absolute state of unreal 2026 – joe wintergreen internet zone
Skip to content
absolute state of unreal 2026
I guess they just had the State Of Unreal at Unrealfest, and as predicted, it was a real fuckfest full of dogshit. Unrealfest is usually a fun event, but by all accounts the vibes at this one were not great, which is what you get when your opening keynote is about the war you’re declaring on creativity and how your event’s most passionate attendees will be first against the wall. What they’ve done – incorporating generative AI, deprecating Blueprint and pushing Verse – is what you would do if your mission was to sabotage Epic’s business as completely as possible while causing as much collateral damage as possible. It’s genuinely absurd, beyond belief, the product of AI psychosis and the regular-degular psychosis everyone saw Tim Sweeney’d started incubating at his Steam Dev Days keynote back in 2016.
This is potentially existential threat level stuff for the engine, certainly for Unreal as a brand. It’s propelled Unreal into the Unity-a-few-years-ago realm of almost total uncertainty. People don’t know if it’s even safe to use it or train in it anymore. Here is the Actual State of Unreal as declared by me, someone saner than Tim Sweeney. Skip to the end for the facts on "should I still use it".
So for a while it’s been generally pretty well-known that people inside Epic are suffering. There’s been thousands of unforced seemingly-random layoffs, and enough gullible higher-ups have taken leave of their senses that there’s now substantial pressure on employees to use "AI" in their work; even if they hate it, even if it’s slowing them down, even if it makes the work worse, even if it costs more money. This is cult stuff, obviously – nobody is even acting in their own self-interest here.
Every company using "AI" has split its workforce into two groups: those revealed as useless hacks, and those who suddenly want to kill themselves. Epic has a lot of the latter, and they’re scared to speak up, because Epic pays better and offers better health insurance than any other job you’ll ever get. Even so, people have started leaving, and that will continue. One of the plainest red flags is Sjoerd De Jong leaving after 12 years at Epic and over 20 years of being an Unreal expert known and respected by all. I don’t know him, but this seems like somebody who doesn’t leave unless things are really bad. If nothing changes, this will become irreversible. The best people will keep leaving and the worst people will remain and be hired, and past a point that’ll be it. It’s possible to get a company into an unsolvable state.
To really understand how badly they’ve lost the plot with Unreal, you have to understand what they had right for so many years, and you probably don’t. So here’s what it is: They started as a small video game company. They made Unreal for their own needs. They showed it off. People wanted to license it, and were willing to pay a lot of money. They realised there was better money sooner in licensing than game development, and made that their primary business, which meant getting as many clients as possible and building an engine that would serve the different (often conflicting) needs of multiple clients at once. They did this better than anyone would expect – by all accounts support from Epic for full licensees was bending-over-backwards excellent for a very long time.
It isn’t glazing them to say nobody else did it like this. The engine got better all the time because it was driven by the development of Epic’s own games and like a billion other massive games. It introduced new features and was very careful about disruption when deprecating old features. It was impressively dedicated to making tools that moved tasks off of programmer’s plates that fundamentally ought not to be programmer tasks at all (their material graph predates anyone else’s by like a decade), it empowered artists and designers. Blueprint ended up kicking off a million programming careers.
This all persisted once the engine was opened up to the public in 2014. Under the public license, you didn’t get UDN access or most of that great support, but you’d benefit indirectly; the changes made for licensees would still end up in the engine. Epic’s main business was still custom licensing, they knew where their bread was buttered.
Once Fortnite blew up, that bread was buttered elsewhere. Longstanding features started to receive less attention; the major at-a-glance indicator of how solid a newer feature was started to be "are they using it in Fortnite". Things never got bad, but there was a vibe change. There used to be indie dev outreach, there isn’t anymore. There used to be community involvement in all sorts of things that there isn’t anymore. Every aspect of Unreal that involves a website started to get consistently worse – the docs, the forums, the marketplace, the Epic Store.
With more money than God, Sweeney started using...