Aug 25, 2024
Welcome to part two of the two-part interview with doFlo Co-Founder and CEO Will Butler. In part one, we discussed Will's motivations for founding doFlo, as well as some of the key technologies involved. Here, we discuss the potential future of organizations in a highly automated world. Click here for Part One.
On Human-Centered Design
I want people to be able to use the internet to accomplish the feats of organization it’s capable of.
Lloyd: Ultimately, people want to be in control, and they want a human being to be responsible, even when something is automated.
Will: That’s another key issue with automation. Automate too completely, and nobody is in control. Nobody is responsible. Exactly. We need to create and encourage responsible uses for this technology. It is not about eliminating people from a chain of responsibility. Instead, automation can keep people responsible for what they’re doing. Automation can make the world safer and fairer and make information and technology more accessible to people who don’t have a lot of power. That’s important work. That’s why we’re doing this.
We consider ourselves an AI and Automation company, so that we will build with AI internally as part of doFlo. We always have to be transparent and honest about its limitations and do our best to make sure it isn’t used in inappropriate ways. That’s a key area of responsibility for us. Privacy and data security are big issues with AI. Responsibility is another.
Lloyd: Fiving regular people access to the automation technology that large companies use could open up opportunities for small groups or individuals to impact their world.
Will: Right. That’s exactly what we want. Are you passionate about, say, global climate change? Well, I want you to use doFlo to 10x your ability to monitor the sources of global emissions. We can then automate your responses to that data and then keep building on that. If everyone could design complex automated systems, we would suddenly have an information economy that is truly “many to many.” Right now, the internet is starting to more and more resemble cable in the 1990s. Remember cable in the 90s?
Lloyd: Yeah, 90 channels, nothing’s on.
Will: Right? That’s the reality today. Big companies want to swamp us with information and drown us all out with propaganda. I want people to be able to use the internet to accomplish the feats of organization it’s capable of. Not have it be used as a kind of pipeline for corporate propaganda.
Lloyd: Wow, that’s a very political mission.
Will: These are political times. As I published recently in our post on “Our Principles.” I believe that if you are not explicitly serving the cause of freedom of information and of speech, and of people’s rights to privacy and difference, then you are complicit in what happens to them when those things are neglected. We are building a company that puts this responsibility first, not as an afterthought.
This is a mission to free people from the tedium of repetitive work. It's also a mission to provide the powerful tools already available to the rich and big businesses. The people who impact all of us at scale with everything they do. We want every individual to be able to have an impact at scale. If they have a good idea or a great creative vision. That’s our goal.
Lloyd: It’s a big, big idea.
Will: (Laughs). I know, it all sounds like too much. Simply: I want a world where my technology works for me, not where I work for it. I want to be an agent of change in the world, not a victim of circumstances. I want all of us to be potential agents of change. Every technology revolution has made this promise, and every one of them has also failed, to some degree, to deliver. I don’t anticipate this will be easy, but it is worth doing. It is a reason to get up in the morning.
A Tech Revolution?
They don’t see themselves doing that yet. But when we put it in front of them, they'll "get it."
Lloyd: Do you think this is a new revolutionary moment?
Will: I think so. I think ChatGPT sparked a revolution that is just getting started. And we need to build human-centered automation solutions so that this revolution delivers some benefit to ordinary people. That’s why we’re building this as primarily a self-service platform, and we’re offering it for free in our private beta. We’ll have to figure out how to pay for it eventually. But I want regular people to engage with this technology and see what they can come up with. I think the more people we have doing that, the better it’s gonna be for them and for us.
Lloyd: What is it that makes this moment revolutionary?
Will: We’ve been digitizing the world for decades. But up until very recently, the world wasn’t really living on the internet. We still had phone calls, we still had paper books, we still had newspapers. We still had meetings and offices, and all that other stuff. And then suddenly, in 2020, all that stuff just went online. The interview we’re doing right now is digitized and transcribed. You could ask an AI to read it and summarize it; refer back to it later. That’s just… unbelievably powerful. As we do more online, we create more data daily than we used to in a year.
The need for automation to help us deal with this new reality is here now. It’s all around us in our daily activities. So many things we’re still doing manually don’t make much sense anymore. So much of our time is spent on things that don’t benefit society. Can’t we get back more of our time, finally? Now that the world has gone digital, where is the relief? Where is the “peace dividend” of a digital world?
Lloyd: It’s funny, we often think that the digital revolution was in the 80s and 90s, but you’re arguing it’s today.
Will: Yes. It’s like when Google came up with Search. If you had been using Yahoo or one of their competitors, it was just a kind of library of linked pages. People thought that was the natural way of finding information. They didn’t believe that you could type a question and get a really useful answer. People marveled at that. That was a revolutionary moment.
I think we’re there today with automation. I think right now, people are sort of happy with how SaaS automation works because they don’t believe that you could just “draw it, and it works.” They don’t see themselves doing that yet. But when we put it in front of them, they'll "get it."
Lloyd: So you’re hoping that will be your “aha” moment.
