星々は私たちに、決してひとりではないと教えてくれる。 https://t.co/B313HuJXPg
— AI_CyberDreams (@AI_CyberDreams) Aug 30, 2024
from Twitter https://twitter.com/AI_CyberDreams
August 30, 2024 at 03:06PM
via IFTTT
星々は私たちに、決してひとりではないと教えてくれる。 https://t.co/B313HuJXPg
— AI_CyberDreams (@AI_CyberDreams) Aug 30, 2024
Apple Eye-Tracking Patent Earlier this year Cornell revealed a research project called GazeTrak (video below) that uses sonar (ultrasonic transducers or sound waves), for eye-tracking with advantage of ultra-low-power. Filed in 2021, Apple was just awarded a patent for same. https://t.co/v3SpKFTFrq https://t.co/hYI8WsgDit https://t.co/pHdWKxC11h
— Christopher Grayson (@chrisgrayson) Aug 30, 2024
Here's Sam Altman, Founder of OpenAI, on cold outreach: Very practical advice indeed. Even if you are a tech genius like him, you NEED reliable lead gen, https://t.co/ElJN9rUjKI
— Tibo (@tibo_maker) Aug 29, 2024
Dylan Field (Figma CEO) on product management https://t.co/ZhBsrtFNKR
— George from 🕹prodmgmt.world (@nurijanian) Aug 29, 2024
Box CEO Aaron Levie on why AI agents are a $1 trillion opportunity "AI will not just be serving back an answer for you, it will also go and do something for you ... as tokens come down in cost, we can scale up intellectual labor and knowledge work [with AI] orders of magnitude… https://t.co/rY84apqacf https://t.co/Pw5Bm4JNqI
— Overlap: Business & Tech (@Overlap_Tech) Aug 30, 2024
Morgan Housel knows the best Warren Buffett stories. Here's one about patience and leverage from his recent appearance on the "Behind the Memo" podcast. https://t.co/hZD0H7SZ45 https://t.co/MdUDQs9Vz4
— Kevin Carpenter (@kejca) Aug 29, 2024
We made distillation and spec decoding work with Mamba (and linear RNNs in general)! Up to 300 tok/sec for 7B🚀. Spec dec is nontrivial as there's no KV cache to backtrack if some tokens aren't accepted, but there's an efficient hardware-aware algo to recompute the SSM states https://t.co/hBuyaDm7Ia
— Tri Dao (@tri_dao) Aug 29, 2024
BREAKING: Palantir $PLTR gets a NUMBER #1 ranking for its AI/ML platform by Forrester. The Forrester Wave is crucial for the tech industry since it provides standardized evaluations, boosts market visibility, aids decision-making, and offers strategic insights that influence… https://t.co/SCwENPxgS0 https://t.co/RAXSgfhe1j
— amit (@amitisinvesting) Aug 29, 2024
Sam Altman's life advice: https://t.co/ESAp5m3wFW
— Ben Lang (@benln) Aug 28, 2024
You can Crawl entire website with Claude 3.5 or GPT4-o with this open-sourced tool firecrawl. 💯 Turn entire websites into LLM-ready markdown or structured data. Scrape, crawl and extract with a single API. Crawls all accessible subpages and give you clean data for each. No… https://t.co/so9drrm8dz https://t.co/XiXCRNpjwx
— Rohan Paul (@rohanpaul_ai) Aug 29, 2024
Map of the system of human knowledge https://t.co/FsG9wdCVVh
— Prof Lennart Nacke, PhD (@acagamic) Aug 28, 2024
feels right. https://t.co/ukj3kmRpYj
— 🍓🍓🍓 (@iruletheworldmo) Aug 26, 2024
The State of Claude Artifacts - August 2024 With Artifacts generally available across Claude dot ai and our mobile apps, I figured now is a good time to record a walkthrough of everything that you can do with Artifacts today. Artifacts has so much potential as a new UX… https://t.co/ctLySb6sNj https://t.co/89XPRCsupx
— Alex Albert (@alexalbert__) Aug 28, 2024
City of Yggdrasil. https://t.co/ab3i4jACcI
— Craftian (@Craftian_BK) Aug 28, 2024
