a rabbit using a laptop. a sticker on the laptop reads “Agents”
“Introducing Spellcaster: AI Grammarly, for documentation proof reading Spellcaster is an open-source CLI tool that uses LLM agents to improve your codebase’s documentation. It scans for: * Grammar errors * Spelling mistakes * Bad code examples Here’s how it works (🧵):
“Game changer for scraping. This GitHub repo lets you easily scrape web pages and have the output in LLM-friendly formats (JSON, cleaned HTML, markdown). Features • Supports crawling multiple URLs simultaneously • Extracts and returns all media tags (Images, Audio, and Video)
The Most Capable Open Source AI Model Yet Could Supercharge AI Agents | WIRED
“🎉🎉 CrewAI version v0.63.6 is out! 🎉🎉 🤖 New LLM class to interact with LLMs 🧠 Support to custom memory interfaces 🦾 Adding support to o1 family models 🎤 Updated logs format 📃 Docs updated 🪲 Bugs fixed 💇♀️ Oh and small logo update 🙂 RT please? 🙏⚡️ we move fast!” / X
“Colossal-AI now makes it possible to reduce model training costs by 30% with one single line of code! Achieved by upgrading mixed precision training that now supports BF16 (O2) + FP8 (O1).” / X
“5 Predictions on AI Agents (summary slide from my speeches) 1) AI agents bring the internet to humans: 1) AI agents centralize information, handling internet tasks and purchases. 2) Content delivered in a multi-modal format humans choose. 2) Media and revenue models shift
AI agents invade observability: snake oil or the future of SRE?
“📰😅I’ve wasted hours scrolling through Hacker News for good posts. Now it takes me minutes. I built an AI agent with @composiohq x @crewAIInc 1⃣ Looks at your Twitter Personality 2⃣ Suggests 5 must read HN posts. here’s the link:
“YC W22’s @spinach_ai automates meeting workflows using AI. Their new AI meeting agent joins your Zoom and helps run the meeting. Congrats on the launch, @talmixed, @joshwillis2000, and @GrossmanYoav!
[2409.12147v1] MAgICoRe: Multi-Agent, Iterative, Coarse-to-Fine Refinement for Reasoning
“AutoAgents: Autonomously generate LLM agents for any goal! 🤖 This tries to solve the need for strong prompting and role definition by autogenerating agents The code is sparsely documented but readable:
Language Agents: From Reasoning to Acting – Latent Space
Five AI Agent Predictions | Jeremiah Owyang
“AI Agent Terms to know: I’ve interviewed 3 dozen agent experts/founders/VCs: -Primary Agent: The main AI interface handling user requests. (expect big tech to each offer their version) -Meta Agent: Oversees and coordinates other agents to optimize tasks. (aka agent” / X
“We built a startup from scratch with AI Agents! By giving agents access to Slack and custom tools using @replicate, @cohere and @v0, our simulation did market research, created branding, posted a job on LinkedIn and built a full-stack app! Read below to learn how it works.
“Too many of us are under the illusion that AI will only work for big corporations. I’m commited to proving that AI Agents will work for YOU. Never before in history has it been so viable for an individual with no resources to compete against big business as it is today.” / X
“You can now upload audio files to otto and have an agent analyze & extract key details from it
[2409.12089v2] The Impact of Element Ordering on LM Agent Performance
“the “end state” of most ai apps probably wont be a platform its a gonna be an ai agent talking with you on slack doing the work of 10 employees at once and then when its done, it sends you the finished project” / X
“GitHub Copilot is AI-native, right where your code is. Copilot on
“We are currently hiring for an open research scientist position working on my team in multi-agent artificial general intelligence:
“A helpful autonomous agent has access to all the tools you need – and nothing more. But we all have different needs, changing all the time, so… what we need is a self-building autonomous agent. Here, I’ll describe the 3 levels of self-building autonomous agents:
“For years now, we have iterated and improved on a new business model combining the best capabilities of software and human talent, to provide abundant and high quality data for AI. The team’s hard work has been paying off; we hit nearly $1B ARR far earlier than expected this” / X
“Fantastic news! Congrats @NegarEmpr, @julia_kiseleva, @Chi_Wang_ and all! It is such a pleasure working together with you on this work. Evaluation is such an important topic around #AIAgent and has been an important focus of #AutoGen. Hopefully this work can inspire more ideas” / X
“Build a multi-agent app without writing a single line of code 🔥 This app by @MarcusSchiesser is sick – probably the easiest way I’ve seen anyone building multi-agents. RAGApp:
“Five AI Agent Predictions: Here’s what I think will happen, the fifth requires an open mindset.
