of power dynamics and network issue.
Speaker A:Right.
Speaker A:I mean that AI is not a magic pill, but you can start imagining all sorts of ways of like teasing out the inherent knowledge in your organization.
Speaker B:It's going to be a big unlock because right now, and we're seeing it even on an escalated stage ever since the pandemic where we have very close knit circles within organizations where we don't tend to go outside of them and there's, you're significantly more likely to talk to the person who you can physically see in the office than anyone else.
Speaker B:And AI can help democratize some of this knowledge.
Speaker B:Knowledge and surface, you know, who knows what within the organization, which is one of the biggest challenges.
Speaker A:Yeah, when I was at npr, I started an exchange program where literally you would just go spend a week in another department because, like, you'd work with these people, you'd see them in the cafeteria, but you just didn't know what they were doing day to day.
Speaker A:And it really was transformative.
Speaker A:It was crazy to see like, oh, the people who put All Things Considered together are not just sitting around trying to figure out how to mess me up.
Speaker A:They actually have other things they're doing.
Speaker B:Yeah, there's an old, old study from MIT that looked at engineers and found that if engineers are seated in an office six feet away from one another, they're four times more likely to collaborate than if they're seated 60ft away from one another.
Speaker B:And as soon as that distance increases past about 100ft, you're less likely to communic and you're less likely to communicate, not just through face to face conversation, but also through email and other virtual channels.
Speaker B:And we long rely on this visual proximity bias that I think AI can help in many ways.
Speaker A:It's interesting to wonder how in person, I could imagine that for some industries, some businesses in person becomes more important, that AI is taking care of all the junk when you're in your office and, and collaboration.
Speaker A:I mean, it reminds me, you know, when the Internet, I'm older than you and I remember, you know, when we started getting email and da, da, da, it was common to say, oh, cities won't matter anymore, like New York, San Francisco, et cetera.
Speaker A:Just people are gonna like, live out in a farm somewhere and do everything remotely.
Speaker A:And the opposite has happened.
Speaker A:Real estate is far more valuable because these agglomeration effects, having proximity is so important.
Speaker A:And I can imagine, I would imagine at least for some industries that, and this is exciting.
Speaker A:I don't want to say this is going to be true for everyone, but like most of your day could be those creative, exciting, fun, engaging with other co workers collectively solving problems.
Speaker A:And AI is doing all the other stuff and facilitating those.
Speaker A:Who knows?
Speaker A:That's a nice idea.
Speaker B:And maybe the last point is the area we didn't talk about that is fascinating is how AI can help nudge us in the direction.
Speaker B:So there's fascinating research that came out of Stanford and Berkeley a couple years ago where they essentially measured what they called the discursive diversity of conversations to be able to predict whether people should be more in a collaboration mode or be in more coordination mode.
Speaker B:You can easily imagine AI can help.
Speaker B:Okay.
Speaker B:If our connection to our colleague has waned, we haven't seen them in person in three, three months.
Speaker B:It's time for an in person checkpoint.
Speaker B:You can also imagine AI nudging you in that direction and potentially making work even more human.
Speaker A:Yeah, I'm rooting for that.
Speaker A:We'll see.
Speaker A:All right.
Speaker A:All right, Rebecca Hines, thank you so much.
Speaker A:Asana Future of work.
Speaker A:What.
Speaker A:What is.
Speaker A:What is.
Speaker A:What is your job called again?
Speaker B:I lead our Work Innovation Lab.
Speaker A:Work Innovation Lab.
Speaker A:So we'll put some stuff in the Discord, but you're very reachable and readable.
Speaker A:Thank you so much.
Speaker A:That was awesome.
Speaker B:Thanks so much, Adam.
Speaker A:I have no doubt Rebecca Hines would love to talk to you if you have any interest in her work and maybe, just maybe would be open to offering some kind of partnership around studying how AI deployment works at your company.
Speaker A:No pressure, that's up to you.
Speaker A:But she's a really amazing resource, and when I told her about the feedforward community, you can see that research mind salivates.
Speaker A:You are the people who are actually creating the future that she is trying to figure out and study.
Speaker A:I am Adam Davidson, and this was the FeedForward podcast.
Speaker A:I look forward to seeing you on the Discord and in our member meetings.
Speaker A:Please reach out if you have any questions or thoughts about this podcast or really about anything else.
Speaker A:Thanks so much.
