Berg

@Gltchdctr Senior is realizing code is rarely the constraint 😂

@thbrgo To me it's are you broken in to corporate It enough yet? 😂

Junior to senior used to mean better code. Now it means better prompts. We’re not just coding anymore. We’re thinking in strategy.

OpenAI price cuts are being discussed. Not surprising. Token costs are squeezing companies. Shift toward cheaper models like DeepSeek feels inevitable. GPT-5.5 + Codex hype is there but enterprise adoption is still weak. Anthropic is ahead here.

Most founders are still pricing AI-native products like traditional SaaS. That's a mistake. If demand scales, pricing should too. Legacy SaaS pricing can cripple an AI-native business.

Mocking has always felt strange to me. Why am I writing my own mock data and mock functions just to assert outcomes ? Shouldn’t the systems I interact with already be mocked ? It feels like I’m testing myself.

This is less about regulation slowdown, more about infrastructure standardization for AI. obernolte.house.gov/media/press-re…

Right now, Opus feels like Fable's student.

Sometimes I feel like I’m married to my agent. We argue for hours but agree in seconds.

I am officially dusting off that old mobile project sitting on my hard drive and porting it to WebGL to try my luck on Yandex Games and YouTube Playables

I thought the project was done. Then I realized I hadn’t written docs → no marketing → no discovery. Nothing is ever really finished. We just hit ship and hope for the best.

@guileful_reed Different stack, same suffering 😭

@thbrgo I’m the same when it comes to copywriting 😅 agonizing over every sentence

Working on frontend margins brings out my obsessive side. Top, 2px left, bottom...

Repeat, repeat, repeat…..

Models look similar on easy tasks. As problems get harder, the gap widens. GPT-5.5: more compute → same performance (~5%) Fable 5: more compute → big gains (11% → 31%) Some models can buy more thinking time. Others just get more expensive. That’s where Fable 5 stands out.

Reply and I’ll follow you. I want to connect with ambitious people building something great.

@paulg That’s not the full signal imo. what matters is gross margin after inference costs + whether demand still scales with compute.

I talked to a founder of an AI startup generating about a 40% annual return on the cost of the GPUs he was using. I.e. he could make $400 in annual revenue for every $1000 worth of hardware he used.

@audrlo so basically an expensive way to realize you already had food at home 😅

Startup idea: Smart refrigerator that uses AI to order groceries when you run out

@eden_ true, but also most people only learn how to use AI after they start using it 😅

Using AI is useless if you don’t know how to use AI

@IricMidel vibe coding didn’t kill ideas, it killed excuses 😅

Vibe coding made us realize most idea guys have ideas that suck.

@Official_Adrop we connected...