Tools and Platforms

Here’s a question I’ve been asking myself when faced with a new (to me) piece of technology: Is this a tool or a platform? In some cases, the answer is so obvious that I don’t need to ask; an old film camera is pretty clearly a tool. But I’m thinking more of those products for which the words “tool” and “platform” are interchangeable. Is Substack a tool or a platform? What about Figma, or an iPhone?

Maybe it helps to place tools and platforms at opposing ends of a spectrum. Here is a list of objects in order of increasing platforminess:

• A screwdriver.
• A ratcheting screwdriver with interchangeable heads.
• A corded electric screwdriver.
• An electric screwdriver that shares a rechargeable battery with a suite of other tools.

Each variant does what the thing above it does with an additional benefit. But each variant also has a greater degree of lock-in. You can’t use the interchangeable screwdriver heads unless they fit the hexagonal socket. You can’t use the electric screwdriver unless you also have a power supply. You can’t use the rechargeable one unless you have a specific power supply from a specific company. By this definition, even the first object, a plain manual screwdriver, has its own degree of platforminess in the types of screws – Phillips-head, etc. – with which it can interface.

A theory: the further down the list we go, the more we think of it as Technology.

Platforms often emphasize benefits like convenience and ease-of-use, while also claiming to solve concerns with less-platformized variants. If you keep cash in a physical wallet, you could misplace the wallet. Someone could steal it. Your house could burn down while it’s sitting on the counter. But while a digital wallet (whether that’s crypto or an online bank account) can protect you from the specific concerns of keeping your physical money in a single physical location, it still carries similar categories of concern, often with compounded effects: you could misplace your access to it (forget your password or seed phrase), someone could steal it (from halfway around the world, by phishing your info, by hacking the bank or exchange linked to your account, by hacking the hardware device you use to protect it, by air-dropping a token into your wallet that on interaction drains your entire account). True, your dogecoin will be safe if your house burns down – unless, that is, your passwords and seed phrases are stored on an internet-disconnected hard drive in your house when it burns down, in which case you, my friend, basically have a digital mattress.

When we talk about platforms in 2022, we mean networked platforms, and – I’m not sure who needs to hear this – network effects are not always positive. An AWS outage can bring down your design studio’s communication system, your bakery’s cash register. A musician selling their music in person fifty years ago – or even online five years ago – didn’t have to worry in the same way about subsidizing a podcaster spreading misinformation. A person writing a weekly pickle newsletter didn’t have to worry about their newsletter platform giving six-figure advances to anti-trans voices. Yes, there are instances where traditionally-published authors have pulled their books from their publishers to protest similar advances, but that just shows that with platformization, concerns that were once relegated to specific industries can now affect people who think they’re signing up for mere tools.

Big Tech likes platforms because they’re more ownable, monetizeable. Lock-in is a competitive advantage. Venture-pathed startups start by building tools (or more-tool-like products), and as they accrete users grow more and more platform-y. Dropbox goes from a tool to sync your files to being a creative collaboration platform. Instagram goes from a tool for applying vintage photo filters to being a visual sharing platform. These companies’ descriptions for what they do (and even their names) become more and more vague, because they become less and less tool-like (for tools, particularly good ones, are predicated on exactness).

Maybe it would be more accurate to say that platformization transfers attention away from the tool and its uses to a growing space around the tool – from the sword to the blacksmith shop to the surrounding town square, eventually the whole bustling town – torch-wielding mobs and all. Tools want to be used; platforms want to be hung out in.

This is not to say that we should go back to only using the most tool-like of technologies. I’m more interested in questions like: How can we better discern between tools and platforms? When does it make sense to choose a platform over a tool? How can we build healthier platforms? How can we anticipate and hedge against some of the those unintended platform concerns?

Maybe those of us who are enthusiastic about technology but also skeptical of platformization could benefit from a looser, more-curious version of Wendell Berry’s criteria for tech adoption – less hard-and-fast rules and more a set of sensible defaults. Such as:

• Favor tools over platforms, at least to start. If learning to draw, better to start with pen and pencil than a full-fledged iPad illustration app. Figure out what you want out of the tools first, before getting a whole art studio.
• Favor platforms that gracefully degrade. Ask what happens if some aspect of the tool/platform stops working. Or if the company that makes the tool/platform goes out of business, or if you want to leave. Can you export your data (and do something meaningful with that data, ie. can you easily import it somewhere else?) Speaking of film cameras, the ones I’ve come to favor have batteries, but only for their light meters. If the battery dies, the camera is still usable. Your exercise bike should still work if the internet goes down, or if Peloton goes bankrupt.
• Favor platforms built on open standards. Related to the above, I’ve been journaling and writing early drafts in Obsidian, which layers its features onto human-readable markdown files saved into a folder of my choice. I can simultaneously search through and edit these files using other tools like NVAlt. Should Obsidian ever fold, I can open the files with any text editor, without needing to export them from a locked library.
• Treat platforms as though they were tools – with specificity of purpose. Ask what you hope to get out a particular tool/platform. (Sometimes, your answer might be: a place to hang out!) Facebook has become a much more pleasant experience for me since I started unfollowing accounts that weren’t local friends and businesses. Glass is an example of a photo-sharing platform that encourages specificity of purpose; its limited tag set dissuades illustration and memes, which makes me want to share my more-artistic film photography there (and use Instagram instead as a hang-out space).
• Ask: How would I react to sharing this tool/platform with someone with opposing beliefs? I’m guessing that many folks wouldn’t be bothered by learning that an archenemy rides a bicycle, or even that they use Mailchimp for their email newsletters. But you might be more disturbed to find out that you’re alongside them on Patreon or Substack (a tell here is that in these latter instances, your followers’ accounts live on those platforms, rather than directly with you). In which case community guidelines and content policies matter all the more, and you may need to be prepared to exert pressure on the platform itself. Or leave.

Thinking about tools and platforms, I can’t help but also recall the walled-garden Web 1.0 services like CompuServe, Prodigy, and AOL. Like today’s platforms, each service had their own versions of the same features; instead of video Stories it was chat rooms and Bulletin Board Systems – an era/vibe Joanne McNeil captures so well in Lurking. Reading her book transported me to my middle school library, to huddling with friends around a 14-inch CRT screen as one of them signed into a Prodigy chat room using the Hackers-inspired moniker Zer0cool. A few years later, I would be dialing from our family’s home office into AOL, then immediately minimizing the AOL window to pop open Netscape and surf the web.

Sometimes, platforms, when they get large and widespread enough, spawn protocols – protocols that persist long after the platforms themselves have died out. But that’s a topic for another week.