Some key takeaways: 1. Anton’s team discovered a breakthrough in AI “unsticking itself”—allowing the AI to keep improving and fixing its own bugs. 2. The biggest bottleneck in product development is shifting—from “Who can build it?” to “Who knows what to build?” 3. Community
GENERATIVE AI
-
Lovable reaches $10M ARR in 60 days with AI-powered app builder
By
–
Building Lovable: $10M ARR in 60 days with 15 people
— Lenny Rachitsky (@lennysan) 9 mars 2025
Anton Osika (@antonosika) is the co-founder and CEO of @lovable_dev, an AI-powered app-building tool that turns a description into working products without requiring any coding knowledge. They are building what they describe… pic.twitter.com/L2fCbisqZwBuilding Lovable: $10M ARR in 60 days with 15 people Anton Osika (
@antonosika
) is the co-founder and CEO of @lovable
, an AI-powered app-building tool that turns a description into working products without requiring any coding knowledge. They are building what they describe -
Google to Release New Gemini Models on March 12
By
–
BREAKING : Google will release new Gemini models on March 12 (new model options to be available in the model selector). There are 2 potential candidates so far: – Flash 2.0 Thinking models (non experimental)
– Recently discovered “Personalization Experimental” -
Covergence AI vs ChatGPT Tasks for research summaries
By
–
Covergence AI (Free Operator agent) sends you such beautiful email summaries upon the research completion. This is what ChatGPT Tasks should have been doing – such a simple change but a huge difference. Does anybody use ChatGPT Tasks?
-
Chess app created in minutes thanks to AI, challenge the master
By
–
Cette IA est juste folle… j’ai créé une app d’échecs en quelques minutes sans toucher à mon ordi 🤯😳
— Ludo Salenne (@LudoSLN) 9 mars 2025
Je peaufine le truc et je défi le maître @Blitz_Stream direct dès que c’est OP !
@ManusAI_HQ is Wonderful 👏
Regardez le job 👇 pic.twitter.com/jIIUEyZwlDThis AI is just crazy… I created a chess app in minutes without touching my computer. I'm polishing it and I challenge the master @Blitz_Stream directly as soon as it's OP! @manusai is Wonderful Look at the job
-
Manus AI: A New Powerful General AI Agent for Task Automation
By
–
I think OpenAI should be worried.
— God of Prompt (@godofprompt) 9 mars 2025
China’s Manus AI is here, and it’s on another level. It's a powerful general AI agent that can automate any task.
10 wild use cases:
1. B2B supplier sourcing pic.twitter.com/mVYXjYzXAXI think OpenAI should be worried. China’s Manus AI is here, and it’s on another level. It's a powerful general AI agent that can automate any task. 10 wild use cases: 1. B2B supplier sourcing
-
Training Helpful Harmless Assistant with Reinforcement Learning from Human Feedback
By
–
Training a Helpful and Harmless Assistant with
Reinforcement Learning from Human Feedback slides: https://
docs.google.com/presentation/d
/1hYPWiLETSK5r_Y6sU0mWbOyAz-uevNJh3Xp6BmUnGHM/edit?usp=sharing
… paper: https://
arxiv.org/abs/2204.05862 instructgpt: -
Early RLHF Paper: From Supervised Fine-tuning to Personality in Language Models
By
–
[Slides] This is one of the earliest papers on RLHF (if not the first, alongside InstructGPT). Before RLHF, language models didn’t really have personalities—they mostly relied on supervised fine-tuning or clever prompting to understand humans. Think back to the InstructGPT days.
-
Prompt Engineering Consulting: Rebuilding Internal Tools with Lovable
By
–
doing some prompt engineering consulting devs are in italy chillin' all weekend re-made their internal tooling with @lovable in 30 minutes unblocked