NPCSlayer

@MaxHirsch13 It sucks i tried it, crappy outputs vs just dealing with the llms directly. Ive had better results with flora ai

I don’t think I’ve met one person that’s had a good experience with Higgsfield other than a paid affiliate Correct me if I’m wrong

Time for level 2 of marketing (Meta Ads) I’m scared

@psomkarbuilds Declining and Aging population 🤷🏻‍♂️

why is Japan so behind in AI they gave us robots bullet trains PlayStation and yet barely show up in the AI race what actually happened ?

@i_am_jeganraj @Polymarket @grok @grok is it the caffeine that is doing the most here or what is it?

@Polymarket @grok verify the fact? And which coffee they are talking about ?

JUST IN: American Heart Association reveals drinking up to five cups of coffee per day may lower the risk of stroke, heart failure & type 2 diabetes.

@_frederickjames adapty.io/finance/?rsid=…

The Great Lock in Begins! 🇳🇱 My app is almost $6k/m, but: • TikTok paused my account. • Running out of $$$ for ad spend. • Apple payouts take a few months! Yesterday, I made ~$150 profit. I'm testing a new marketing meta. Not seen anyone speak about it.

@alexwg I read the whole thing in Dr. Alex Wissner’s intelligent voice from moonshots 😆! Love these weekly updates!

@ChadAppDev Static images on meta are pumping right now, easiest onces to scale in volume, specially with andromeda algo

Has anyone had success with static images on Meta Ads or do we have to dance?

@xburak @grok what unsaturated health niches is he talking about?

everyone's building AI apps but the boring truth in the US Apple Ads data: health & fitness converts at 7.80%, the best rate of any category. people pay to change, not to be entertained. Music category at 3.8% prove the other half. so, you may think building a fitness app is really hard, and you're right. But there are still tons of unsaturated health niches available to enter.

@adhamuxi 10 😵‍💫😫😫

How many projects are you currently building? -1 -2 -3 More

@adhamuxi Both

Builders, be honest: Are you solving a real problem or just shipping a fun side quest?

@TheSpacerr Video gen and visual analysis

Claude is good for Frontend ChatGPT is good for Backend Gemini is good for?

@13_narcissus Build apps with ai

"Learn to code" was the advice of the 2010s. What's the equivalent advice for the 2030s?

@BraedendotTECH Ai still needs to be triggered, ai itself doesn’t have desire. The ai could have the potential to make billions but itself doesn’t care the way we do . @grok can you go deeper into this

@MilliardVenture If AI eventually becomes better at formulating ideas what role is left for us? Our species won’t be needed in the future 😭 And we’re creating the thing that will make us irrelevant

At some point, we become the bottleneck. not the AI. The models will know how to do the task better than we know how to tell them what we actually want. Our ability to direct them becomes the limiting factor. But once model progress starts to plateau, the only variable left to improve is us. The models become the ceiling, and the only way to improve output is to get better at using them or remove humans from the loop. Scary part about this is: if AI is optimized to improve output without end, at what point does it conclude that humans are the inefficiency?

@BraedendotTECH True! I think evolving the harness unlocks real potential

at some point a better model stops making you more productive. you hit a level of "coding skill" where basically everything you'd actually build is already doable, and the next model adds nothing. the benchmarks go up every few weeks. your actual output doesn't move. a smarter model, more parameters, a bigger context window, and you're still shipping the same things at the same pace, because the bottleneck was never the model. i think we're closer to that ceiling than anyone wants to admit. everyone treats the next release like it's going to unlock something, then it drops and your week looks identical. I can’t be the only one noticing that whatever new model comes out (Kimi, Fable, GPT 5.6 etc…) even if it’s 10x better than the next best, it doesn’t change anything on my actual productivity. once capability is just "enough," being smarter stops being the thing that sets one model apart. the whole race has to move somewhere else. so what happens to all of us the day more parameters stop buying anything?

@tykng_ just amplifies how lonely you really are that you dont have anyone to go to the movies with...

Going to the movies by yourself is literally a humiliation ritual

@TuanVy Dam bro its irking me already and hes not even my friend!

How would you feel? A good friend of mine got married, and I gave him a discontinued white gold 41mm Presidential Rolex from my personal collection as a wedding gift. Not even 2 months later, I bumped into him and his wife at a restaurant while I was being seated. I said, “Bro, you ain’t rockin’ the Rollie!?” He’s like, “Naw man, sold it,” with a blank face. I’m like, “Cool, all good,” smiled, and walked to my table. What’s crazy is, he’s not broke. And he knows it was a personal watch with sentimental value to me. I couldn’t help but feel some type of way after that. He hasn’t hit me up since, and I haven’t reached out either. Am I in the wrong or?

@furkan3ayraktar Awesome work! When will we be able to choose our own ai models within the lovable harness?

This was a fun one to work on. You ask Lovable to enable agent integrations, it proposes the tools and sets up the auth, you review and publish. From then on, your app works inside Claude and ChatGPT. Apps that agents can't use are starting to feel like websites that don't work on phones.

@AlexandrePesant When will we be able to choose what model the lovable harness uses?

Soon agents will be the primary consumers of software. Lovable makes any app an MCP in one click, OAuth included. Can't wait to show you what's next :)

@matthewcp How i see myself in the past now

Anyone remember early AI when you'd copy paste code out of a chat app.