JENSEN HUANG WAS JUST ASKED ABOUT THE HYPERSCALERS SPENDING ON CAPEX & HOW THE STREET SHOULD THINK ABOUT AN EVENTUAL ROI FROM AI. This was one of the most important questions during the call, because essentially the analyst asked... "Why should we still believe your clients… https://t.co/6lZMmuyl2y https://t.co/JHwGTjv0Zg
— amit (@amitisinvesting) Aug 28, 2024
A new model from OpenAI called Strawberry is set to come this fall. Here's what we know from @theinformation: • Solves complex problems on the first try without hallucinations • Generates high-quality training data for 'Orion', OpenAI's next flagship LLM • OpenAI is… https://t.co/ltzjVL31EQ
— Dan Shipper 📧 (@danshipper) Aug 28, 2024
並んで戦えば、無敵だ! https://t.co/46QsGZJU6D
— AI_CyberDreams (@AI_CyberDreams) Aug 26, 2024
We've teamed up with @cerebrassystems to bring you lightning-fast LLM workflows ⚡️ With our latest integration, you can build and deploy faster than ever. Cerebras supports 1,800t/s for Llama3.1-8b and 450t/s for Llama3.1-70b. Try out the Cerebras Module on LangChain:… https://t.co/K77Hd36BoI https://t.co/pq4hNqKFYg
— LangChain (@LangChainAI) Aug 27, 2024
6 things I didn't know about Cursor Watched the first hour of Coding with Cursor @shaoruu @shl livestream from last night. Here are the first 6 things I did not know. And i'm only 56 minutes in of 2h46min. More to follow.
— Imrat (@imrat) Aug 27, 2024
Shoshin: The Beginners Mind. In the beginner’s mind there are many possibilities. In the expert’s mind there are few. - Shunryu Suzuki ❤️🇯🇵 https://t.co/pO4eD1EaHZ
— Marc Benioff (@Benioff) Aug 27, 2024
The Mamba Mentality - Kobe Bryant ‼️ https://t.co/5nhingRnQf
— Reads with Ravi (@readswithravi) Aug 27, 2024
Use Cursor + custom prompts to build features instantly. In 60 seconds Cursor Composer builds me an *entire* auth system from a single instruction file. Then I create a Composer Project to handle any work I need done on that feature. Copy this easy trick to build faster. https://t.co/F15csnHgp6
— Mckay Wrigley (@mckaywrigley) Aug 27, 2024
There's an IDE for creating AI applications (from the creators of PyTorch Lightning) https://t.co/f94geCMRqK
— Tom Dörr (@tom_doerr) Aug 26, 2024
🔥Releasing RAG Course on @OReillyMedia 🚀Super excited to release a course on @OReillyMedia on Building RAG Applications with @llama_index: From Basic Components to Advanced RAG Systems. This 8-module course takes you from the basics to advanced techniques in RAG system… https://t.co/tVMEueWHh4 https://t.co/2cyVIOJ2t9
— Ravi Theja (@ravithejads) Aug 26, 2024
Great tutorial by mckaywrigley on making a call to Vercel v0 from Cursor AI. This is useful because, once configured, Cursor can create a link with prompt that can be launched directly. Since Vercel v0 accepts query parameters, it works seamlessly. https://t.co/fz5Ig49tPA https://t.co/N9QmPKZdc4
— AshutoshShrivastava (@ai_for_success) Aug 26, 2024
I rewatch this once a month for motivation https://t.co/L5NKPvWA7V
— Yasser (@yasser_elsaid_) Aug 25, 2024
Cursor + Claude Sonnet 3.5 is good, but combine it with Vercel v0, and it becomes even more powerful. Now, you have both a backend and frontend end-to-end solution stack. https://t.co/Qwewwul7Y6
— AshutoshShrivastava (@ai_for_success) Aug 25, 2024
ideas on how to compete with $100B saas companies: - intentionally limit features (focus on core value) - give affiliate program 50% cut until $1M ARR (then lower) - shift from monthly subscription to "quantum pricing" (dynamic based on usage/value) - kill your best-selling… https://t.co/0nhACfXUgn https://t.co/b2C3b14R5g
