Alex Payne writes online here.

See also the archive, publications, events.

An individual post follows.

What Pro Computing Could Be

Since the announcement late last week of revised MacBook Pro models it has become clear that a certain type of person is not happy with the first major revision to Apple’s flagship portable computer in some years. Those complaints have sparked broader speculation amongst the twittering classes as to whether Apple is still a company that makes computing devices for “pros”: software developers, graphic designers, video editors, 3D modelers, audio engineers, data scientists, and others whose livelihoods depend on fast machines with plenty of memory and myriad ways of moving data in, out, and around them.

I have one friend in particular who I talk about these sorts of tech product announcements with. Last Thursday’s Apple event was just the latest in our long chain of messages on the general topic of pro computing. It’s a subject we have strong feelings about.

In conversation over two years ago, we converged on an assumption: Apple and Microsoft will taper off their investments in pro hardware and software while they chase bigger, easier money in the consumer and enterprise spaces. History being known to repeat itself, we saw a moment on the horizon not unlike the one in which NeXT and Be emerged. We figured there were some instructive lessons from the histories of those upstarts. We wondered if the right team could move pro computing forward today, and by measures beyond small increments.

My friend and I then started sketching a wholesale rethinking of professional computing in the form of an integrated hardware and software platform. We quickly became obsessed. We tore through journal articles and whitepapers on topics like heterogenous computing. My friend started designing a motherboard. We sourced components. We wrote the beginnings of an operating system (for the trainspotters: in x86_64 assembly and Rust). We worked on the system’s concepts up and down the stack, from UI to networking to a new toolchain. We tried to get into the heads of pro users while throwing out preconceptions of how an OS works and functions. It was probably the most interesting thing I’ve ever worked on – and we were just getting started. Judging from the feedback we got when we shared our ideas with friends and colleagues, we might have been onto something; it got people excited.

After what was essentially an involved thought experiment, I remain convinced that pro computing is up for grabs. We’re in a moment where comparatively cheap hardware is underutilized by aging, bloated operating systems and software. Microsoft made some slick announcements last week, to be sure, but those are more about improving the ergonomics of some types of digital creative work (mostly visual). There are such bigger opportunities to be had. As thinkers like Bret Victor have envisioned, the way computers and the devices that interoperate with them function is so ready for new approaches. What justifies having a general purpose computer running a general purpose operating system in your workspace in an age of connected embedded systems? A pro machine better be doing something special to earn a place on your lap or under your desk in such a world.

It should be no surprise as to why nobody has attempted the sort of ground-up overhauling of pro computing that we mapped out: it’s expensive, slow, and risky to do something big, new, and different. What’s more, we lack the innovative institutions that previously provided cover for the future to take shape: PARC, Bell Labs, et. al. It’s not as if the ideas and talent that will move pro computing forward aren’t out there; it’s a matter of bringing the resources together and having the patience to see something incredibly complex come to fruition. You don’t need me to tell you why that confluence of factors is a rarity in technology today.

After several months of intense collaboration, life pulled my friend and I in different directions. He’s applying his interest in high-performance computing to the essential problem of curing a vicious disease. The team he works on is making great strides, which in turn underscores the quandry of why pro computing isn’t liable to improve dramatically any time soon: it doesn’t necessarily need to, because pros can still get shit done with the tools they have.

Pros don’t quit because their tools are suboptimal. That’s practically the definition of “professional” – a pro gets the damn thing done. A pro user might gripe about the new MacBook Pro, but the next time she needs a new machine, she’s going to buy one anyway because that’s the path of least resistance and she needs to get back to work. That cycle of dependence, along with the need for stability and predictability in one’s tools, makes product incrementalism the norm in pro computing.

To break that cycle, a company needs to introduce a product that promises such dramatic improvements in productivity and expansion of creative possibility that it’s worth the investment of a pro’s money and time into something unfamiliar. Breakthroughs happen, but rarely, and are usually allowed to come to market only once every last dollar has been wrung out of extant innovations. Don’t have a category-defining new product? You can always graft the smart watch platform you developed onto the keyboard of your flagship laptop. Your customers will be forced to upgrade to it eventually so it’ll look like a win in a few quarters. Good enough!

Apple will only see an exodus of pro users if it turns out they’ve shipped a machine that truly can’t meet the needs – the actual working limits – of their customers. Armchair grumbles about misfeatures, memory limits, and the wrong ports aren’t the same as being totally unable to do your job because your tools have utterly failed you. I don’t think that’s where most pro users are today, but some are starting to recognize that today’s professional computing tools aren’t likely to carry us forward into new ways of working.

What I see in the frustrated blog posts from the past few days could be the first rivulets of a sea change … or we could be right back here four years from now. As was clear several years ago, the opportunity for a new approach to pro computing is opening up again. Whether anyone will take it on is another question entirely.