Tokaint

I'm actually making the retro @Sidemen fighting game that @Tobjizzle and @ZerkaaHD asked for last week. v1.0 is 93% done😎 were they serious? probably not but too late lol I attached some WIP pics 🔥🔥 Retweet so they can see it. #indiedev #GodotEngine

@emmanuel_2m @hervenivon @emmnoirhomme No offense, it looks like a scam/cheap now. I remember the old days when it had brand identity, that seems to be gone now. I wonder, could Claude not reference that original look or do you find people respond better to this new look for AI related stuff?

We've rebuilt scenario.com from the ground up in 10 days. What used to take months (or $40-50K with an agency) now happens in a fraction of the time & cost. You can parse all our 500+ models across 50+ providers, plus all apps, workflows.... Thanks, Claude Code :)

@henrydaubrez This was alright, which is crazy because that's really impressive for AI progress lol

Junkyard King - Ep. 0 | MAKING OF ⚙️ HERE’S HOW I MADE THIS, STEP BY STEP 🧵👇 1/14

@turtlekiosk Let me put you on fonthero.com

i wish fonts didnt cost $100000000 to use legally so if you use one in an app you're basically forced to use Inter or something on google fonts or nothing at all

Would that make it dumb? If yes, then are dumb people too smart so they can't comprehend simple tasks? lmao

What if AI isn't dumb, it's just too smart and can't comprehend simple tasks?

@eptwts If you try to sleep at home but everyone at home is awake, it's a lot harder. If you try to diet but your fridge has junk food it's a lot harder. If you try to work but your home office has distractions it's alot harder. Otherwise you don't have any of those problems.

does anyone else not understand coworking at all? what is the benefit in a bunch of guys heading to a coffee shop, whipping out their laptops & working on their own individual things?

Should I post more of these?

@tomislav_rupic @flowersslop You can use an entire album as inspo in Suno? Or do you have to merge it together into 1 file first?

@flowersslop that's exactly how I use Suno: I upload my own music as an album and use it as +inspo. I generate probably 100 tracks before picking one that sounds exactly as if I made it ✌️

Current music gen tools still have the wrong interface. I don’t want to prompt lyrics or melodies. I want to feed it my favorite songs so it learns my taste, then give a vague vibe prompt. It generates 10 songs, I pick the 2 closest to what I want, it generates 10 more from those, and we iterate. So instead of explicit prompting more like steering the search through latent space using preferences

@Framer_X Sora 2 is the only one in this example the looks consistently like anime

Seedance 2.0 vs Sora 2 ⚔️ I tested the same prompt with Ghibli aesthetics. The results speak for themselves. Quick breakdown below 👇

@robotiguy @plasma_node Does it have a game engine and creator monetization?

I actually think you're spot on: it is possible for Roblox games to run natively in the browser one day (WebAssembly + WebGPU make the technical path real). But pulling it off at Roblox scale with millions of UGC experiences, complex scripting, physics, multiplayer, anti-cheat, avatars, and monetization all performing smoothly would likely involve many years of work and a genuine change of focus for the company. Even in this AI world where tools can accelerate coding and prototyping, the hard parts (performance under browser limits, security sandboxing, asset streaming, cross-device consistency, and making user-generated content feel buttery on the web) still demand deep architectural decisions, relentless optimization, and massive compatibility testing that can't be shortcut. This is exactly the challenge we've been working on at Jamango! building a true browser-native UGC platform from the ground up so creators and players can build, publish, and play instantly with zero downloads. Pudgy World shows the momentum and what’s possible; excited to see where it all heads.

Call me crazy but I predict one day Roblox games will run in the browser

@SmashJT I'm sorry what? Inflation... Means... The money they are getting... Is worth less... It doesn't matter if the product cost them nothing to make. If you're selling Coca-Cola, you can keep it at the same price, but your company can't afford what they usually could anymore.

Fortnite raising the price of V-Bucks might be the most blatant scam in modern gaming. In-game currency tied to digital goods should NEVER be subject to “inflation” - These items cost LITERALLY nothing to REproduce. There is no scarcity, no supply chain, no materials. It’s pure corporate greed. Honestly one of the clearest examples we’ve ever seen.

@ChironTheMage I did this with spelunky 2 for free 😭

If your answer isn't a Roguelike, Esport, or MMORPG, you'd probably fail. You need to be able to play a single game for Three Thousand Two Hundred Eighty Five hours (3285) to get the $5M.

@dangroshev So weird, I was looking for something like this for a really niche app I vibe coded for removing models from my game assets. Specifically picking multiple good colors from an image, will try this tomorrow

You want to pick up a colour that represents an image. Say, the back of a flashcard. You want the colour to look good. So you grab a library or vibe code something. It will likely just resize the image to 1x1. The colour is depressing. Can we do better? Yes! With Okmain 👇

@tunahorse21 Funny cause I do think grok is ass, but it's the best closed source AI model we have access to for some niche use cases like 2D animation. Everything else sucks, and seedance is being stingy.

elon paying billions for essentially a 32b diffusion Chinese clone model is peak cinema

@buffalowlow @blakeir @KaiCenat They don't, X for example wasn't rebranded because it was cleaner, it was rebranded for an actual strategic reason. New owner, new direction, aligns with old brand vision being merged, flows better with other companies. Chatgpt is already widely accepted, no point to change it

@Tokaint @blakeir @KaiCenat i think that’s literally what they do

If I ran marketing at OpenAI I’d do a big campaign with @KaiCenat. The whole campaign would just be: “Chat, are we cooked?” “Chat, be honest…” “Chat, is this real?” Then the camera pans…it’s the ChatGPT app. Fully lean into rebranding ChatGPT to Chat.

@DragonbornZane @Affinity @dartfawx Same thing happened to me twice 😭 v1 bought 1 month before v2 came out then had to buy v2 and only had it for a short time before this shit came out. And it's horrible it always crashes and gets no updates. Rip affinity

@Affinity @dartfawx Only a few months after I pay for a lifetime, yall abandon that release and go free. Good luck getting me to subscribe for the full experience. Could have at least given us it for a few months.

Hey @Affinity, how long will the V2 licence servers remain online, and will the V2 apps continue to receive security updates?

@austingriffith Well, I heard the 4b version is as good as Gemini 2.5 which was a great model. Vibey, satisfying. If they now have that power locally it's a dream. I don't think they care about how good sonnet is. Their priority is open source. If you don't care about that, enjoy sonnet.

⚙️ Are the "local ai" guys basically just larping? 🐏 I have a machine with 128GB of RAM 🤖 I downloaded qwen3.5:122b 🤣 It takes 3 minutes to incorrectly do what sonnet 4.6 can do in seconds 😅 Am I doing it wrong? Do I need more RAM?

@melvynx They're always saying "introducing a open source LLM that's closed source level" but that's never true and we never get open source llms that outperform the closed source ones. I never believe that bs

All these benchmarks are so much BS, and you need to try them for 5 minutes to see how much they are just trained to do benchmarks 💀

@melvynx Ouu can you add these 6 features I needed from handy? -pause button -stop button -window display selection -recording length time -transcribing percent -copy last transcript shortcut

I forked handy.computer and in 1 hour with Claude, I added the 2-3 missing features and created an open-source free fork called "Parler" that has simple and amazing features for free, at the level of WhisperFlow. Available here: github.com/Melvynx/Parler It's so simple to create perfect software with open source now.