Ethan B. Holland

“AI agents are coming. But who’s in control? @mmitchell_ai, one of the best researchers in AI ethics, makes a critical point about autonomy: fully autonomous systems carry unknowable risks because they operate on computer logic rather than human logic. The solution? Build
https://x.com/fdaudens/status/1879540098863157543

“I’m joining forces with @mikeknoop to start Ndea (@ndeainc), a new AI lab. Our focus: deep learning-guided program synthesis. We’re betting on a different path to build AI capable of true invention, adaptation, and innovation.”
https://x.com/fchollet/status/1879583863368032432

“4/ If you have upcoming interview or a programming exam you can delegate ChatGPT to come up with new puzzles to solve to become better at it in Canvas!”
https://x.com/karinanguyen_/status/1879270535475171773

“One of the largest barriers to a future where everyone has multiple AI co-pilots is trust. Trust will come with agents providing consistently amazing experiences over time –– but getting there will be harder if everyday interactions with AI are poorly built / frustrating.” / X
https://x.com/reidhoffman/status/1878859959234170943

“This is the future of UI. I built an AI chat that transforms itself as you talk. Words shape not just the conversation, but the entire interface and tone. Want a new look? Dark mode? Windows 98? Ask, and watch the UI transform✨ Deployed on @Replit for everyone to try. 👇”
https://x.com/skirano/status/1878865450702139824

“We’re excited to introduce Tasks! For the first time, ChatGPT can manage tasks asynchronously on your behalf—whether it’s a one-time request or an ongoing routine. Here are my favorite use cases: 1/ ChatGPT checks stock price every morning!”
https://x.com/karinanguyen_/status/1879270529066262733

“With our upcoming /extract you can capture information from multiple URLs and even entire websites The future of web extraction is coming ✨ Shoutout to @guilherme_rodz + @firecrawl_dev team for beautiful UI work on the beta preview 👇”
https://x.com/nickscamara_/status/1877796423267156390

“Send this to all your developer friends in NYC. @tryramp @googlecloud @AgentOpsAI @southpkcommons @knocklabs @floxdevelopment Are hosting a hackathon on January 14, 6-9 PM at Ramp HQ in NYC.  Limited spots are available.  Register below.”
https://x.com/AtomSilverman/status/1876645864225354031

“I built a tool that can help open the black box of AIs. It is called ‘Programmed Conceptual Deconstruction’. It systematically breaks down any initial concept (Ci) into its constituent, related concepts through an iterative process until further decomposition is no longer”
https://x.com/gabmfrl/status/1877369479790997949

“2025 is the year of agents. baby steps in that direction:” / X
https://x.com/gdb/status/1879327050819104778

“In the future, it’s agents all the way down.” / X
https://x.com/dharmesh/status/1876013330529501652

“I made our AI hedge fund customizable. You can now select specific agents to run. Choose from: 1 • technical analyst 2 • fundamentals analyst 3 • sentiment analyst 4 • valuation analyst This is made possible by @LangChainAI By default, all agents will run. But if you”
https://x.com/virattt/status/1877497641522835714

“Our AI hedge fund now trades multiple stocks One step closer to a real-world system. Our system has the following: • valuation analyst • technical analyst • sentiment analyst • fundamentals analyst Given some tickers, each analyst uses tools to analyze every ticker. Then,”
https://x.com/virattt/status/1880031667873583556

Building in the Era of Autonomous Software Development
https://backchannel.org/blog/autonomous-software

“We’re excited to announce AI SDR-Kit, a comprehensive suite of app integrations and starter templates to build highly customizable AI sales agents. As a developer, I only realized how tough it can be to crack sales after founding my startup. It’s a lot of grunt work,
https://x.com/KaranVaidya6/status/1877739775916196223

MiniMax-01 series – MiniMax
https://www.minimaxi.com/en/news/minimax-01-series-2

“Job applications without the grind This GitHub repo automates job applications with LLMs for resume personalization, handling API keys, YAML configurations, and OpenAI/Gemini endpoints, saving time by auto-filling forms and generating targeted cover letters. What it offers: →
https://x.com/rohanpaul_ai/status/1879493784284860497

“Quite a brilliant paper. 👏 Search-o1, lets models pause reasoning to search for missing knowledge, just like humans do Search-o1 framework enhances LLMs by integrating agentic search and knowledge refinement during reasoning, enabling dynamic knowledge retrieval while
https://x.com/rohanpaul_ai/status/1880173989143921058

“Launching today: Organize your inbox with AI Our new v3 AI Assistant is fully agentic & can reason about how to manage your inbox — including archiving & deleting junk, identifying important emails, adding labels, and creating todos for action items.
https://x.com/startupandrew/status/1879257093846552641

