You can now create realistic videos of AI characters in seconds. They will narrate any script you provide. Here’s how: https://t.co/6kQcA03cSH
— Alvaro Cintas (@dr_cintas) Jul 31, 2024
Three basic rules https://t.co/SPBEqlh9G0
— Shane Parrish (@ShaneAParrish) Jul 31, 2024
$16,047/mo That's how much you can make converting JPG to PNG. - Boring business - One feature - Bootstrappable
— Robin 🚁 (@robin_nomicly) Jul 31, 2024
1/n When ChatGPT Fails: Building Robust Planning Systems with LLMs and Agentic AI Reasoning Large Language Models (LLMs) have captivated the world with their impressive linguistic abilities, leading some to believe they possess the capacity for complex reasoning and planning.… https://t.co/rY4nUAqrr5 https://t.co/WSJxGgabGg
— Carlos E. Perez (@IntuitMachine) Jul 30, 2024
Jensen Huang says the next wave of AI is physical AI, which will require three computer systems to implement: AI, robot and Omniverse https://t.co/q91sQr6Fyx https://t.co/hVTNAWUqts
— Tsarathustra (@tsarnick) Jul 29, 2024
GraphRAG in LlamaIndex! GraphRAG is a technique from Microsoft research for richly understanding text datasets by combining text extraction, network analysis, LLM prompting and summarization into a single system. The process involves two main steps: ➡️ generating a graph that… https://t.co/ZnDtJ731hl https://t.co/mx54Su1gYk
— LlamaIndex 🦙 (@llama_index) Jul 29, 2024
Here is a quick comparison for image-to-video between (1) Luma Labs Dream Machine @LumaLabsAI, (2) @runwayml's new image-to-video Gen-3, and (3) @Kling_ai image-to-video. I used pretty simple prompts for all of them. Let me know what you think. I'll continue to do more… https://t.co/U0aFZLQRnG https://t.co/O28vmrY5S2
— David M. Comfort (@DavidmComfort) Jul 29, 2024
1_ I have been working on a workflow to incorporate @hedra_labs with #liveportrait , this way you can bypass the 1:1 aspect ratio from Hedra output but most importantly its a workaround for those like me who don't really have an affinity for acting LOL. This is can be an… https://t.co/QkAeSXc8VP https://t.co/qp19VUcDLb
— the ALTERVERSE (@Alterverse_AI) Jul 29, 2024
As agentic applications become more popular, we are seeing longer traces getting sent to LangSmith. Oftentimes, you want to search for a particular event within a single long trace. Our new feature enable exactly this, allowing you to filter for runs within the trace view.… https://t.co/8PlV5ZqEC0 https://t.co/vS3EfQkA7H
— LangChain (@LangChainAI) Jul 29, 2024
Good read on Apple's foundational models. Like may things Apple does, they are in a position to do this in a unique, very Apple, way. Which allows them a lot of differentiation. https://t.co/ZaAXW8OsOV
— Ben Bajarin (@BenBajarin) Jul 29, 2024
I updated prices for my stock market API. Each of these endpoints is now $0.04: • GET income statements • GET balance sheets • GET cash flow statements For power users, this is ~$400 / month. Here’s the math on monthly cost: • say you care about 10,000 tickers • each… https://t.co/q9lJzxXF2B https://t.co/mQv2lxnanL
— virat (@virattt) Jul 27, 2024
After a few days of testing, I can safely say that the Llama 3.1 405B is the best model when it comes to prompting a personality and staying in character. I had a 2 hour long chat with "Richard Feynman" that put a big smile on my face. Here's the prompt I used 👨🏻🏫
— Pietro Schirano (@skirano) Jul 28, 2024
Recommended Reading: Superconvergence: How the Genetics, Biotech, and AI Revolutions Will Transform our Lives, Work, and World https://t.co/hT7NOmjcUY
— Vinod Khosla (@vkhosla) Jul 28, 2024
eventos celestes https://t.co/ADc26OA97s
— ☁ (@canekzapata) Jul 27, 2024
📢Interested in fully interpretably machine learning using Large Language Models? Come chat with @TimZXiao and me about "Verbalized Machine Learning" at the ICL workshop (Saturday) #ICML2024! ⏲️Poster session: 10:25 - 11:15 AM & 3:20 - 4:15 PM 📍Location: Lehar 4 📰Paper:… https://t.co/brZfnZ5k95 https://t.co/8OwgtaeH5m
— Weiyang Liu (@Besteuler) Jul 27, 2024
Here’s how the LinkedIn Engineering team reduced the error rate from ~10% to ~0.01% when generating structured outputs with LLMs. Turning natural text into structured outputs is a cool use case for LLMs. But getting LLMs to do this reliably is quite challenging. https://t.co/FOgEWE4K3Y
— Leonie (@helloiamleonie) Jul 28, 2024
Building a platform for generative AI applications https://t.co/zcKebaJ7SU After studying how companies deploy generative AI applications, I noticed many similarities in their platforms. This post outlines these common components, what they do, and implementation… https://t.co/6ym1OE1JR2 https://t.co/M5lik70z1M
— Chip Huyen (@chipro) Jul 25, 2024
1. Ruthless Focus For an early-stage startup, the definition of focus could be acquiring customers and shipping on a constant basis. Working on anything else is pretence, not impactful work. https://t.co/dSkdX7ICLn
— Jonathan Swanson (@swaaanson) Jul 26, 2024
