Tell that to Tim Sweeny please. Epic is integrating Ai right into Unreal Engine 6. And they criticize everyone who wants a disclosure of Ai on the shop page, when purchasing the end product.
From what I’ve heard, even expert computer scientists point out that vibe code isn’t actually better code than human code. It’s not faster or more efficient or more elegant. It just looks better.
As for visual art, the process is still to describe in exact detail what you want (using LORAs and quantifiers to indicate how much you want of a feature) then generate dozens of examples, and have a human being curate the desired results…
…often picking those out from among numerous cosmic horrors that the human curator then cannot unsee.
Way to bury the lede. 🖕🏼
The executives “want it to be” because they’re paid to, and any outlet failing to state the simple fact outright belies their hidden bias. (ie. funding, influences, kompromat, etc.) The shareholders want it to be, and likewise up the chain.
Follow the money trail. It all leads to billionaires, fellow plebs of the world. Guillotines, roll out. ✊🏼
How is it burying the lede when anyone with basic reading comprehension knows that’s what the headline’s obviously suggesting? Like sorry your reading comprehension is apparently trash, but that doesn’t mean GamesRadar did something wrong.
ew. that other place’s leaking.
The execs really want it . . . . so payroll can be eliminated.
Take what’s happened as a warning to prepare yourself for the day humans really are replaceable. Start making significant changes to our societal structure now, or stash yourself some cyanide (or similar means of escaping this shitty plane of existence) - your choice.
Currently, the technology still needs human beings to filter out the unusable examples. And generating the content takes far more resources than hiring a human being to do the same thing.
An industrial revolution historian on Wired noted that when we invented the power loom, it produced better textiles than workers weaving it by hand. When Ford created the assembly line, it commonly produced better material goods (like cars) than a single person. The massive steel factories produced higher quality steel.
We’re not seeing the same increase of quality yet, and it appears we’re a long ways away from being able to scale up AI use for its implementation in most of the economy. It’s the difference between a aeolipile and a 19th century steam locomotive. Or between a manned mission to the moon, and an occupied Mars colony.
I’m pretty sure this technology will never have its moment in the spotlight because it can never, ever live up to the expectations that marketing has hyped up.
After multibillion investments and a database that encompasses nearly everything humans have ever produced, the lofty promises have, as expected, failed to materialize. It’s simply nothing more than a tool that can certainly be used profitably. But its areas of application have absolutely nothing to do with what was promised to investors. It is simply not artificial intelligence, but a search engine that reproduces existing data or, upon request, assembles it into something that looks like it’s new, but is actually just a collage of what the model was fed.
The fact that humans, in their way of being creative, naturally never operate outside the bounds of their own experience was exploited by marketing here to portray the probability calculations performed by LLMs as comparable to what we understand as intelligence. For LLMs, however, this is never achievable. It is merely an illusion built on the lack of understanding among so-called decision-makers.
A dragon age vet is the person who puts the band aid on paarthurnax’ boo-boo.
