Steam Machines must be more than a hobby

Steve Jobs was in the habit of describing the Apple TV - the real Apple TV, that is, not the hypothetical uber-device that's been stalking the imagination of tech pundits and the nightmares of TV manufacturers for years - as the company's "hobby". It sells a few million units here and there, but it's no iPad, no iPhone, not even a Mac. It's a casually dangled toe in the water of a new market whose primary purpose is to extend the functionality of iTunes and iOS devices, rather than being a significant product category in its own right. "Hobby" summed it up; lots of noise and light around the topic accompanied all of Jobs' later keynote events, but really, Apple was just dabbling.

Steam Machines, then, are Valve's hobby. Admittedly, Valve is a company with a lot of hobbies, but Steam Machines fit a similar profile to Apple TV in this regard. The rest of the world, or its more credulous denizens, are waiting with bated breath for Valve to sally forth with a device that's going to cut a swathe through the games market - yet for all the world, everything Valve does looks like little more than casual dabbling. They're mucking around with a custom version of Linux (saying "SteamOS" sounds really impressive until you realise that most people's family pets have their own custom version of Linux at this stage) and experimenting with an intriguing controller design, both of which are fine hobbies - but the much vaunted Steam Machines themselves, thus far, are little more than an underwhelming branding exercise.

"Valve's not about to get into hardware manufacture any time soon. It's not what they do and it wouldn't make sense"

Of course, Valve's not about to get into hardware manufacture any time soon. It's not what they do and it wouldn't make sense. However, the company has a deep interest in ensuring that the gaming PC, as a platform, is in robust health. The name "Steam Machine" is a giveaway, if one were required; Valve needs lots of machines out there for Steam to run on. It has, as I've argued before, become the de facto champion and caretaker of the PC gaming sector, a role long since abandoned by Microsoft. Steam itself is the biggest pillar of Valve's support for the PC, and Steam alone has done a great deal to ensure the continued flourishing of this market. The company's gamepad efforts are an interesting sideline, its dabbling with SteamOS little more than tinkering for now; the Steam Machines, though, we earnestly expected to take a rather more dramatic form when they emerged at CES this week.

In the end, Valve managed scarcely a handful of minutes on stage to introduce the dozen "Steam Machine" manufacturers, each of which is producing its own versions of the system. Gabe Newell deflected all questions to the device manufacturers. Despite carrying the Steam name, it's almost like Valve isn't entirely happy to be associated with the project right now - perhaps wincing at the heavy responsibilities which being seen as a platform holder will inevitably bring. Perhaps the Apple TV comparison isn't fair after all; Apple TV always felt like an under performing but beloved hobby. Not much feels beloved about Steam Machines. Not yet, anyway.

It's not hard to see why the Steam Machines might be unloved, though. They're an ugly and rather ramshackle lot. Their prices range from a console-matching $499 up to an eye-watering $5000, while their case designs range from the functionally ugly through to the kind of howlingly awful rig that inspires mass eye-rolling even at LAN parties. The specifications of the devices, which one might have expected to conform to some kind of standard, or a number of standard "steps" at different price points, cover the whole spectrum of PC performance. This is perhaps the most disappointing aspect of the announced devices - if these were actually meant to attract less hardcore gamers (most core gamers will still build their own systems, of course), then by doing nothing to reduce the mind-numbing complexity of figuring out specifications and component codenames, they have already failed in their most basic task.

"by doing nothing to reduce the mind-numbing complexity of figuring out specifications and component codenames, they have already failed in their most basic task"

If I sound disappointed, it's because I am. I'm disappointed on an entirely personal level, I confess. 2014 was going to be the year I got myself a gaming PC again. I've missed too many games and experiences through not owning one, and I'd love a reasonably small, low-profile box with enough grunt to play PC titles comfortably. I haven't followed PC specifications and components for about a decade and I'd rather perform my own open-heart surgery than build another of the damned things myself, so a Steam Machine looked perfect; yet after this week's CES reveal, it appears that the actual advantages of such a system in terms of reducing complexity (let alone cost, which was always unlikely to be a major factor) are negligible. I'm left wondering who exactly these boxes are for - the core audience will ignore them and build their own systems, while the more casual audience who are eager to engage with PC gaming won't find any advantage in a "Steam Machine" that doesn't exist at any other pre-built PC box-shifter.

Valve may have reason to wish that it was taking this hobby a bit more seriously. In spite of the robust health of the PC games market right now, there are structural issues with the PC market as a whole which present a major challenge to its continued growth and success in the coming years - structural issues which only Valve is likely to be in a position to solve, and for which a Steam Machine style venture may well be necessary. To wit, a primary advantage of the PC platform, namely its sheer ubiquity, is winding down. It used to be the case that nearly everyone owned a PC and thus, nearly everyone could play games, at least to some extent. In recent years, the PC benefitted even more as a gaming platform from the inverse of that argument. Gaming PC purchases were justified in part by the prowess of the system as a multi-function device. A gaming PC was expensive, but also served as the user's primary computer.

Today, the desktop computer is an increasingly rare beast. A great many households only have laptops; a great many more are supplementing laptops with tablet computers that perform much of the functionality that once belonged to PCs. Laptops, too, have changed. Apple's Macbook Air and Google's Chromebook, followed by a steady parade of Ultrabooks and wafer-thin, solid-state imitators, have refocused the desires of buyers away from power and towards size, weight and battery life. When I bought my last laptop, the ultra-thin one with 10 hours of battery life put forward a case that simply couldn't be answered by any hankering for a powerful gaming system. Sales figures suggest that I'm far from alone in making this choice. The resulting device can run some games (it's fine for lots of indie stuff, and Civilization V just about works) but it's certainly not a gaming system. I've never even bothered installing Steam.

"I am not, in any sense, predicting the "death of the PC" - but there's no doubt that this switch away from desktops presents a challenge to the existing PC market"

I am not, in any sense, predicting the "death of the PC" - to do so would be nonsensical - but there's no doubt that this switch away from desktops and towards tablets and Ultrabooks presents a challenge to the existing PC market. I believe that gaming PCs will increasingly have to make a case for themselves as gaming devices alone; a subtle but important change from "here's your next PC, and it's great at games" to "here's your next gaming device, and it can do PC stuff too if you want, which you probably don't". Core gamers won't change their outlook at all, of course, but beyond that group there's a vast hinterland that was once the domain of PC gaming and which now risks disappearing as the technological landscape shifts.

Steam Machines ought to be at the vanguard in terms of counteracting that shift - accessible, attractive, easy to understand gateways to PC gaming designed perfectly to fit into the lives of "lapsed" PC gamers using Ultrabooks, or console gamers looking to branch out, or former core gamers who want to stay in touch but don't have the time or money any more to be deeply involved. Valve, as the operators of essentially the only PC game software distribution platform that matters a damn, ought to be leading that movement. On the strength of this week, Valve knows it ought to be doing something, but doesn't have the stomach for doing much of anything - while left to their own devices, it seems, PC manufacturers aren't capable of seeing beyond their own narrow world of hilariously arcane specifications and desperately ugly boxes. There's an enormous opportunity waiting here to be grasped; so far, Steam Machines have only fumbled.