guys..... i just created a website 100% copy and pasting from claude: 1. front end 2. back end - meaning... anyone can sign in / sign up with a google account. - I can assign admin access. - admins can add "lessons" that get updated - people can add comments and i filmed the… https://t.co/FZ4xbkkFSx
— Riley Brown (@rileybrown_ai) Jul 27, 2024
Just finished training Llama 3.1 for @HyperWriteAI It's... better The writing style is just so damn natural https://t.co/3ycrgeynkQ
— Matt Shumer (@mattshumer_) Jul 26, 2024
If you're into AI, this is the coolest website of all AI companies and startups. AI landscape. It's been constantly updated and you can read about each and what each AI tool does. https://t.co/oRlAv26FxJ https://t.co/5X9CtuTsue
— Kris Kashtanova (@icreatelife) Jul 26, 2024
20min talk I gave at the Berkeley AI hackathon a few weeks ago, on how hacking around makes its way to real-world impact in my experience. While True: build and publish projects. Accumulate 10,000 hours. Snowball your work. https://t.co/dril50VAU5
— Andrej Karpathy (@karpathy) Jul 26, 2024
Most interesting results from SELF-DISCOVER: Large Language Models Self-Compose Reasoning Structures: - SELF-DISCOVER improves LLM performance on challenging reasoning tasks by up to 32% compared to Chain of Thought (CoT) prompting. - The method requires 10-40x fewer inference… https://t.co/nbk6ClDQyc
— Tom Dörr (@tom_doerr) Jul 26, 2024
Wagyu is the most expensive meat in the world. This is how it's served at Azabu Note, a restaurant in Japan. [📹 space_tamnik] https://t.co/3qCjLdLnuS
— Massimo (@Rainmaker1973) Jul 25, 2024
6 statistical methods that can be used for A/B Testing (and when to use them). 🧵 https://t.co/qZsTBzIK0m
— 🔥 Matt Dancho (Business Science) 🔥 (@mdancho84) Jul 25, 2024
If you want to grow your startup faster while spending less on SEO: 1. Go to free tool called “Termsuggest .com” 2. Type the most generic keyword of your business 3. Copy suggested keywords 4. Paste these terms into “Searchvolume .io” 5. Filter the keywords with high volume 6.… https://t.co/HCUW0Sa7mk https://t.co/jsM1FpAy5a
— Natia Kurdadze (@natiakourdadze) Jul 25, 2024
Dwarkesh Patel (@dwarkesh_sp) is on a quest to know everything. The most important new tool in his learning arsenal? AI—specifically Claude, Claude Projects, and a few custom tools he’s built to accelerate his workflow. They enhance how he reads, learns, thinks, and conducts… https://t.co/bJs1lG4iuF https://t.co/zi4f9lKEe2
— Dan Shipper 📧 (@danshipper) Jul 24, 2024
Another lovely day in open-source AI. This time it's a 1 trillion tokens multimodal interleaved dataset. It has 3.4 billion images! Includes sources such as PDFs and ArXiv papers. https://t.co/f3j87pZkeK https://t.co/6kGJXvjpwa
— elvis (@omarsar0) Jul 24, 2024
Incredibly impressed with @bardeenai. They just saved me $10k in 10-minutes. I explained a complex workflow to them. In 10-minutes they had a fully runnable and executable agent running the workflow. I was quoted $10k from a developer to build out the automation.
— Adam Silverman (Hiring!) 🖇️ (@AtomSilverman) Jul 24, 2024
Yesterday’s Google’s earnings may have been bit of “meh” at first glance, but there were nuggets in the call that may have important implications for the broader market. My highlights from $GOOG earnings. Full piece below. https://t.co/kJbqT8MOcE
— Mostly Borrowed Ideas (@borrowed_ideas) Jul 24, 2024
@mfishbein - ChatGPT - custom gpts - custom agents - custom ai tools - https://t.co/YGMk1khxbe - and more
— John Rush (@johnrushx) Jul 24, 2024
I laid off entire sales & marketing department of my enterprise SaaS a year ago. My goal was to fully automate their work with AI and switch into organic and inbound traffic. They didn’t believe it was possible and didn’t wanna make it happen. 12 months later we have more… https://t.co/rdTp2RLAFB
— John Rush (@johnrushx) Jul 24, 2024
A playbook on how to stand out in a crowded market explained in 2 minutes: https://t.co/8cq3qQqqvQ
— Hiten Shah (@hnshah) Jul 23, 2024
Fave shot from my two days with Mark Zuckerberg. In exchange for a wake surfing lesson and a wide-ranging interview, I gifted him a gallon of sunscreen.🌞 Full episode: https://t.co/0bL0cWrCLH https://t.co/YcOE0kyefP
— Emily Chang (@emilychangtv) Jul 24, 2024
Meta is using this compositional approach to add multimodal capabilities to Llama 3.1 https://t.co/LDN4aTb3L8
— Tom Dörr (@tom_doerr) Jul 24, 2024
Excited to post my full overview of Llama 3.1 here on X! Includes the key insights and a few interesting tests: https://t.co/rHljFRkgAi
— elvis (@omarsar0) Jul 23, 2024
Want to start a data company? Here's your '24-'25 roadmap https://t.co/zfQ4zbal1R
— Greg Kamradt (@GregKamradt) Jul 23, 2024
7-23-2024 •Netflix Earnings •The Status of Netflix Advertising https://t.co/ybqB2YeBAe
— Stratechery (@stratechery) Jul 23, 2024