— GREG ISENBERG (@gregisenberg) Aug 25, 2024
I think this google research proposal to add a pipe operator to SQL is really good in my opinion. There is a lot of bang for the buck. This cleans up so many parts of SQL in one go. I'm a fan. Wonder if we should implement it for ShopifyQL? https://t.co/gL0lZIrn2v and… https://t.co/7XCZ2vL5Ix
— tobi lutke (@tobi) Aug 25, 2024
Can’t recommend The Iliad enough! Best as Penguin audiobook at 1.25 speed. https://t.co/gh7VEytlRe
— Elon Musk (@elonmusk) Aug 24, 2024
If it’s good enough for Andrej, then it’s good enough for you. Learning how to use Cursor and build software is literally the highest leverage thing you can do in the entire world right now. Embrace the future and build. https://t.co/0cwsLgVDgR
— Mckay Wrigley (@mckaywrigley) Aug 24, 2024
Meta silently dropped Sapiens: a series of vision foundation models specialized on humans 👯 all weights open, they release segmentation, depth and keynote models with various sizes on @huggingface 🤗 also h/t they uploaded torchscript compiled models too! 😍 https://t.co/Fp9KmxOtcO
— merve (@mervenoyann) Aug 24, 2024
Breaking News—Ideogram 2.0 just dropped! It’s already being called the future of text-to-image AI. Here are 12 Insane examples you won't believe 🤯 https://t.co/KX041uEbku
— Alamin (@iam_chonchol) Aug 24, 2024
Navy SEALs are the most elite warriors on Earth. Here's one thing their taught early on... The Backwards Law (how to get what you want by letting go): https://t.co/Nwjo3pIDIZ
— Josh Gessner (@GessnerJosh) Aug 24, 2024
Ads will never be the same. People are creating next-level AI commercials. 12 wildest creations: https://t.co/fFcZbaLmSh
— Madni Aghadi (@hey_madni) Aug 24, 2024
Meet OPEN SOURCE AND FREE SakanaAI/ The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery. I have been running a lot of tests on this for quite a bit. Enjoy uncensored SCIENCE. https://t.co/wXco5ZrlWF
— Brian Roemmele (@BrianRoemmele) Aug 23, 2024
For finetuning workflow, Linkedin's just launched Liger Kernel will immediately give you - increase multi-GPU training throughput by 20% and - reduces memory usage by 60%. And you need to a single line of code change `apply_liger_kernel_to_llama()` https://t.co/MTIYvyZcBu
— Rohan Paul (@rohanpaul_ai) Aug 24, 2024
@karpathy Shameless plug, but I just released a 3hr 48min course on Cursor on Wednesday. It’s what I’d want for myself if I were just starting out - 100% coverage of all features. All content shown here: https://t.co/P2XHOGeV3P My YouTube video gives you a taste: https://t.co/pymV4m61qJ
— Mckay Wrigley (@mckaywrigley) Aug 24, 2024
“Big opportunities in life have to be seized. We don't do very many things, but when we get the chance to do something that's right and big, we've got to do it.” — Warren Buffett https://t.co/hU4n8WFfms
— Reads with Ravi (@readswithravi) Aug 23, 2024
Sakana AI, the automated research scientist, is an open-source project! Should I make a tutorial for it? https://t.co/NcFisgEGhI
— MatthewBerman (@MatthewBerman) Aug 23, 2024
Think ads need to break the bank? AI is changing the commercial game. 10 wild examples: https://t.co/oiXRRDGCrC
— Min Choi (@minchoi) Aug 23, 2024
Fitting together lumps of irreducibility? https://t.co/iRcgolgwkR
— Carlos E. Perez (@IntuitMachine) Aug 22, 2024