“MotleyCrew: a pragmatic approach to AI agents AI agents are all the rage these days. An AI agent is simply a wrapper around a Large Language Model (LLM) that allows it to request actions, such as a web search, from the outside world, and feed the results back to the LLM, and so
“Menlo VC just dropped a deep dive on AI agents Some insightful images: 1. A nice market map x-axis = vertical, horizontal y-axis = more or less LLM autonomy
“Gemini’s Real Superpower – It’s 10x Cheaper Than o1! The new Gemini is live on ChatLLM teams if you want to play with it. Predictably, the new version is better than the old version, and Gemini now trails behind o1, Sonnet, and gpt-4o but is significantly better than
“Exclusive: Google just released two upgraded Gemini 1.5 models, 1.5-Pro-002 and 1.5-Flash-002 — achieving new, state-of-the-art performance across math. I sat down with Logan Kilpatrick (@OfficialLoganK) to discuss the models, AI agents, AGI, and more. Timestamps: 00:00 Intro
“🕴️🖥️ AI Powerpoint Assistant: Automate your Presentations using AI agents 🎮 No more sleepless nights drafting those decks!!! In just 50 lines of code, the agent: 1⃣ Reads and analyzes the Google Sheets data 2⃣ Creates awesome visualizations and tables 3⃣ Drafts a Presentation
“Exclusive: Google just released two upgraded Gemini 1.5 models, 1.5-Pro-002 and 1.5-Flash-002 — achieving new, state-of-the-art performance across math. I sat down with Logan Kilpatrick (@OfficialLoganK) to discuss the models, AI agents, AGI, and more. Timestamps: 00:00 Intro
“Just launched a new AI sales tool for Google Sheets 🚀 We got frustrated by growing complexity of almost every sales tool we use. So we built a simple one, using AI in Google Sheets. Get prospecting in seconds. Start by entering target company URLs. An AI research agent
“If you can describe it, Agentforce can do it. Watch Chief Digital Officer @RudiKhoury build his first AI agent in minutes using product support data from @FisherPaykelUS’s site. Learn how to build yours here:
“Agentforce + Slack = Game Changer! 🚀 Agentforce in Slack brings your CRM-based agent right to where you’re already working, so you can interact with your CRM data without ever leaving Slack.
Meta
“With Llama 3.2 we released our first-ever lightweight Llama models: 1B & 3B. These models empower developers to build personalized, on-device agentic applications with capabilities like summarization, tool use and RAG where data never leaves the device.
“Zuck’s AI strategy in a nutshell: – Free frontier model for devs – Undercut closed-source rivals – Crowdsource best use cases for multimodal AI – Skip cloud wars; instead monetize biz/creator agents – Harvest fresh data/content for Meta ecosystem – Profit Zuck’s XR strategy in
“Today we are launching Microsoft 365 Copilot agents. There’s a lot I could (and will) say about it, but first, this is why I exchange my labor with Microsoft. The power of our company when pulling all together in the same direction is unmatched anywhere. So, what are agents? ⏬
Microsoft updates its AI suite with more agents and Copilots
O1
“We are entering the 3rd phase of LLM Development. 1st phase was early tinkering, Transformer to GPT-3 2nd phase was scaling 3rd phase is an innovation phase: what breakthroughs beyond o1 get us to a new proto-AGI paradigm
“Updates to OpenAI o1 API availability: – We’ve expanded access to developers on tier 4 (100 requests per minute for both models). – We’ve 5x’d rate limits for developers on tier 5 (1000 requests per minute for o1-preview and 5000 for o1-mini). More expansion coming soon.” / X
“O1 preview scores 98% on a planning benchmark, open up new category: the Large Reasoning Model (LRM) The king is dead long live the king” / X
“8). A Preliminary Study of o1 in Medicine – provides a preliminary exploration of the o1-preview model in medical scenarios; shows that o1 surpasses the previous GPT-4 in accuracy by an average of 6.2% and 6.6% across 19 datasets and two newly created complex QA scenarios.” / X
““You can see it questioning common conventions” “It’s spiritual but oddly human in its behavior” And that is why OpenAI won’t let you see the unfiltered chain of thought of o1; not competitive reasons or whatever
“OpenAI staff who worked on the new o1 model say that the AI is “spiritual” and “oddly human” in how it can reason, reflect and question itself
Introducing OpenAI o1 | OpenAI
“A research note describing our evaluation of the planning capabilities of o1 🍓 is now on @arxiv
“Jony Ive is back with his first major tech project since leaving Apple, teaming up with Sam Altman for an AI-powered hardware device. The project already raised private funding and plans to raise up to $1 billion by the end of the year. iPhone meets o1?
“🚀 SEAL Leaderboard Update 🚀 OpenAI’s o1 is dominating SEAL rankings! 🥇 o1-preview is dominating across key categories: – #1 in Agentic Tool Use (Enterprise) – #1 in Instruction Following – #1 in Spanish 👑 o1-mini leads the charge in Coding
“Finally, a nice paper that evaluates whether o1 can plan. Here is a quick summary of the interesting bits: A domain-independent planner (Fast Downward) can solve all instances of Mystery Blocksworld. LLMs struggle, even on small instances. But Large Reasoning Models (LRMs) like
OpenAI
“I used ChatGPT Advanced Voice to improve my sales pitch. Here’s how you can too👇 1. Give your sales pitch and ask for feedback about: • persuasiveness • clarity • value 2. Remind AI to not be an echo chamber, but actually challenge your ideas. 3. Ask follow-up questions
Rabbit
Rabbit’s web-based ‘large action model’ agent arrives on r1 on October 1 | TechCrunch
Replit
“In the two weeks post launch, our AI team has reduced adverse experiences in the Replit Agent by over 80%. If you’ve been holding out to give it a shot, now is a great time.
“Using replit agent feels like talking to a really eager APM. Also now I have a clean GUI for YT-DLP to download YouTube videos, and never need to deal with a sketchy app or website again.
Leave a Reply