In the day that I have had access, I gave the new model a wide range of tasks.
A friend asked me to make a memorial image of her recently deceased cat along with two favorite toys. It crafted an image that looked like a highly personalized sympathy card.
It elegantly took two photos from my wedding and made it appear as if they were in an old-style photo album with photo corners.
My colleagues suggested a poster for a fictional event. I decided to create a Mike Allen look-alike contest in Washington Square Park this Sunday. (Of course, it’s only fictional if no one shows up.)
It also made a handy infographic making “the case against candy corn” which I used unsuccessfully to convince two colleagues that the treat, which is neither candy nor corn, is also not good.
The full pages it designs are scary good. I’d go so far as to say this is definitely a shot across the bow for design work.
If you’re used to absurd lettering and poor design decisions, the output included in the story suggests otherwise.
People who are not writers think AI can replace writers.
People who are not translators think AI can replace translators.
People who are not developers think AI can replace devs.
You think AI is “a shot across the bow for design work” since you are not a designer.
Everyone thinks AI can do other people’s jobs because they’re not experts in those areas.I wonder what your profession is, and whether you think AI could replace you and deliver the same quality of work.
I spent two decades as a newspaper writer, editor and designer and have won national awards for all three. Try again.
Also, it doesn’t matter if it’s 100% as good as what I can do. Corporate will readily settle for 80% for a fraction of the price and time.
And the public was wowed by seven-fingered hands when image generation first came out. Their threshold is even lower.
Those examples are instantly recognizable as slop to me
It doesn’t count if you were told first.
Need to try again with double blind AB testing.deleted by creator
That’s asking which one someone likes more, not which they think is AI. It’s a subtle but important distinction. For me it was a tossup. I picked the AI 3 to 2.
In reality all the passages felt like over-written attempts to say something simple, as poetical as possible. I’ve never really liked that style of writing. I try to be much more direct.
But generally, yah. That’s what I mean. You need that kind of test to know if you can tell something is AI or not.
deleted by creator
OK. What are the clues?
Mostly the art and the fonts if I had to guess? It was more of an uncanny valley visceral reaction though.
I’ve been out of the design game for six years at this point. Given that I did broadsheet design and wince at most everything for both art and typographical choices, it wasn’t really the sort of thing that stood out to me. There is a lot of bad design being happily gobbled up … hell, I didn’t look at rave flyers in the '90s and think “this is great design,” but rather “how the hell did they have the budget for C2S in CMYK with spot fluorescent and a top coat?”
I didn’t say it was bad, per-say, I don’t know enough about design to judge that, it just immediately triggered the “this is ai” bell in my head when I looked at it.
Would it have if you’d not been told it was ahead of time, though?
I think so based on how much it stuck out at me but I can’t know for sure obviously
Both had “I’m a junior designer” vibes about them. If you saw the first few pages I designed in college, those were far worse. Of course, we didn’t have the luxury of CMYK, let alone just coughing up RGB shit. I’m used to seeing execrable design.
Your bell was primed because they told you it was AI in advance. That why mentioned AB testing. If you saw these, side by side with other human made posters and flyers, how reliable would your bell be then? We don’t know.
