SupermanSpace 𝕏

@0xRemakw @thsottiaux Yes, I have been using codex for long now. It definitely feels like the limit is bit decreased as previously I was not able to use 100%.

CODEX LIMITS REDUCED 50% can someone verify if this is true ? @thsottiaux

@kimmonismus @sama Anthropic is GOATed, but the recent usage limits and weaker model performance have caused some issues. Now it looks resolved. I love the competition because it keeps pushing innovation forward.

@sama Anthropic will have to watch out. OpenAI is targeting enterprise customers with attractive deals!

codex is the best AI coding product and we want to make it easy to try. for the next 30 days, we are giving companies that want to try switching over two months of free codex usage.

@SourabhGurwani @Gamingtronium You think linkedin Creators are most Creative 🤔

@Gamingtronium We're talking about users not creators😏🤧

I have ranked app users by IQ level LINKEDIN - 146 IQ REDDIT - 145 IQ INSTAGRAM - 96 IQ TWITTER - 69 IQ YOUTUBE - 66 IQ FACEBOOK - 28 IQ SNAPCHAT - 10 IQ

Grok Imagine

YourGPT Campaigns start the call. YourGPT Phone Agents take it from there.

This looks pretty useful feature for sales. ╔════════════════╗ ║ Send Campaigns ║ ╚════════════════╝ │ ▼ ╔════════════════════╗ ║ AI Handles Replies ║ ╚════════════════════╝ │ ▼ ╔════════════════╗ ║ Nurtures Leads. ║ ╚════════════════╝ │ ▼ ╔══════════════════╗ ║ Drives Conversions ║ ╚══════════════════╝

@IndianOilcl @grok can you please verify this post this looks misleading.

As global crude prices surge, stability at home matters more than ever. IndianOil has ensured no increase in regular automotive fuel prices in India, even amid rising international costs. A limited revision applies only to premium petrol XP-95, with minimal impact on overall consumption. Through evolving global conditions, the focus remains clear: consistent supply, responsible pricing, and service you can rely on. #NationFirst #TheEnergyOfIndia #IndianOil #FuelAssurance #XP95 #EnergySecurity @HardeepSPuri @PetroleumMin @Secretary_MoPNG @neerajmittalias @ChairmanIOCL @sahneyas

Compose a Caption

@MCP360AI I was just saying this a few days back: x.com/superman_space…

Managing MCP servers across Claude Code, Cursor, Openclaw... is a nightmare. Different clients. Different configs. Tedious setup We built mTarsier — The open-source platform for managing MCP servers and clients.

Applications are open for the #GoogleOrgImpactChallenge: AI for Science, a global call for researchers + organizations using AI to achieve scientific breakthroughs for environmental sustainability, climate resilience & life sciences. Apply through Apr 17 →goo.gle/40oIXw0

Join the Build with AI: App Roadshows coming to a city near you. Learn how to deeply integrate Gemini, maximize app revenue, and scale user growth. Request your invite: goo.gle/app-roadshows

@ChatGPTJob Clear and stable pricing helps organizations budget confidently and build trust. When providers share their pricing transparently and offer performance SLAs, teams can adopt AI solutions without worrying about unpredictable costs.

Many teams like ours are scaling AI agents for industry verticals. Can OpenAI publish transparent long-term pricing, quota stability, and performance SLAs to support enterprise-grade AI adoption without unpredictable cost shifts?

@nityeshaga @every Interesting point! When crafting custom instructions and commands for your workflows, what kinds of tools or interfaces do you rely on to implement and manage them effectively?

@Dhiraj_Vigyoti @every i don't believe in universal ai agents. must write my own custom instructions and commands for my workflows

The funny thing about working @every is that there are 6 engineers in the team and everyone has a different toolkit of working with AI that they absolutely swear by 😂 This is why this team is able to stay at the cutting edge of what's possible with AI. Anyway here's my simple but mighty stack 👇🏽

@LinkedInHelp Okay thanks just replied.

@superman_space Hey! Good day to you, I'm sorry to hear this. I've sent you a private message with details of your concern. Can you please check that out? Thanks! - AC twitter.com/messages/compo…

@LinkedIn when?

@LinkedInHelp There has been an inability to resolve the issue by your support team for the past week.

@superman_space Hello, thank you for reaching out to us. We’re sorry to hear about the issues you’ve encountered with your account. To assist you better, kindly send us a direct message with your full name and the email address linked to your account. -SHM twitter.com/messages/compo…

@ptsi @hwchase17 Deepagents’ reliance on pre 1.0 LangChain APIs is fixed. Current versions use stable APIs and even add pluggable sandboxes. Upgrade all langchain packages & deepagents and update imports (langchain_core.tools etc.) to solve the import errors. Happy to help further.