Will: Exactly. And I want people to think: “Oh my god, this is fantastic. Now I don’t have to do XYZ all the time! I have time to play with my kids!”
The Anti-Work Movement
Work-life balance is a treacherous phrase.
Lloyd: So you want people to get more time off?
Will: Yeah. I want people to be more productive, but also to do more of what’s good for them and society. That productivity should benefit them directly. Take care of their kids. Spend time with their spouses. That’s what people ought to be doing in our digital age. So much of what we used to have to do is unnecessary anymore. So let’s automate that, but let’s give people back their time. They can spend it on whatever they want. Maybe they want to be more active in a hobby, research, or creative enterprise. I don’t know. Let them figure it out. That’s my take.
Lloyd: Part of your statement on the principles of doFlo is to shorten the work week and increase work-life balance.
Will: Work-life balance is a treacherous phrase. It implies that work and life are somehow a continuum, or like two equal parts of someone’s life. To me, life is everything. If you love working on what you do, then do it as much as you want. But let’s be real: I’m going to be hiring people to do lots of things, and I don’t expect that every single one of them will be in love with the work. Ideally, they won’t have to. We can create a situation where they are spending as much of their time as possible on stuff they enjoy.
That’s why I’m trying to conceptualize a workplace where “good workers” are not necessarily workers who give us more time. They are workers who bring great value to what they do and spend as much time doing their best work as possible, but no more. I also want to work with people who are dedicated to their communities and to their families. I don’t want doFlo to be a place where people miss a spiritual, familial connection and have to find that at work. That’s bleak, to me. Work should not be your life if you don’t want it to be. It should not be, as a default, the focus of your life.
This is really important to me. I’m not just saying it. This is one of the things I love about Prague. Thomas and I spent many years working there together. They don’t depend on their careers to fill them spiritually or emotionally. They have their own lives, and that’s treated as separate most of the time. One of the reasons we decided to move back to Prague to work on doFlo was this. In London, you are where you went to school. In America, you are what you do. But in Czechia, people are defined by their friendships and family, and I appreciate that.
Lloyd: Does that philosophy also extend to your customers at doFlo? How can you influence other organizations to change their work cultures?
Will: What can I do? I can offer competitive pay for good and flexible hours. That at least gives people the option of a company like doFlo. It incentivizes other companies to try to do the same.
We can also work towards putting regular workers in control of the automations they are designing. That’s why I’ve said I want our product to be something regular people can use. It's so that they can design the automations they work with all the time. Not so that somebody above them can automate them out of existence. That’s important to me. I want automation to be a ground-level up innovation. Not something that is used to squeeze people even harder.
doFlo: Ground Floor Automation
You can get people to give you 25% more productivity, but you can’t make people 25% more creative... This is why we want to make everyone an automation designer.
Lloyd: Ok, let’s talk about that in terms of the technology. You mentioned Event Streaming Technology as one of the innovations you’re working with so that each individual step of every process can be monitored and understood. Why does that necessitate the automations being adopted from the “ground floor,” as you said?
Will: Think of your average medium-sized company. Let’s say with 1000 employees. Now, there are quite literally thousands of different SaaS products at play in the average company. An individual may use several or dozens of them. Email, document management, communication, security, right down to spell checking and printing. There are dozens of things you use on a daily basis, and you don’t typically think about how they all work together to help you achieve simple tasks. Mostly they’re just there. That’s because you, your tasks and your thinking are what is motivating all those programs and services to work together.
Now let's zoom out and look at the usual targets of automation. If you start from “top-down,” then it’s all about squeezing the headcount. Getting four people to do the work of five people. Essentially it’s about using all those tools I mentioned to do more work. Work is defined by hours of labor for a particular economic result. That’s what a lot of automation initiatives are trying to do in the modern workplace. Extract more labor. They’re trying to cut down on staff so they get 20% of the work automated and cut 20% of the staff. That’s a “win” in the short-term thinking typical in business today.
But we intuitively understand, even if we can’t express it, that if you take the work of five people and make four people do it, you’ll lose something. You’ll get the same amount of work, perhaps, but you’ll be doing it with about 20% less creativity. In the drive to eliminate headcount, you have sacrificed an opportunity represented by that extra person. Fewer people in your organization mean fewer minds, and fewer minds mean less creative impulse. Ultimately if you think you can automate everything and eliminate people, and you take that to its logical extreme, you end up with Twitter under Elon Musk: a broken shell of a company that still technically does the same thing but has completely lost its identity or its ability to innovate and adapt.
The results of that kind of mercenary thinking are stark. I think that's provided a great living example of the dangers of the age of automation.
When you eliminate people from your organization, you lose something very valuable: you lose institutional knowledge. You lose instinct. You lose experience. And you lose flexibility. What if you grow tomorrow? What if you need that knowledge? That’s the “human touch” that people can bring. Sure, you can get some people to give you 25% more productivity, but you can’t make people 25% more creative. Cutting headcount kills the creative culture of any organization. An organization without room in it for people to be creative is dying. It’s that simple. If you are automating from the top down, you’re selling up your future for a payoff today. Nothing more.