Llama 3.1 Now Available in VSCode as a Code Assistant via @GroqInc 🚨 You can now use this new @AIatMeta model directly in VSCode using the @codegptAI extension My first impression: The model is incredible 🚀 https://t.co/PaocpOQ7kl
— Daniel San (@dani_avila7) Jul 23, 2024
My live cohort starts again in 2 weeks. Let me explain what you'll get when you join: 1. 18 hours of live sessions where I'll teach you how to build machine learning systems. (These sessions are not prerecorded. They are live. You can interrupt, ask questions, and have a… https://t.co/lzL18d4gv6 https://t.co/2ar3QKOGpk
— Santiago (@svpino) Jul 23, 2024
The Importance of Being Unimportant in Startups in the AI & Defense age. https://t.co/MXg3GVL6mw
— sam lessin 🏴☠️ (@lessin) Jul 23, 2024
Crashes and Competition The recent Windows crashes are downstream from trying to encourage competition in areas Microsoft should have never made open to begin with, highlighting the challenge for regulators. https://t.co/YOaKoOndyK
— Stratechery (@stratechery) Jul 22, 2024
CodeAct is the default OpenDevin agent. Are there better ones? https://t.co/Yj3Hu4KNTy
— Tom Dörr (@tom_doerr) Jul 22, 2024
Wow... this AI SaaS makes $1M PER MONTH: – Problem: Creating a website is hard/scary – Solution: Generate your website in minutes by typing one sentence. I tried it out for Starter Story. Not bad! https://t.co/MFl8rewGMA
— Pat Walls (@thepatwalls) Jul 22, 2024
Train Loras in ComfyUI #ComfyUI https://t.co/DdhH2I33yJ https://t.co/AIqZexA2cr
— Machine Delusions (@Machinedelusion) Jul 22, 2024
Apple presents LazyLLM, which introduces a novel dynamic token pruning method for efficient long-context LLM inference. It can accelerate the prefilling stage of a Llama 2 7B model by 2.34x and maintain high accuracy. Idea: It selectively computes the KV for tokens that are… https://t.co/U4qzNGpvA6 https://t.co/hCWljqGMaO
— elvis (@omarsar0) Jul 22, 2024
Levels of AI by OpenAI 1: Chatbots, Al with conversational language 2: Reasoners, human-level problem solving 3: Agents, systems that can take actions 4: Innovators, Al that can aid in invention 5: Organizations, Al that can do the work of an organization https://t.co/QWr21xNCI5
— Smoke-away (@SmokeAwayyy) Jul 11, 2024
Apple joined the race of small models. DCLM-7B Released a few days back is open-source in every respect, including weights, training code, and dataset! 👀 🧠 7B base model, trained on 2.5T tokens from open datasets. 🌐 Primarily English data with a 2048 context window. 📈… https://t.co/b09s4ZtpLT https://t.co/mCkJHpcGQW
— Rohan Paul (@rohanpaul_ai) Jul 20, 2024
$100k monthly revenue in 9 months. First-time founder. Sonnet 3.5 changed everything. https://t.co/lANC1ZSDKK
— Pietro Schirano (@skirano) Jul 20, 2024
This was a crazy week for Agent Launches. Do you know about the 10+ major launches this week? Here is a detailed breakdown of everything you need to know.🧵 https://t.co/IlvvSX235f
— Adam Silverman (Hiring!) 🖇️ (@AtomSilverman) Jul 19, 2024
Apple has entered the game! @Apple just released a 7B open-source LLM, weights, training code, and dataset! 👀 TL;DR: 🧠 7B base model, trained on 2.5T tokens on an open datasets 🌐 Primarily English data and a 2048 context window 📈 Combined DCLM-BASELINE, StarCoder, and… https://t.co/owFZ5KvlAd https://t.co/pMoZV9EvLk
— Philipp Schmid (@_philschmid) Jul 19, 2024
99% reduction in cost in 2 years is crazy https://t.co/PG3j9cSRH3 https://t.co/ppAV38jg1E
— vittorio (@IterIntellectus) Jul 18, 2024
The big story with GPT-4o-mini is its cost. And how quickly the price of intelligence (of a sort) has dropped. From @simonw https://t.co/eylhJWXxV1 https://t.co/wyYTEIh8ju
— Ethan Mollick (@emollick) Jul 18, 2024
The Transformers architecture clearly explained 👇🏻 https://t.co/PXKLGUKZge
— Josep Ferrer (@rfeers) Jul 18, 2024
This is what the future of software development will look like. AI writes 80% of the code. Human devs finish the last 20%. The next wave of models will begin to unlock the potential of these tools. Massive AI codegen wave incoming. https://t.co/VTcQQu9jhs
— Mckay Wrigley (@mckaywrigley) Jul 17, 2024
Mistral releases 12B open LLM! 🤯 Mistral Nemo comes as a base and instruct version with a 128k Context window and is multilingual in 9 Language with a new tokenizer. 👀 TL;DR: 🧠 12B Base and Instruct drop-in-replacement for Mistral 7B 🪟 Supports 128k context window and new… https://t.co/BG2Yo79Hik https://t.co/emxGlMBahV
— Philipp Schmid (@_philschmid) Jul 18, 2024
Building a multi-agent customer support bot, from scratch Any LLM-powered customer support use case is a great fit for a multi-agent architecture. You want to simultaneously have a constrained guided flow through the support experience, but at the same time dynamically route to… https://t.co/Yzp8vlNWx0 https://t.co/Y5UzsL6fSr https://t.co/7Vulr1L3oU
— Jerry Liu (@jerryjliu0) Jul 18, 2024
Multi-agent system in ComfyUI https://t.co/cR609otTp0
— Tom Dörr (@tom_doerr) Jul 17, 2024