“🤖 Build an LLM-Powered Data Pipeline Transform your S3 files (text, PDFs, images) into AI-searchable vectors using LangChain and Qdrant. Create intelligent semantic search with LLMs in just 30 minutes! Learn more 👉
https://x.com/LangChainAI/status/1877747452486320610

“LLM and agents are still brittle but many researchers are already experimenting with them in different areas of medicine including mental health care. Here is an example of a new paper that proposes a multi-agent framework, AutoCBT, for Cognitive Behavioral Therapy. The work
https://x.com/omarsar0/status/1880283025595867631

“Your AI Agent startup is doing well when your customers use it and ask you: “and this is all done by code?”” / X
https://x.com/garrytan/status/1877806051187184075

“🔥 Learn how Decagon built their AI Agent Engine 🔥 Join us for an evening meetup in San Francisco hosted at Pear VC, featuring a fireside chat with @hwchase17 (CEO of LangChain) and Bihan Jiang (Product Lead at Decagon). Bihan and Harrison will take you behind the scenes of
https://x.com/LangChainAI/status/1879235381545386309

“🤖 🧠 Build AI Agents with Persistent Memory Create enterprise-ready LangChain React agents that remember conversations across sessions using PostgreSQL and Claude-3-haiku LLM, featuring both Python & Node.js implementations. Learn more:
https://x.com/LangChainAI/status/1880299047178715244

“Yet another huge announcement for CrewAI! AI Agent Week is coming March 2025. The first global summit for production AI: • 3 Cities • 5 Days • 10+ Countries • 1000+ Use Cases • 500 Industry Leaders • Wrap up week with a bigger AI Agent Summit in NYC This changes
https://x.com/joaomdmoura/status/1877321421246050692

“AI Agent which control’s your Browser! Multi-tab handling! Agent task = ” open 3 tabs with elon musk, trump, and steve jobs, then go back to the first and stop. ”
https://x.com/PratikKadam_/status/1876881767535092096

“Agent Laboratory: Using LLM Agents as Research Assistants Enables you to focus on ideation and critical thinking while automating repetitive and time-intensive tasks like coding and documentation
https://x.com/arankomatsuzaki/status/1877195634878353680

“AI agents are able to perform autonomous tasks today They can log into various systems like Salesforce, PayPal and Confluence and perform the tasks humans do Every white collar knowledge worker can deploy a dozen agents TODAY and increase productivity by 50% Your 5-day work” / X
https://x.com/bindureddy/status/1878606861433463240

“Github repo to build, serve, and interact with LangGraph-based agents using a FastAPI service, Streamlit UI, and Docker for streamlined development. Your LangGraph agent in minutes Supports streaming responses, multiple agents, feedback, and content moderation. What it offers:
https://x.com/rohanpaul_ai/status/1879494777026306287

“We just released Codestral 25.01: – Joint #1 on LMSys Copilot Arena while being much faster than other #1 models. – 2x faster than Codestral 22B – Strong FIM performance Try it now for free in VSCode:
https://x.com/dchaplot/status/1878952042498334944

“Multiagent Finetuning Introduces multiagent finetuning, a novel approach for improving language models through self-improvement. Unlike traditional single-agent finetuning methods that often plateau after a few iterations, this approach uses a society of language models derived
https://x.com/omarsar0/status/1878816276312989821

“Interesting attempt to build an AI agent-based research assistant to automate machine learning paper writing by acting as PhDs, post-docs, & professors working in a typical lab. It doesn’t autonomously produce high-level work but it looks promising as a copilot for researchers.
https://x.com/emollick/status/1877377502751985825

HAL: Holistic Agent Leaderboard
https://hal.cs.princeton.edu/

“In a world of AI agents and video models that shatter the Turing test, proof of humanness and identity will become crucial. How soon until your iPhone is constantly using Face ID to make sure it’s actually *you* using your device? The “attention awareness” feature on iPhone is” / X
https://x.com/bilawalsidhu/status/1879320013275853193

“The three areas of need are: 1. Agent Authentication & Authorization 2. Prompt Injection-related attacks 3. Secure Agent Architecture For #1, Agent authentication can likely be put into existing stuff like okta etc. Authorization is the really hard part and will need to be” / X
https://x.com/rez0__/status/1879557690101260681

“📩Open source email assistant agent An email assistant is a great example of this – so we built and open sourced it! It checks emails and triages them. It can either ignore, notify, or respond. The human is very much in the loop – approving all outgoing emails” / X
https://x.com/LangChainAI/status/1879218070008570213