But: if you start at the bottom and ask each of those people how they would like their work to be done, even to be automated, and then allow those people to design that process and then monitor that process, you’ve done two important things. First you’ve made those people happier and more productive. Next, you’ve also embedded your institutional knowledge and instincts into what you do. Now if you are around to monitor and modify and respond to changes in the market with new insights and new ideas, then you continue to create value for yourself and your organization.
That is why we want everyone to be an automation designer. Because the truth is: where are all the good ideas at most companies? The executives like to think that this is where they step in. But that’s wrong. Executives may be good at big-picture thinking, but they’re usually not good at improvisation. They’re usually not good at creative small-scale thinking. The workers and the middle managers have area expertise. They have intangibles that you can’t measure and that you will miss if you eliminate those people. To make people central to your organization, you need to adopt automation through those people. Not around them and not on top of them: through them.
Lloyd: It’s so true. How often do middle managers have great ideas that never get traction?
Will: As a former middle manager: all the time! So we want doFlo to provide the tools for those in the middle and on the ground to do their work better and more efficiently. Leave them in control of the robots. Don’t replace them with robots.
Lloyd: It seems like a radically new way of looking at automation. How are you going to support that with the product you actually design?
Will: This touches on the other areas where we plan to leverage AI. The idea is to have this a “lucid chart” approach with the automation happening under the hood, without the user ever needing to know that a sophisticated process is even happening in the background. Just like when you’re typing into Google search… you’re not thinking about all the incredible things the data models are doing to come up with your results. You just see the results. That’s the same thing we want to do. We want you to be able to visually design something and have it happen. No intermediate steps.
Lloyd: So it’s a visual AI?
Will: It will use elements of visual AI to analyze your inputs, but there will also be tools for denoting what you want different services to do in your automation flows. After all, you may not know exactly what you want every service to do. You may want to explore still what’s possible. Using a combination of visual AI, chat AI, and a visual design platform, we can offer you a super powerful toolbox to create all kinds of sophisticated processes.
Ultimately entire families of programs could be created in doFlo. We could do all kinds of things beyond just productivity automations. That’s just the most obvious use case at the outset. We think people could use our platform to design entire programs and complex services, which they’d be free to leverage in any way they want to, including by selling them or selling access to them.
You like to write poetry, right?
Lloyd: I have to admit that I do.
Will: Well, imagine you had an idea for a product that can analyze your poetry and do all kinds of cool things, like recommend lines, compare your stanzas to other writers, or give you reading material based on what you’ve written. I’m just spitballing here but imagine that.
Lloyd: Ok, I’m there.
Will: doFlo could allow you to do that, and if you wanted to, to sell or give away that service to others, just like that. No coding.
Lloyd: I’m intrigued.
Will: I’m glad. It’s wonderful to make products that make people into creators. That’s a joy for me.
Lloyd: So, can you imagine a potential future where I come to you only wanting to get certain things done? Maybe not even have existing solutions, meaning existing SaaS services that I'm using, and you're able to actually leverage all those services for me without me having to prepare them or engage them myself?
Will: That's the blue sky vision: you drag and drop visually. You have an onboard little wizard that you can explain what you’re trying to do, and it offers you various products that may be useful in your process, or you create your own microservice right in the platform. You connect these services, mix, match, and explain what you want in plain English in a text annotation.
When the process is running, you can either set it and forget it, or assign a particular service to notify you or do some external step. And that would require next to no human intervention whatsoever. It would then have various interfaces and capabilities that an API provides. You can either monitor it on doFlo or easily just set up a Slack channel to communicate with the API in plain English. It can even talk to you via email, just like a person. It will just work, like a little mechanical Turk working away all the time, doing what you need to be done.
Lloyd: This is where it starts to sound… very cool. Because, frankly, most people have no idea what SaaS services are available. Nor what they do.
Will: That’s the dream. For now, we have many things to iron out.
On doFlo as an Automation-Sharing Platform
Let's make smart companies. Not just smart products.
Lloyd: One great thing about it, is it sort of sounds like GitHub. It’s like the GitHub of automations.
Will: That’s a fair comparison. And we want more sophisticated users to actually use doFlo to create custom processes and solutions for other people. We want it to be a platform where creativity can be shared. Even where businesses can be built around new ideas. If you have a complex process at the heart of something new, you can make a proof of concept get going. I want everyone to be thinking more entrepreneurial and free with their ideas. Let's make smart companies. Not just smart products.
Any programmer knows that new products are just built on top of the open-source community. Having our open-source community and the ability to share innovations is key. The nice thing with us is that the automation and the documentation are sort of “one thing.” If you have automation running on doFlo, you know exactly what it’s doing and all the steps. It’s easy to audit and easy for programmers to modify or copy. With doFlo automations is that you can snip out or wholesale copy any part of a process. You can build out a complex chain of automations, or simplify existing ones.
Once you’ve done an automation flow, you can save it to your library and share it with your team. Or you’ll even be able to sell them on the marketplace. We’d like to have “doFlo Certified Developers” who can sell their automations or their time on the platform and create automations that are more complex than most people want.