I don't know why nobody tells you this, but you can easily create your own Chrome extensions. First, the boilerplate: https://t.co/BlzhqMiY8x Then copy the code into Claude and go nuts: https://t.co/RdBIarXoSH Your entire browser-- every site, everything you see, all of it--… https://t.co/376SBjJGxP
— M̵̞̗̝̼̅̏̎͝Ȯ̴̝̻̊̃̋̀Õ̷̼͋N̸̿ (unity/🇺🇸) (@spatialweeb) Jul 17, 2024
15 websites you've never heard of that make over $1M/year: https://t.co/RB0yYnRysw
— Pat Walls (@thepatwalls) Jul 17, 2024
Generative AI is paving the way for a shift to skills-based hiring. This evolution in recruitment focuses on capabilities over credentials, promising a more inclusive workforce. #AI #Hiring #SkillsBasedHiring #FutureOfWork https://t.co/i6xgmAiBXc
— Bryan Talebi (@BryanTalebi3D) Jul 15, 2024
Warren Buffett's secret to success? Look for painters. https://t.co/3C23xqeS9t
— Kpaxs (@Kpaxs) Jul 17, 2024
Been using all the AI-coding tools I can find to help build charts for https://t.co/2NmPYRSErj. I recommend Supermaven (really fast) and Cursor (code RAG is good), but what's crazy is copying code in and out (& in and out) of https://t.co/jCtd6oQVET still beats everything. Why?
— Hrishi (@hrishioa) Jul 17, 2024
These autonomous drones sit on power lines to recharge, allowing them to fly indefinitely - no humans required. Xiaomi's new factories are 100% autonomous and self-improving - and there are no humans working there anymore. Why this matters: Skeptics say "we can just unplug… https://t.co/qdS7NK0YeP https://t.co/W3gHQkeYyr https://t.co/CvAahG7mLg
— AI Notkilleveryoneism Memes ⏸️ (@AISafetyMemes) Jul 17, 2024
Claude Sonnet 3.5 Opensource artifacts UI that you can use with ChatGPT, Gemini, Llama-3 or any other LLMs. https://t.co/K83NkoKZM3
— Unwind AI (@_unwind_ai) Jul 17, 2024
In just a few years, the majority of AI usage might happen on-Device. Excited to see SmolLM, a series of small language models from 135M, 360M, and 1.7B parameters under Apache 2.0, trained for pushing on-device advancements and innovations! 🔥 TL;DR: 🚀 SmolLM models: 135M,… https://t.co/HGfdQjRCZL https://t.co/gkAGo1eHvf
— Philipp Schmid (@_philschmid) Jul 16, 2024
Elon Musk: “A company is a group of people that have gathered together to create a product” “If you think about it, what’s the point of a company existing? The point is that it’s a group of people that have gathered together to create a product. If the product is good, the… https://t.co/9dYfaFMtq9 https://t.co/9imkHRvYSk
— Startup Archive (@StartupArchive_) Jul 17, 2024
AI platform that uses LLM with RAG to transform customer feedback into actionable insights in seconds. Discover critical customer pain points on autopilot without manually going through all customer interactions. And the best part? No coding required. https://t.co/A7j62IZlDs
— Shubham Saboo (@Saboo_Shubham_) Jul 17, 2024
https://t.co/RFA1OChBdL
— Psyche Wizard (@PsycheWizard) Jul 15, 2024
AI-generated computer games become a faster reality than imagined https://t.co/ZbOcS94cy5 https://t.co/mrpk3uWaTE
— Chubby♨️ (@kimmonismus) Jul 16, 2024
Peter Thiel: "Many of the great companies that have been built over the last two decades were founded by people whose identities were somehow deeply connected to their company. It was their life's project. They had some kind of idiosyncratic, different vision of what they were… https://t.co/GsH95NB81T https://t.co/DX9gd9qLkw
— BUILD OR DIE (@BUILD_OR_DIE) Jul 16, 2024
🐒🐒Tribe AI ✨ Low code tool to rapidly build and coordinate multi-agent teams ✨ Lot's of folks have been asking for a low-code builder for multi-agent LangGraph applications This open source project does exactly that! Code: https://t.co/ZeOLPacBWx https://t.co/OEpap0zj7G
— LangChain (@LangChainAI) Jul 15, 2024
Introducing Claude Engineer 2.0, with agents! 🚀 Biggest update yet with the addition of a code editor and code execution agents, and dynamic editing. When editing files (especially large ones), Engineer will direct a coding agent, and the agent will provide changes in batches.… https://t.co/Wn3UiVXLSC https://t.co/iuajZCDTVQ
— Pietro Schirano (@skirano) Jul 15, 2024
The future of document RAG is multimodal RAG 🖼️📑. In our brand-new cookbook we highlight a new multi-modal RAG architecture for processing a slide deck (heavy in text, diagrams, charts, tables) using LlamaParse, @llama_index, and gpt-4o. At the core is a hybrid text/image… https://t.co/KWsVGwT3jD https://t.co/nPiXl2PGCd
— LlamaIndex 🦙 (@llama_index) Jul 15, 2024
A tiny fraction of all investors will 20x their money. Most will lose. Even among those who return, those returns will be driven by a tiny fraction of their investments. What will differentiate the two outcomes is smart investing/having an actual thesis on the trajectory of the… https://t.co/jNJoNTegIo
— Vinod Khosla (@vkhosla) Jul 15, 2024
The Top ML Papers of the Week (July 8 - July 14): - RankRAG - RouteLLM - FlashAttention-3 - Internet of Agents - Learning at Test Time - Mixture of A Million Experts ...
