#358: Back in your box

In probably the least surprising news of the year so far, this week Valve quietly delayed the release of the TV-friendly PC Steam Machine, the wireless VR headset Steam Frame, and the redesigned Steam Controller over cost concerns. The explosion of component prices, driven by the tech industry’s genAI death march, has caused a subtle shift in Valve’s language around its plans — from “early 2026” to “the first half of the year” — though the blog post announcement leaves plenty of wiggle room for further delays. “We have work to do to land on concrete pricing and launch dates that we can confidently announce,” the firm says, “being mindful of how quickly the circumstances around both of those things can change.”

Mmm. Hard to say we didn’t see this coming, really, what with prices of memory, storage and graphics hardware going through the roof thanks to AI hyperscalers snapping up everything in sight as they speedrun the global economy, and indeed the globe itself, towards collapse. When Valve announced all this last year I wrote of my hope that it would be aggressive on price; that it would seize the obvious opportunity, stroke the ball into the open goal that Microsoft and Sony had left for them with their price rises and reputational crises, and transform TV-based PCs from a niche proposition to a mainstream one by making Steam Machine as affordable as possible. I can obviously forget about that now. Really, I should have never suggested it in the first place.

In fairness to Past Nathan, I doubted at the time whether Valve would share my thinking. Its previous hardware initiatives have hardly given the impression of a company that thinks too deeply about pricing, or even cares all that much about sales figures. It just likes making stuff and putting it out there, and that’s why we can probably expect it to make good on its commitment to launch Steam Machine et al before the end of June. Valve seems destined to be the first company to release a console with a four-figure price tag and let’s be honest, that would be a pretty Valve thing to do! It will arrive on schedule and sell like cold porridge, the status quo will remain unchallenged, and people like me will lament the missed opportunity. Never mind. At least we’ll get Half-Life 3 out of it I guess (lol as if).

Clearly this is a terrible time — the worst, I would venture, in gaming’s short history — to be thinking about launching a new videogame system. To do so would require you either to not really care about price or profits, a la Valve, or to care about those things a lot but be really very dim indeed. Which presumably explains why Microsoft appears set on releasing a next-gen Xbox next year. Rumours of this have been building for a while now, and this week we got our clearest indication yet that it’s going to happen. Lisa Su, CEO of chipmaker AMD, broke all manner of confidentiality agreements by telling investors that her firm’s work on the new Xbox’s innards is “progressing well to support a launch in 2027.” Lisa! I’m pretty sure you weren’t supposed to say that! Hit Points thanks you nonetheless for your support. This newsletter relies on the unthinking blathering of people just like you. Welcome to the club! Tell your friends.

A new console is a bad idea all round right now: component costs have not gone through the roof so much as breached the planet’s atmosphere, and everyone is skint in any case. But it’s a particularly daft premise for Microsoft, since the value proposition of an Xbox, whatever the price of it, has never been lower. There are already a number of very good reasons not to buy a console from Microsoft: AI, genocide, enshittification, layoffs, price rises, multiplatform exclusives, Phil Spencer's t-shirts, Satya Nadella admitting he gets Copilot to summarise podcasts for him... honestly I could be here all day. Perhaps they just want to complete the set by launching a new console at the sort of price you’d normally expect to pay for a family car? Dunno, you wouldn’t put it past them. A new Xbox console in the current, AI-obsessed climate will waste a lot of money and time and be a complete failure, which is why it is almost certainly going to happen.

Sony, too, is apparently beavering away on its next-gen effort, and while I recognise that it takes several years of work to bring a console to market — and the rumour mill suggests Sony doesn’t plan to pull the trigger on the next PlayStation until 2028, by which point costs might have stabilised a bit, possibly — this all feels a bit premature, does it not? I was telling the eldest on the drive home from school last night about how console transitions have historically worked. You get this brief, magical window where developers have been working on a machine for years and understand it deeply, pumping out games that would once have seemed impossible. Then there’s a period where they actually are impossible — where you can see that developers’ ambitions have begun to exceed the potential of the hardware, things are starting to creak a bit and you’re like, ah, yep, time for a new one. We saw that on Switch, sure. But there’s no suggestion, that I’ve encountered anyway, that either Series X or PS5 are even approaching the peak of their powers, let alone surpassing them.