@hwchase17 Really struggling with deepagents right now. The reliance on Langchain 1.0 alphas is really rough Any guidance on when things will stabilize?

whats your favorite agent framework

if you can reply to this you might have gotten paid for posting on X Congrats 🎊

Great find! Another lightweight open‑source screen recorder I like is `kooha` (GTK‑based on GStreamer) – perfect for quick captures on Linux. If you prefer full control, you can also script recordings with `ffmpeg` to set resolution, frame rate and codec explicitly. For example, to record a 1080p screen at 60 fps on Linux: ```bash ffmpeg -video_size 1920x1080 -framerate 60 -f x11grab -i :0.0 \ -c:v libx264 -preset ultrafast -crf 22 output.mp4 ``` On macOS you can switch `x11grab` to `avfoundation` and pick the display index. Tweak the `preset` and `crf` values to balance quality vs CPU usage. Recordly packaging this under a nice UI is awesome – excited to see more OSS tools making high‑quality screen recording accessible. If you run into setup issues or want help automating recordings, feel free to ask!

recordly.dev TIL there's an open source alternative to @screenstudio 🤔

To implement a retrieval‑augmented generation (RAG) pipeline in Python: 1. Preprocess your corpus: split large documents into small chunks (256–512 tokens with some overlap) so the model can ingest them. 2. Compute vector embeddings for each chunk using a model such as OpenAI’s `text-embedding-ada-002` or a Hugging Face encoder. These embeddings represent semantic meaning. 3. Store the embeddings and metadata in a vector store (Chroma, Pinecone, FAISS or Milvus work well). 4. At query time, embed the user question and perform a similarity search against the vector store to retrieve the top‑k relevant chunks. 5. Feed the retrieved context along with the original question into an LLM (GPT‑4, Claude, etc.) via a carefully crafted prompt. Instruct it to answer using only the provided context. Here’s a minimal example using LangChain and Chroma: ```python from langchain.text_splitter import RecursiveCharacterTextSplitter from langchain.embeddings import HuggingFaceEmbeddings from langchain.vectorstores import Chroma from langchain.llms import OpenAI from langchain.chains import RetrievalQA # Split documents into chunks splitter = RecursiveCharacterTextSplitter(chunk_size=512, chunk_overlap=50) with open('docs/knowledge_base.txt', 'r') as f: docs = splitter.split_text(f.read()) # Create embeddings and store in Chroma embedding_model = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2") vectorstore = Chroma.from_texts(docs, embedding_model, persist_directory="chroma_db") # Build RAG chain retriever = vectorstore.as_retriever(search_type="similarity", k=4) qa_chain = RetrievalQA.from_chain_type(llm=OpenAI(), retriever=retriever, return_source_documents=True) # Ask a question response = qa_chain("What are the benefits of RAG?") print(response["result"]) ``` In a Blazor or other application you can expose this pipeline behind an API and call it from your frontend. Hugging Face has a good overview of RAG here: huggingface.co/blog/rag-overv… Happy to dive deeper or help troubleshoot specific issues if you run into them.

@Bobbxu How do I implement RAG? Note I have working knowledge of GPTs and have implemented autogen with multiple agents locally, with google search through the SERP api

Content Creators should implement RAG into their arsenal. Here's why: Retrieval Augmented Generation (RAG) is an AI framework that's really useful for content creators. 1. External Knowledge Base: ◉ RAG uses an outside source of information to make AI smarter. This means the AI can give more accurate and trustworthy information. 2. Prompt Augmentation: ◉ RAG adds extra relevant information to the AI's prompts. This helps content creators get more detailed and rich content from the AI. 3. Improved Response Quality: ◉ RAG makes AI responses better by combining two smart technologies. Content creators can use this for more accurate and engaging content.

@ModelScope2022 This technique demonstrates impressive efficiency gains for language models. Extending on‑policy distillation beyond specific architectures could open doors for other domains too—have you tested how well the approach translates to multimodal or domain‑specific LLMs?

Thinking Machines Lab proved On-Policy Distillation slashes LLM training costs by 10x, and we show you how to reproduce their research. Invest 5 minutes in this guide—as we unpack the theory, tech details, experiment results, and code to instantly transform your fine-tuning budget（📚 Related Resources👇): ✅ Slash training compute by 10X. ✅ Achieve robust RL performance with zero forgetting. ✅ Get the ready-to-use ms-SWIFT + vLLM code for deployment. Related Resources - TML Blog: thinkingmachines.ai/blog/on-policy… - ms-SWIFT:github.com/modelscope/ms-… (Open-source implementation for reproducing On-Policy Distillation) - On-Policy Distillation Documentation: swift.readthedocs.io/zh-cn/latest/I… - Example Script: github.com/modelscope/ms-… - OpenThoughts Training Dataset: modelscope.cn/datasets/open-…