— DAIR.AI (@dair_ai) Jul 14, 2024
Claude Dev, an autonomous software engineer https://t.co/9mKhrCgigu
— Tom Dörr (@tom_doerr) Jul 15, 2024
@HBCoop_ Checkout our AI generated character Eva performing "Finding You" https://t.co/k4Tl2B5vGu
— Studio Magic (@studiomag1c) Jul 15, 2024
Build an AI Movie Production Agent with Claude Sonnet 3.5 in just 30 lines of Python Code (step-by-step instructions):
— Shubham Saboo (@Saboo_Shubham_) Jul 15, 2024
PersonaRAG: Enhancing Retrieval-Augmented Generation Systems with User-Centric Agents Enhances RAG models by incorporating user-centric agents, adapting retrieval and generation based on real-time user data. 📝https://t.co/sbMJXwEfnh 👨🏽💻https://t.co/J1UwjWKkha https://t.co/8e8s2jPaix
— Sumit (@_reachsumit) Jul 15, 2024
The strongest country/empire of every century. [🎞️ sepsi_history] https://t.co/WtkT4PBzOn
— Massimo (@Rainmaker1973) Jul 14, 2024
✅OpenAI's Assistant API is a great example of agentic infrastructure: - Built in persistence - Built in background runs - Easily configurable ❌BUT it doesn't give you full control over the cognitive architecture of your application. Little blog: https://t.co/xj9ouZxurH https://t.co/IzRSWzbABL
— Harrison Chase (@hwchase17) Jul 14, 2024
𝗠𝗼𝘀𝘁 𝗖𝗼𝗺𝗺𝗼𝗻 𝗦𝗼𝗳𝘁𝘄𝗮𝗿𝗲 𝗔𝗿𝗰𝗵𝗶𝘁𝗲𝗰𝘁𝘂𝗿𝗲 𝗦𝘁𝘆𝗹𝗲𝘀 Software architecture styles are the foundational blueprints for constructing various software systems, ensuring they meet specific requirements and quality attributes. An architectural pattern, on the… https://t.co/cddF9fJFnm https://t.co/E8xDXt16v4
— Dr Milan Milanović (@milan_milanovic) Jul 13, 2024
Every AI engineer should learn how to build an advanced RAG flow from scratch in a step-by-step fashion. This is a fantastic guide from @kingzzm showing you how to use @llama_index query pipelines to compose basic to advanced RAG and agent modules together, with full visibility… https://t.co/nK4dFRLQOZ https://t.co/z5WnxXyZgh
— LlamaIndex 🦙 (@llama_index) Jul 14, 2024
Gemini 1.5 Pro with 2 Million context summarizes an hour long video in just 30 seconds. It hugely surpasses GPT-4o, Claude, Llama or other existing LLMs for videos by processing almost 1.37 million tokens in 30 seconds. https://t.co/NEw5ZjBBOA
— Shubham Saboo (@Saboo_Shubham_) Jul 14, 2024
@thepatwalls I feel like there are so many successful SaaS firms that simply re-package existing solutions for a tailored niche. Recently worked with a firm that re-sells a white label website builder (with their own apps+themes) but for gym owners, and they're doing really well.
