24 Jan 2023

Funding software innovation

I'm Jacob O'Bryant, maker of Yakread and other tools for online speech. I write about practical ways to make the internet better.

Startup accelerators like Y Combinator leave a lot to be desired, and not just because I'm salty about my past 8 applications being rejected. For one thing, accelerators obviously don't do much good for open-source developers or anyone else who is mostly interested in building public goods. But even if we restrict our focus to those who want to be founders, the reality is that it can take years for a nascent idea or desire to mature into a real business. YC may be focused on early-stage companies, but sometimes even an early-stage company takes a long time to build.

That's been my experience, at least. My 4-year anniversary of quitting my job and going full-time on entrepreneurship was sometime last week. I've only been able to go at it this long thanks to, in order of significance, 1) living with family since the pandemic started, 2) my wife continuing to work until the pandemic started, 3) savings from the one year I spent as a regular software engineer.

This has been workable in my case, but it points to a lack of institutional support. Not everyone has the resources (or stakeholder buy-in) to take a 4-year sabbatical. We have institutional support for research—what if there was something like "going to grad school," but for open-source developers? Could it be scaled up enough so that people can actually get in without applying 8+ times?

I like the idea of grant programs like Emergent Ventures and O'Shaughnessy Fellowships, the latter of which I just discovered and will apply to shortly. But even these, while not limited to growth-ready startups, are still focused on a shorter time horizon than what I'm envisioning. No one expects a new grad student to know what the subject of their thesis will be. I'm also not confident that there will ever be enough grants to meet demand.

What if part-time jobs were more of a thing? Work 2-3 days a week, get the equivalent of a grad student's stipend, and spend the rest of your time incubating your own ideas. If you get a grant, or get into YC, or get sponsors, then great; you can go full-time. But even if you don't, you don't have to just give up or live on savings.

Gergely Orosz wrote a good thread about why part-time work isn't very common in the software industry. The concluding tweet:

If you want to work part-time as a software engineer, your options are: find one of the very few places like Gumroad, contract few days/week, or start your own, part-time business.

As a company: it’s possible. But there are tradeoffs: your competition will likely move faster.

I don't know if part-time jobs are the panacea I'm looking for, but Gumroad's example is encouraging. Maybe it's a good fit for situations where trying to get big fast doesn't work anyway. At the very least, I have a hunch that it'd be worthwhile for more people to build companies around asynchronous + part-time work, and we'll see how far the model can be pushed. That's what I'm planning to do with my own business.

If part-time work can be institutionalized to a greater degree, I think it will be especially helpful for people and ideas that tend to get filtered out by other funding methods. Not every worthwhile project looks ambitious at the start. But no one can reject your idea if it's not part of the application—or if you haven't even thought of it yet.

Discuss this post on Discord.

Subscribe to this newsletter.


I like newsletters more than tweeting

I've been reading Indie Microblogging by Manton Reece, the author of Micro.blog, which from what I gather is kinda-sorta like an open version of Twitter. A few snippets from the introduction:

We focus on microblogging because it’s the easiest form of web publishing.

and

Sometimes it’s hard to put words down. Staring at a large, empty text box from blog software that wants you to give your post a title, as if every thought is fully formed before you know what to say. Doubt and imposter’s syndrome creep in.

The solution is easier, quick posting. Just start writing without the pressure of getting it all right.

and this from Part 2:

This is why we should start with short posts. They represent the majority of content on silos like Twitter and Facebook, and they’re easy for anyone to create, without the often daunting task of thinking about a whole web site. With better tools and platforms, people can have their own web site as a default outcome when microblogging, rather than as a chore and technical hurdle.

That all seems quite reasonable to me, though for whatever reason I feel like an exception to this model. Short posts are harder for me to write than longer posts. I love the format of a weekly newsletter, putting my ideas together in batch. In contrast, I never posted much on Twitter, and that hasn't changed now that I've switched to Mastodon.

For that reason, I've often wondered if having an open-alternative to e.g. Twitter actually matters—maybe we should instead try to get more people to do low-key weekly newsletters. Email is already open. Back-and-forth discussion can be relegated to small individual online communities.

An interesting benefit of the weekly newsletter format is that it introduces an additional layer of curation. I only write about thoughts that are interesting enough to survive in my head until the next time I'm writing the newsletter. Similar for links I share. So there's less content total than if I were to post everything as it came to me, with average higher quality for the things I do share. Maybe a world where everyone writes a weekly newsletter has less information overload than a world where everyone tweets all the time.

Maybe what people need to get started is structure. Instead of getting a blank canvas after you click "New Post," there could be some kind of default template. I've thought about that a little bit in relation to Yakread—it already has a feature where you can easily copy-and-paste the links you've thumbs-upped in the past week. I use it for the "Stuff I read last week" section below. Maybe Yakread could eventually have a newsletter product built-in to make that process more automatic.


Blog posts as search queries

This Escaping Flatland post was a good reminder about what to aim for when writing and how to think about distribution as a non-internet-famous person. It seems particularly relevant to people for whom writing is a supplement to their work (like me), instead of being a revenue source in itself. But maybe it's just as relevant for aspiring full-time writers too. My summary-paraphrase is "don't worry about getting lots of subscribers."

From the author's bio:

I’m an anthropologist, programmer and father of two daughters who I home educate with my wife and our friends.

Part of my ethnographic work is studying the extremely online—in particular, people who use the internet as a research platform. I look at how they collaborate, form communities, and process information. These cultures of researchers and tinkers are some of the more powerful vehicles for learning I have seen. Can we scale them? Can we figure out what the bottlenecks are that keep people from accessing high growth cultures, and redesign or tools and norms to support more people?

No wonder it was an interesting post! I'm even in the same demographic, minus being an anthropologist. My two daughters aren't old enough for school yet, but so far we're planning to homeschool them since I had a great experience being homeschooled myself.


Memrise is great

Not strictly related to tools-for-online-speech, but I downloaded a language-learning app called Memrise a couple days ago and have been impressed. I've wanted to learn Chinese for a while since it's my wife's native language. I've tried a couple language-learning apps in the past, but they felt like they had been optimized for engagement/retention rather than being optimized for getting you to actually learn the language. You look at the pictures, click the buttons, it feels like a game. You get a sense of progress as you go from level to level as long as you never think too hard about if you're actually becoming fluent. Whereas Memrise feels like they're actually optimizing for student outcomes.


Stuff I read last week

Also check out my recommended newsletters.

Practical steps towards a better internet.
Made with Platypub. This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.