Admittedly things are different now. It’s harder to reach the old levels of experience with a single console in an era where games take so long to make — where a triple-A studio making two games in a decade can be fairly described as prolific — and when so much of the industry’s dynamism and creativity comes from the indie scene, which typically lacks the resources, the need, or even the desire to push a piece of hardware to its limits. But all that just reinforces my point: I’m fine with the consoles I have, thank you very much, and so is everyone else I know. If the cost apocalypse tells us anything it’s that the industry should probably put its next-gen plans in the drawer for a bit — and hope, like the rest of us, that this insanity might one day conclude.

MORE!

• Shortly after I hit The Big Scary Send Button on last week’s edition, shares in game companies took a tumble following the release of Project Genie, a Google AI tool that lets users generate, and explore, virtual worlds from text prompts. Creations are limited to 60 seconds in length, run at N64-era framerates and invoke all manner of copyright-infringement risks, but investors took one look at it all and decided the sky was falling in. Idiots. (Later that evening, I bought a couple of hundred quid’s worth of Take-Two stock, thinking, that’ll show ‘em. At the time of writing I have lost £20. If GTAVI doesn't come out this year, we riot.)
• Switch is now officially Nintendo’s best-selling console ever, with the Mario maker’s latest fiscal results revealing lifetime sales of 155.37m units by the end of 2025, taking it comfortably past the DS’ paltry 154m. Switch 2 sales continue to outpace those of its predecessor despite a minor slowdown over Christmas; with 17.37m units sold by year’s end, it’s already beaten the lifetime sales of Wii U lol. Lovely stuff — just try not to think about the share price.
• It has been precisely 14 seconds since our last AI-flavoured outrage, so let’s hand things over to Sony’s R&D department. The PlayStation firm has been granted a patent for what it calls “LLM-based generative podcasts for gamers”, which seems like a strong contender for the worst sentence ever written until you read the next one. Sony intends to have its language models “generate a podcast presenting the news in the voice of a videogame character of a videogame played by the videogame player.” Horrific.
• Ubisoft union representatives have called for the head of CEO Yves Guillemot. “It’s his company at the end of the day,” rep Marc Rutschlé said. “I think the level of hate people have for him [means] he should move on. Then we could build some sort of trust again.” Seems highly unlikely if you ask me but hey, the French know a thing or two about guillotines.
• Goichi Suda put his foot right in it this week by claiming NetEase, parent company of his studio Grasshopper Manufacture, had abandoned efforts to crowbar genAI into its games. “They originally had a section researching and developing AI-related stuff,” he said, “but at one point they decided not to do that any more. They folded that section and told their studios to not use AI in games, to not use it at all.” Sounds cool! What say you, NetEase? “The assertion that NetEase Games has closed an AI department or that we've mandated teams or studios not to use AI is not true,” a spokesperson said, before politely pointing out that two NetEase staffers will give AI-flavoured talks at GDC next month. Awkward!
• Lastly, hats off to Hit Points chum and The Guardian games editor Keza MacDonald, whose new book Super Nintendo: How One Japanese Company Helped The World Have Fun is finally upon us. (I think she first told me about it four years ago? I often wonder whether I should turn my hand to writing a book but, lawd, that's a very long time to be doing one thing, isn't it.) Keza is a very nice person and a wonderful, accessible writer who knows more about Nintendo than just about anyone, and while I haven’t read it yet — she didn’t send me a copy? Weird? I suppose I should have asked for one — I heartily recommend it nonetheless. It’s on sale in the US already; it doesn’t hit UK shelves until next week, but you can tide yourselves over with a lovely Keza guest spot on the latest instalment of the ever-essential Back Page Podcast.

There you go! Please insert the usual waffle about paid subscriptions and the obligatory Tip Jar shoutout, I am tired. That stuff really is very important though. If you've enjoyed today's edition, it'd be great if you'd considering supporting it. Maybe forward it to a friend and suggest they sign up? Every little helps. Anyway, have a cracking weekend, and I’ll catch you next time.