— Роман (@ruskayapolitika) Jul 8, 2024
$1.3M PER MONTH saas idea you've never heard of: This SaaS helps churches create custom apps so they can collect more money. 37,000 users. $16M ARR. Brilliant. https://t.co/oA4bEiQeX6
— Pat Walls (@thepatwalls) Jul 8, 2024
Build a Llama-3 based RAG pipeline to chat with your local files including citations without writing a single line of Python code (100% free and offline): https://t.co/BsHVGERs7l
— Shubham Saboo (@Saboo_Shubham_) Jul 13, 2024
Five of the top Agent companies launched massive updates this week. Here is a detailed breakdown of everything you need to know. 🧵 https://t.co/RCC3uHXG8x
— Adam Silverman (Hiring!) 🖇️ (@AtomSilverman) Jul 12, 2024
At the @Axios Future of Defense launch event: @demarest_colin: What does the future of the defense industry look like? @RepMcCaul: “What we need are smaller, more innovative, efficient models. Anduril is a good example…they try to anticipate what the defense industry will need… https://t.co/VL6Q6O2STT https://t.co/9u4fir6uhx
— Anduril Industries (@anduriltech) Jul 12, 2024
Plan for this weekend: dive into GraphRAG a bit more. Two new things I want to explore: 👨💻 Implementation of Microsoft's GraphRAG paper from @tb_tomaz: https://t.co/e4yRxzSYih 📃"The GraphRAG Manifesto" - a high detailed piece on GraphRAG from @neo4j: https://t.co/iM8j6GM5Vg https://t.co/RhF4Iw5Ril
— Harrison Chase (@hwchase17) Jul 12, 2024
https://t.co/fnJQkquU9a Is long context >> RAG
— Gopi Vikranth (@GopiVikranth) Jul 12, 2024
“Books allow us to fully explore a topic and immerse ourselves in a deeper way than most media today”, says Mark Zukerberg. 15 books recommended by Mark Zukerberg: 1) The Hard Thing About Hard Things by Ben Horowitz https://t.co/CwiMrbwAUb https://t.co/i3frk3h8jV
— Reads with Ravi (@readswithravi) Jul 11, 2024
Principal Component Analysis (PCA) is the gold standard in dimensionality reduction with uses in business. In 5 minutes, I'll teach you what took me 5 weeks. Let's go! 1. What is PCA?: PCA is a statistical technique used in data analysis, mainly for dimensionality reduction.… https://t.co/BqTdmK3K6R https://t.co/4oLKctc6Kz
— 🔥 Matt Dancho (Business Science) 🔥 (@mdancho84) Jul 11, 2024
Build a Multimodal LLM app with GPT-4o and Opensource Claude Sonnet 3.5 artifacts UI that can execute Python Code and render React Components on-the-fly. https://t.co/bsDCg3aCnj
— Shubham Saboo (@Saboo_Shubham_) Jul 11, 2024
My friend and Neo4j CTO @prathle has written an outstanding blog post that summarizes the recent buzz around GraphRAG, what we've learned from a year of helping users build systems with Knowledge Graphs + LLMs and where we believe the space is going. Thread below. 👇🧵 https://t.co/TrwkJEH8hL
— Emil Eifrem (@emileifrem) Jul 11, 2024
I attended Citi’s Gen AI Summit in New York yesterday and continue to believe we are early in the AI adoption cycle (but like other technology / innovation cycles, the trajectory will be non-linear)… Atif Malik and team put together a list of takeaways that I thought were worth… https://t.co/mXZvlytDPw https://t.co/ONj7TFamvt
— Radnor Capital (@RadnorCapital) Jul 10, 2024
// A Survey on Mixture of Experts // Great survey paper on Mixture of Experts (MoE), including the technical details of MoE, open-source implementations, evaluation techniques, and applications of MoE in practice. (link in ALT) https://t.co/VKPjqKhmsg https://t.co/ADgoz2ChoQ
— elvis (@omarsar0) Jul 10, 2024
From @natfriedman's personal site (former Github CEO) https://t.co/BYuNGDLV30
— Ben Lang (@benln) Jul 9, 2024
Self-recommending -- read this book! https://t.co/nZu4u5TubQ 🔥🚀💥 @m2jr https://t.co/35MLwOADsq
— Marc Andreessen 🇺🇸 (@pmarca) Jul 10, 2024
A 3 hr masterclass distilled the secrets of an Amazon VP's 15-year rise. I mind-mapped every insight. Here's how you break through from senior manager to true executive according to @EthanEvansVP → Transitioning to an executive role requires new behaviors and strategic… https://t.co/2paLqwrhXD https://t.co/KTdQPPhDQN
— Jason Shen · The Outlier Coach (@JasonShen) Jul 10, 2024
The idea behind Apple Intelligence is a stroke of genius! They built a large foundation model and fine-tuned it on multiple tasks. But they are doing something very clever: They built adapters they plug into the large model to fine-tune it on the fly for specific tasks.… https://t.co/O2Q0dgzgqY https://t.co/VJ3em3EpsZ