“Things will get INSANE as Computer Use Agents continue to take off…. You can wire up an AI agent to log in as you and perform tasks on different systems. You can also respond to Slack messages and emails from folks. WHY BOTHER WORKING WHEN AN AI AGENT CAN DO IT FOR YOU 😅😂” / X
https://x.com/bindureddy/status/1878983646532825320

AutoGen v0.4: Reimagining the foundation of agentic AI for scale, extensibility, and robustness – Microsoft Research

AutoGen v0.4: Reimagining the foundation of agentic AI for scale, extensibility, and robustness

“Micro Agent: an AI tool that generates unit tests and iterates on code until the tests pass, ensuring reliable function creation in languages like JavaScript, TypeScript, and Python
https://x.com/tom_doerr/status/1875884853935100092

“Think it, Agent Laboratory builds it, a research assistant that never sleeps Agent Laboratory enables researchers to automate tedious research tasks while maintaining control over ideation, reducing research costs by 84% compared to existing methods. —– 🔧 Methods in this
https://x.com/rohanpaul_ai/status/1879126054255956271

“”Build AI agents that think in code. A TypeScript framework for building reliable AI agents – generate execution plans, run them as ASTs, and debug with full code visibility”
https://x.com/tom_doerr/status/1877142409252934112

“Exciting news from @CopilotArena! The latest Codestral 25.01 release is now topping the Copilot Arena leaderboard (joint #1, +12 points over previous Codestral!). Congrats to @MistralAI🎆 Try out the new model today in the @CopilotArena VSCode extension.
https://x.com/lmarena_ai/status/1878872916596806069

AI Agents Are Here. What Now?
https://huggingface.co/blog/ethics-soc-7

““We don’t care about professional coders anymore,” Masad said. – Replit CEO Instead, he says it’s time for non-coders to begin learning how to use AI tools to build software themselves. Replit CEO Amjad Masad revealed that their AI tool “Agent” can create full software
https://x.com/rohanpaul_ai/status/1879838210563666321

“Here is why I think the process for building AI Agents is broken🧵 (save for later)
https://x.com/AtomSilverman/status/1877507585256923450

“Honestly, AI is the best thing ever for curious people. I find 10 new rabbit holes to go down every week and keep building and expanding my map of the world. Here are my 5 daily drivers: I obviously love Perplexity Pro but have been digging Gemini 1.5 Pro’s deep research” / X
https://x.com/bilawalsidhu/status/1878175807060783258

“👑 The King RAGent Meet your new open-source research assistant that combines document processing and web search for comprehensive answers. Powered by LangChain’s RAG patterns, it features PDF handling, Streamlit UI, and a developer-friendly dry run mode for testing. Discover
https://x.com/LangChainAI/status/1876010792115462433

“What’s Agentic RAG? Let’s explore how @huggingface’s variant of Agentic RAG works. Its idea is to transform users’ queries into retrieval-friendly statements, critique and retry retrieval, making the system more autonomous and accurate. So the workflow of Agentic RAG includes:
https://x.com/TheTuringPost/status/1878590804325011727

“Today, we’re excited to announce the general availability of the Contextual AI Platform. This is the first enterprise platform designed for building specialized RAG agents to support expert knowledge work. What is a specialized RAG agent? First, a general-purpose AI agent is one
https://x.com/ContextualAI/status/1879563224192077993

“17/ @LangChainAI launched The King RAGent: an OSS research assistant that combines document processing + web search for comprehensive answers. It features PDF handling, Streamlit UI, and a developer-friendly dry run mode for testing.
https://x.com/AtomSilverman/status/1877844403152200158

“Microsoft just launched Copilot Vision and it’s INSANE It can see what you do on the web and talks with you about it 7 wild examples so far (Don’t miss the 5th one):
https://x.com/CodeByPoonam/status/1876279050249085009

Copilot for all: Introducing Microsoft 365 Copilot Chat | Microsoft 365 Blog

Copilot for all: Introducing Microsoft 365 Copilot Chat

“Build a multi-agent system that researches, writes, critiques and refines a blog post for you with @NVIDIAAI! Content creation is a huge time sink, especially when it involves laborious research against a large knowledge base. This is the kind of task LLMs are great at
https://x.com/llama_index/status/1876486877479649312

Pratik Kadam on X: “AI Agent which control’s your Browser! Multi-tab handling! Agent task = ” open 3 tabs with elon musk, trump, and steve jobs, then go back to the first and stop. ” https://t.co/m5NEgcbwf7″ / X – https://x.com/PratikKadam_/status/1876881767535092096
“Agents Google’s whitepaper covers the basics of llm agents and a quick Langchain implementation

https://x.com/Hesamation/status/1875299361284124938
Mercedes to integrate Google Cloud’s new Automotive AI Agent