Heroku's Adam Wiggins on making computers better

The Heroku co-founder is an advocate for bringing out the best in computers, but he's not shy about what's not working in the tech industry today.

Vision, vista, open road
Gustavo Frazao/Shutterstock

Adam Wiggins co-founded Heroku, the pioneering PaaS company, and is the founder of Muse App. He's also had a hand in interesting projects like Ink & Switch and has written extensively on the philosophy of computing at Making Computers Better

I recently had the chance to talk with Adam about the impact of computing, serverless deployment, online identity, and more.

Matthew Tyson: Adam, you are involved in computing as both a builder and a thinker. Did you start as a builder and then become more interested in the philosophy of computing?

Adam Wiggins: Builder first! Early in my career, I was skeptical of Aristotle-style deep thinkers. But I eventually came to appreciate the value of an academic approach to understanding the past, present, and future of computing in our lives. A Bill Buxton quote I like is: “Head in the clouds, feet in the mud.” Dream big, but ground yourself in reality.

Tyson: Your writing in Making Computers Better is thought-provoking. Often, we just want to work on something interesting and make a living but it seems obvious to consider the impact of what we are doing as software developers. How can we nurture that kind of thinking? 

Wiggins: It’s entirely reasonable to be focused on providing for yourself and your family while doing work that you enjoy. But I found myself aware of the privilege granted by having marketable skills in the computing industry and wanted to think about the overall impact of my career. “Making computers better” is a call to action on the spiritual quest I found myself on, hoping that might inspire others.

And I’m not alone in choosing to trade off the comfortable job and earning potential against having more impact and meaning in my work. Andy Matuschak’s experiment as a crowdfunded computing researcher and Brett Victor’s Dynamicland are two examples that I find inspiring.

Tyson: Why is work in computing likely to have a big impact on the world?

Wiggins: It’s now trite to say “software is eating the world,” because we can just look around and see computing as part of everything from commerce to education to politics. But the wild thing is that we may only still be at the beginning of this integration of digital technology to human life and society.

So if you work to change the trajectory of the computing industry just a little bit now, my hunch is the effect may be felt dramatically 50 or 100 years from now.

Tyson: You talk about trying to improve the way identity works online. How do you think we’re doing on that front?

Wiggins: Online identity is a catastrophe happening in slow motion. Internet fraud keeps getting worse—it’s a multibillion-dollar “industry,” by some measures exceeding theft of real property. In the meantime, the security measures our industry has rolled out—email verification, two-factor authorization, password managers, face and fingerprint recognition, CAPTCHAs, public/private key encryption—make the user experience of doing even minor transactions online a confusing maze of steps.

It’s getting worse year by year, all the big attempts (OpenID, Microsoft Passport, etc.) to solve it have failed, and not even that many people in tech seem to be working on it. It’s concerning.

But as a counterpoint, Passkeys / WebAuthn is a promising recent development!

Tyson: Can you talk about your current project, Muse?

Wiggins: Muse is an infinite canvas for thinking. It gives you a mixed-media canvas: imagine a combination of Figma, Notion, and a physical whiteboard.

You can use it privately to do research, strategy, and rumination; or together with your team for planning, retrospectives, and roadmaps. It’s built as a high-speed native app for Mac and iPad, and uses local-first sync so that it works offline but also allows for realtime and asynchronous collaboration.

Computers are very good for the production steps of knowledge work—like typing out a paper in a word processor or editing a video. But we still tend to fall back to analog tools (sketchbooks, Post-its, whiteboards) for the earliest stages of ideation. As more teams move to remote work, we need thinking tools that live in the digital realm.

Tyson: You are a serial entrepreneur. What about startups keeps you engaged? 

Wiggins: I love the process of taking something from just a spark of an idea to something that exists in the world. I love working on small teams. I love to solve problems for people (users/customers) through technology, and at the same time make an artistic statement about how I think computing could or should be.

All of that adds up to enjoying entrepreneurship generally and the startup approach specifically!

Tyson: Any advice for people who want to get into the startup world?

Wiggins: I usually say that entrepreneurship shouldn’t start with “I want to start a company” but rather “I want to solve a problem that exists in the world, and starting a company is the best vehicle for doing that.”

More practically, a good place to start is lurking in communities of builders like Indie Hackers, Y Combinator’s Startup School videos, or following creators of your favorite products on Twitter or Mastodon.

Tyson: I was alerted to the idea of “local-first” applied to software by Denjell (one of the creators of Tauri) and I know you are involved in that movement, as well. Would you mind telling us more about local-first software and what it hopes to achieve?

Wiggins: Local-first software is our answer to how we can improve on “the cloud.” Cloud software (web apps, or native apps which connect to internet APIs) gives us all kinds of benefits such as easy sharing and collaboration. But these apps also come with some significant downsides in terms of ownership over your work as a creator.

My colleagues and I wrote Local-first software: You own your data, in spite of the cloud four years ago, building on a decade of computer science work in academia on CRDTs [conflict-free replicated data types]. Since then we’ve seen that term explode in popularity, with dozens of products labeling themselves “local-first” in their marketing and many developers interested in moving in this direction with their work.

What we hope to achieve is a world where authors, scientists, students, designers, and anyone else who invests a lot of themselves into their creations will have more control and ownership over their work product. My colleague sometimes describes this as “Google Docs without Google.”

Tyson: Heroku was a groundbreaking platform. You describe how you were trying to improve the way we develop software and connect it to an ongoing story including things like Vercel. What do you think about how we develop and deploy software in the present day?

Wiggins: If you write a website or web app in 2023, it’s vastly easier to get it online at a publicly-available URL with HTTPS in comparison to when we founded Heroku (2007). This is in part because you can use Heroku, but also because many products such as Vercel, Netlify, Replit, and the whole serverless movement have expanded on the idea that you shouldn’t have to spend hours futzing with servers to run a simple website.

We also have app stores (iOS and Android) which have their own deployment story, which isn’t too bad, although it has its own challenges with developer certificates and the app review process.

But I think developing software is still too abstract, especially for tinkering on a small scale. I’d like to see development tools that include more direct manipulation and built-in collaboration like we have with other categories of creative tools. Ink & Switch is researching programmable ink as one move in this direction.

Tyson: What are your thoughts on AI like ChatGPT?

Wiggins: I’m not as energized by recent AI advances as a lot of folks in the tech industry, either in the sense of what it can be used for or the potential longer-term risks. But it’s obvious that large language models, image generation, and computer vision are likely to massively change more than one segment of knowledge work. It’s also going to open many new doors for spam and fraud that we’ll have to contend with.

I’m most excited about how these AI advances can be incorporated into our creative tools. Not to have computers do our thinking for us, but as a cognitive aid (or robot collaborator, if you like). This illustration process is one example.

Tyson: What is your greatest hope for the future of computing?

Wiggins: My greatest hope is that computers continue to be the best tools for creative expression and problem-solving that humanity has ever created.

Standing in the way of this are market forces, government forces, and individual user preferences. In aggregate, these push us in the direction of computers as vectors for fraud and social manipulation; or simply turn computers into attention-stealing appliances rather than bicycles for the mind.

I hope that people in the industry who care about this will work together to counteract these forces—and to develop computing in ways that enhance humanity’s prosperity and happiness overall.

Copyright © 2023 IDG Communications, Inc.

InfoWorld Technology of the Year Awards 2023. Now open for entries!