— Santiago (@svpino) Jul 9, 2024
One of our fintech clients was quoted $105k USD by a consulting firm to automate their KYC with LLMs 🤯 Instead, they used @StackAI_HQ to build it on-prem in a week. They broke down their workflow like this: 📩 Processing customer: Receive an email with the customer documents… https://t.co/NrMKZ43j5p https://t.co/Q9pQSNKw7F
— Bernardo Aceituno (@BernAceituno) Jul 8, 2024
Chrome's new `window.ai` feature is going to change the web forever! 🤯 It allows you to run Gemini Nano, a powerful 3.25B parameter LLM, 100% locally in your browser! We've also added experimental support to 🤗 Transformers.js, making it super easy to use! 😍 Check it out! 👇 https://t.co/O5B2uVZyqr
— Xenova (@xenovacom) Jul 8, 2024
Just created the first DSPy nodes graph that uses more than one processing step https://t.co/AThHJQHFzP
— Tom Dörr (@tom_doerr) Jul 8, 2024
We are entering the new SaaS era, moving from Software-as-a-Service to Service-as-a-Software. Until now, software access has helped analysts do their job better. In finance, Bloomberg or Capital IQ improves employee productivity. But in the not too distant future, an AI agent… https://t.co/ZXEyz9kpEP https://t.co/rV3u8YD74Z
— Alex Immerman (@aleximm) Jul 8, 2024
A few more stills from our AI generated cinematic in development - Heaven's Magic. You can view the first episode in development: https://t.co/xROlbvcqIy Read the original story on @wattpad : https://t.co/eCSGLjgbKa Tools: @midjourney @Magnific_AI @elevenlabsio @runwayml https://t.co/iunnsQv60Y
— Studio Magic (@studiomag1c) Jul 8, 2024
I replaced a $2000/month predictive analysis software with Julius AI. In this video, I'll show you how to create a complex B2B marketing dataset using GPT. Then, I'll identify companies with a high likelihood of churn, determining their reasons for doing so. Julius AI achieves… https://t.co/9R9Jmj4Mze https://t.co/OEubDqpPq0
— Muratcan Koylan (@youraimarketer) Jul 7, 2024
Pluggable Copilots https://t.co/9RvivKJe7m
— Samuel Timbó (@io_sammt) Jul 7, 2024
Everyone should go through the whole “explorations of capabilities” section at least twice. It’s hard to fully absorb all of it the first time. I’d bet that 99% of the people who watched the May 13th livestream don’t understand how cool GPT-4o actually is https://t.co/knuGKCjcMs
— Spencer Schiff (@SpencerKSchiff) Jul 6, 2024
Nice Sunday read for an overview Paper on RAG - Basics of RAG workflow 👇 📌 Query Classification : Classify queries to determine the necessity of retrieval. Queries requiring retrieval proceed through the RAG modules; others are handled directly by LLMs. 📌 Chunking :… https://t.co/PcEUPhqNfu https://t.co/eXkysNnyjh
— Rohan Paul (@rohanpaul_ai) Jul 7, 2024
[Transformer] by Hand ✍️ Make Your Own 🛠️👉 https://t.co/3gHYexraoS Over the past few months, I've collaborated with several AI educators to customize my AI by Hand exercises. I am glad that my materials are being used and appreciated in many classrooms around the world!… https://t.co/CGKPvLUQpy https://t.co/LZvaTTLluz
— Tom Yeh | AI by Hand ✍️ (@ProfTomYeh) Jul 7, 2024
1/n Scaling Synthetic Data Creation: A Billion Personas Approach A recent technical report from Tencent AI Lab Seattle introduces a novel approach to synthetic data creation that could potentially revolutionize how we generate diverse, high-quality data at scale. The paper,… https://t.co/woRN3lP0ZG https://t.co/34eQNNBYpW
— Carlos E. Perez (@IntuitMachine) Jul 6, 2024
Just read this paper "Grokked Transformers as Implicit Reasoners". The authors really went all-in on understanding how transformers learn to reason implicitly through this "grokking" phenomenon. They focused on two key reasoning tasks: composition (like chaining facts) and… https://t.co/lYKBqsGn46 https://t.co/CsKK8g6sUh
— Avi Dhaliwal (@champydaku) Jul 5, 2024
https://t.co/RbZSSZvtRn
— Ate-a-Pi (@8teAPi) Jul 6, 2024
AppSumo is the leader in terms of offering business deals. However, they take a HUGE commission >70% on each sales. I’m thinking about building an AppSumo’s competitor more respectful for creators. Do you think it's a good idea?
