@mattwelter 50-100 instantly, as long as it's actually warmed up
English
Anton Medviediev
486 posts
Anton Medviediev
@AMedviediev
Software engineer, gamer, and pop culture enthusiast. Currently building https://t.co/Q3BBLSnj9p
Katılım Ekim 2010
135 Takip Edilen57 Takipçiler
how much would you pay?
if reelfarm provided fully warmed up US tiktok pages ready for you to post to
$10 per page?
$50 per page?
$100 per page?
English
@MattVidPro The problem is that as soon as this gets normalized, turning it off will mean PS2-level graphics. Same thing happened with DLSS 1 - devs just stopped spending effort on optimizations and now nothing does 60fps without dlss
English
Did you know? You can just turn DLSS 5 off? Did you know it won’t be enabled by default? Did you know people used to whine and complain about DLSS 1 until the technology matured and now it’s just normal?
Follow the crowd, make an emotional tweet and get free likes.
English
@CMoney4Ever Lmao, of course a corporation is on board with something that saves them money, what does that have to do with us not liking the end result
English
DLSS 5 is Literally not changing the models champ. Just adds the realistic lighting that devs like the ones working on Resident Evil Requiem are OBVIOUSLY trying to achieve. Considering that Capcom, the actual developers, are on board, you might want to rethink your take 😬
K-Med@K__Med
There's a absolutely no way that DLSS will ever know "the artist's true intent". DLSS 5 might look okay in some games, but this is definitely crossing into different territory.
English
As an AI enthusiast, my problem with dlss 5 is that it changes the image instead of enhancing it. Look at Grace here - she gets a different lip tecture, dark roots, and a ton of extra wrinkles that age her 10 years.
I get the argument that developers get control, but they only get to change the intensity of the effect, not what it does
English
@jonathanzliu Like cal ai wasn't just a random number generator lol
English
bro made a cal ai clone but forgot to make the core feature actually work 💀
Alex Nguyen@alexcooldev
My first failure of 2026. I was very confident about my AI calories app, but I made a mistake. The onboarding was too short and the core features didn’t stand out enough compared to other calorie tracking apps. I started organic TikTok marketing and it actually worked pretty well. Traffic was coming in. But almost 100% of users started the trial and then canceled. Conversion = 0. So I decided to stop marketing for now and focus on shipping v2.0.0 to make the product much better. Sometimes the problem isn’t distribution. It’s the product. 🥴
English
@levelsio For automated stuff on the backend where you want the behavior as consistent as possible, my prompts are still long-ish.
When I am iterating, it's always 2-3 sentences max.
English
My prompts are really quite short and it just understands what I mean most of the time, I talk to it like a friend, I see other people use giant prompts and wonder if it actually makes a difference?
Joan Capdevila@joancapdevila
@levelsio the good thing about talking to him is that you can give him a lot of extra context that really makes a difference and helps understand what you want for that task.. i also feel weird about it but it's way better than typing.
English
hyperfocusing on writing a new book with my own app hits different
English
@eliana_jordan I was supposed to be building my writing app, but instead I'm writing a book in my writing app
English
What's it called when you're smart enough to know something is bad, but not smart enough to actually fix it?
Because I think I just annoyed Claude Code with so many revisions it had to tell me to fuck off and leave it alone.
English
@shawmakesmagic People can't even be bothered to make their own coffee and you're telling me they're going to make apps. Yeah right.
English
Here’s the deal
You now have godlike powers
But nobody wants to buy anything you make
Because they also have godlike powers
It’s fine. Get used to it.
English
@Gandalf_da_Grey @ExtropianEric @Alicoh1 Sure, there's many theories, my favorite one is the preferred universal frame
English
@ExtropianEric @AMedviediev @Alicoh1 Please explain how you resolve the paradoxes created by FTL. I’m all ears.
English
the speed of light is a hard limit and any sci-fi future that involves FTL travel or even communication is simply off the table for our species.
English
The hubris of mankind truly knows no limits.
Go read the replies.
We gained sentience like 5 minutes ago in galactic terms. Our understanding of the fundamental laws of the universe has been proven wrong time and time again. And yet we have a bunch of people confidently stating that our current theories (which have gaping holes in them) are the final complete and correct ones. We can no longer be wrong and nothing new will be discovered ever again.
Truly baffling.
Chris 🇺🇲💦🍑@Alicoh1
the speed of light is a hard limit and any sci-fi future that involves FTL travel or even communication is simply off the table for our species.
English
@MeisterBanker @Alicoh1 Well math shows that an Alcubierre drive is somewhat plausible, so this would be an engineering issue too
English
@AMedviediev @Alicoh1 There was no law of physics that said this wasn’t possible, this was just something they thought andnis more of an engineering issue
English
@kilianhekhuis @ChitandaOSRS @fear_nd_loathin @KadyMuhammad1 @TheSkullTrumpet @Alicoh1 Member of species who has constantly been wrong before confident that they cannot be wrong ever again
English
@AMedviediev @ChitandaOSRS @fear_nd_loathin @KadyMuhammad1 @TheSkullTrumpet @Alicoh1 Like I said, severe D-K.
English
@kilianhekhuis @ChitandaOSRS @fear_nd_loathin @KadyMuhammad1 @TheSkullTrumpet @Alicoh1 Sure, and you know everything there is to know about the universe. This is fundamentally different than the leading physicists of the 19th century declaring that there is nothing left to discover in physics. NOW WE FOR SURE FOR SURE KNOW EVERYTHING
English
@AMedviediev @ChitandaOSRS @fear_nd_loathin @KadyMuhammad1 @TheSkullTrumpet @Alicoh1 Look, there's no way your advanced Dunning-Kruger is letting anyone get through to your skull with arguments, but we know infinitely more today than any of your old examples. In order for FTL to be possible, about 90% of our fundamental understanding of the universe must be wrong
English
@ChitandaOSRS @fear_nd_loathin @KadyMuhammad1 @TheSkullTrumpet @Alicoh1 Alright, different example - until ~1897 it was considered fundamentally impossible to split the atom
English
@fear_nd_loathin @KadyMuhammad1 @AMedviediev @TheSkullTrumpet @Alicoh1 The difference is that scientists never believed these things were physically incompatible with the laws of the universe, they believed them to be beyond human engineering. These are two very different types of limits, most notably because one of them can be surpassed.
English
Bro, we can't even reconcile GR and quantum mechanics, but suuuuuure, that's just a minor hiccup and otherwise our knowledge of physics is full and correct. Totally different than when Lord Kelvin said 'there is nothing new to be discovered in physics' right before Einstein blew the whole thing up.
English
@TheSkullTrumpet @Alicoh1 In what way? The current understanding of physics in the 19th century confidently assumed different things were possible than the current understanding of physics today. Assuming that THIS TIME we definitely 100% know everything correctly is mega arrogant.
English
@mauricio_460 @Alicoh1 All I'm saying is that "the current understanding of physics" evolves all the time
English
@AMedviediev @Alicoh1 Lorentz transforms, which form the foundations of Einstein's theory of relativity, imply the impossibility of faster than light traveling.
English