— Pauline - SaaS, AI & data (@Pauline_Cx) Jul 5, 2024
Did you pass the test? https://t.co/Js7HAm8kMI
— ᴀʀᴛ ᴏꜰ ᴘʜʏꜱɪQᴜᴇ (@Artofphysique_) Jul 6, 2024
IMO, open datasets are more impactful than open models these days! You can now filter almost 200,000 of them on HF by modality, size and format: https://t.co/gqmXtazVuw https://t.co/lRDTVws7mv
— clem 🤗 (@ClementDelangue) Jul 5, 2024
Vector Calculus https://t.co/kJsqXgnBiD https://t.co/eBW1qrTG7l
— Jean de Nyandwi (@Jeande_d) Jul 4, 2024
@HBCoop_ This is our first cinematic visualized entirely with AI: https://t.co/Du43oqeVHf
— Studio Magic (@studiomag1c) Jul 6, 2024
LangWatch for DSPy looks interesting https://t.co/MEamzW0yRa
— Tom Dörr (@tom_doerr) Jul 4, 2024
Annual Dividends From $1,000 Invested: 🍎 $APPL $4.50 🛒 $COST $5.40 💳 $MA $5.90 🖥️ $MSFT $6.50 🌍 $V $7.70 🍕 $DPZ $12.10 🛍️ $WMT $12.20 👟 $NKE $19.70 👷🏼 $LOW $21.60 🍔 $MCD $26.70 🛠️ $HD $27.00 🦺 $MMM $27.60 ☕️ $SBUX $29.90 🥤 $KO $30.60 🛢️ $XOM $33.10 🥤 $PEP $33.30 💉… https://t.co/0KQzQ59asH
— Dividend Talks on YouTube (@DividendTalks) Jul 5, 2024
Last week a rare interview with Bernard Arnault came out I read it 3 times, loved it, and decided to make an episode about it. Available in your podcast app now! https://t.co/25NvIBRune
— David Senra (@FoundersPodcast) Jul 5, 2024
GPT-2 is done It took a LOT, but pretty much every little feature and my thought process are documented, and I learned a ton. It only waits for a full pretraining -- end of week I promise. All credit to @karpathy, I only added comments making it 2head accesible. https://t.co/fBvRgbXUJJ
— Lazarz (@Laz4rz) Jul 3, 2024
Introducing Code Assistant that Runs Code in VSCode! We've enabled code execution for Python 3.10 Agent in a sandbox environment, thanks to the integration with @e2b_dev 🙌 Debug your code before integrating it into your project with just a click of "Run Code" 🤩 Currently… https://t.co/247TXZikUZ https://t.co/cEU0snrePb
— Daniel San (@dani_avila7) Jul 4, 2024
Proofs raised a $2.6M Pre-Seed to build an AI agent for APIs > 10 slide pitch deck https://t.co/zj8ERVF9BG
— Chief AI Officer (@chiefaioffice) Jul 4, 2024
the year is 2024 and you're still not using Penrose notation (pic from: https://t.co/IvUisA8q1W) https://t.co/tF5RS3VheJ https://t.co/Ik8yedl0XE
— doomslide (@doomslide) Jul 3, 2024
Yeah, semantic search is cool. But have you tried combining it with filters? In Weaviate, you can combine your semantic search query with filter operations like: - is equal or contains, - wildcards, - greater than, less than, and many more! https://t.co/HDfgWUewFh
— Leonie (@helloiamleonie) Jul 4, 2024
How to find your purpose. Ikigai—a Japanese concept that will change your life: Image Credit: Justin Wright https://t.co/KRHAcsj96v
— Ben Meer (@SystemSunday) Jul 3, 2024
The longest book I’ve ever read for the podcast was a 1000 page biography of Enzo Ferrari The author just released an updated — and shorter version that I’m going to read In case you want to join me: https://t.co/eNnMWVSJPW
— David Senra (@FoundersPodcast) Jul 3, 2024
AI Visual Studio Code Extension that lets me write optimized Python code for LLM apps, AI agents and RAG apps without running into errors. Traycer is an AI extension that provides me with smart, real-time feedback as I code. It stays in the background & appears when I need it. https://t.co/c14q9Zq1f8
— Shubham Saboo (@Saboo_Shubham_) Jul 3, 2024
Anthropic CEO Dario Amodei says another generation or two of scale is needed before the error rate is low enough to have autonomous AI agents https://t.co/ubu5QQe0Gf
— Tsarathustra (@tsarnick) Jul 2, 2024
how to get lucky https://t.co/AJ8aHEcAyU
— Z Fellows (@zfellows_) Jul 3, 2024
OpenAI Co-founder Andrej Karpathy explains the new computing paradigm: "We're entering a new computing paradigm with large language models acting like CPUs, using tokens instead of bytes, and having a context window instead of RAM. This is the Large Language Model OS (LMOS)" https://t.co/FupH3dVfV7 https://t.co/Gk9G0Tu6aU
— Haider. (@slow_developer) Jul 3, 2024
Speed of execution is the moat inside which live all other moats. Speed is your best strategy. Speed is your strongest weapon. Speed has THE highest correlation to mammoth outcomes. Those who conflate speed w/ 'thoughtlessness' haven't seen world class execution @ speed. E.g.:
— Hemant Mohapatra (@MohapatraHemant) Jul 2, 2024
Incredible stuff https://t.co/sJmbZDqA36
— Kris Kashtanova (@icreatelife) Jul 2, 2024
“RAND signed a contract with John von Neumann to produce a general theory of war, to be completed during a small slice of his time: that spent shaving. For his shaving thoughts, von Neumann received $200 a month” I first read that story in this book. Seriously, von Neumann was… https://t.co/HWVMnKzz7Y https://t.co/xhNz0vYvI2 https://t.co/ntGUjUxqE7
— milton (@tensor_fusion) Jul 1, 2024
Writing a literature review (annotated example) https://t.co/3ROAhHYVPb
— Prof Lennart Nacke, PhD (@acagamic) Jul 2, 2024
Claude Sonnet 3.5 transforms a simple PDF earnings report into an interactive dashboard in just 30 seconds. It goes beyond the capabilities of GPT-4o, Gemini Pro, Llama and other existing LLMs. Future of work will 10x more productive with AI. https://t.co/7bo6fGWSPJ
— Shubham Saboo (@Saboo_Shubham_) Jul 1, 2024
Some audiobook recommendations: The Story of Civilization by Durant Iliad (Penguin Edition) The Road to Serfdom by Hayek American Caesar by Manchester Masters of Doom by Kushner The Wages of Destruction by Tooze The Storm of Steel by Junger The Guns of August by Tuchman The… https://t.co/DnFA4JSygP
— Elon Musk (@elonmusk) Jul 1, 2024