The most fun & exciting part of Agentforce isn’t just that Atlas is outperforming other agents by 2x—it’s that we’re reimagining Salesforce as the first, large-scale, “Agent-first” enterprise software company. Humans with AI Drive Customer Success. Embracing Shoshin, with… https://t.co/M6Eg4YoGBs https://t.co/i7f8BZ1QEv
— Marc Benioff (@Benioff) Aug 22, 2024
Graph-based memory is going to be huge. Just a matter of time until engineering catches up to layer this onto LLMs and agents. Awesome work Mem0 team! 💪 https://t.co/iaTC8cabGR https://t.co/VuNQNT9JIJ
— Ted Werbel (@tedx_ai) Aug 21, 2024
Being unable to do one Pull-up reflects a massive leak in our bodies. Star players are missing from your posture's lineup. This weakness affects muscles all over your back and the front of your arms and wreaks havoc from your elbows to your lower spine. It's never too late to… https://t.co/7JqYeUWs3Y https://t.co/hWzO346uLk
— Alex Bernier (@mythoughtfood) Aug 22, 2024
How Pieter Levels launched https://t.co/WKqu2fEkCu with no automations and got 1000+ paid customers in the first 24 hours. Landing page, stripe payment link and a typeform to submit your photos. A great example of "doing things that don't scale" https://t.co/cFGQTad7jY
— Anatolii (@quasa0) Aug 22, 2024
@gunsnrosesgirl3 10 incredible facts about the octopus. https://t.co/C0G4f8EwW6
— True Crime (@TrueCrimeNFT) Aug 22, 2024
🎵 Google publishes about a new way to recommend music using Transformer models 🔍 Adapting to changing user preferences based on context 💡 Solution: • Combines Transformer with existing ranking model • Processes sequences of user actions to understand context 📊 Input… https://t.co/Ym7oJlEgn8 https://t.co/KVODtn8ypy
— Rohan Paul (@rohanpaul_ai) Aug 20, 2024
A Redditer automated Leetcode using Claude’s 3.5 Sonnet API and Python. The script completed 633 problems in 24 hours, completely autonomously. It had a 86% success rate, and cost $9 in API credits. Now the point is, in 3rd quarter of 2024, it's so unsurprising that LLMs will… https://t.co/EvHiQFQNRi https://t.co/39MHwEKHZ5
— Rohan Paul (@rohanpaul_ai) Aug 21, 2024
Introducing AI Youtuber (https://t.co/QOK1afNpZM) - make 100X cheaper ❌ $100+/video ✅ $2/video Use it for making a not "faceless' automated channel. AI Avatar + AI Caption + AI Brolls 🔥 https://t.co/F2FygpFyxc
— Brivael (@BrivaelLp) Aug 20, 2024
Tiers of web scraping (easiest to hardest): 1. You already have the URL, parse HTML (beautiful soup) 2. You already have the URL, javascript spaghetti, selenium) 3. You don't have the URL, clicks needed (playwright) ~~AI zone~~ 4. Few clicks needed, but not sure which… https://t.co/Eu5hvehdan
— Greg Kamradt (@GregKamradt) Aug 20, 2024
.@MultiOn_AI Offers an agent API https://t.co/3bwZ1PxXv8
— Tom Dörr (@tom_doerr) Aug 20, 2024
AI Software Engineering Agent that builds custom AI agents with RAG from simple English prompts. https://t.co/rA69xcZkbm
— Shubham Saboo (@Saboo_Shubham_) Aug 20, 2024
RT @balajis: Ok, wow. We received thousands of applications from 80+ countries for the Network School! We’re adding capacity and will send…
— Gopi Vikranth (@GopiVikranth) Aug 20, 2024
Ok, wow. We received thousands of applications from 80+ countries for the Network School! We’re adding capacity and will send admission decisions soon. https://t.co/axy0kBh9Tq
— Balaji (@balajis) Aug 19, 2024
Stop training your own model! Salesforce's newest version of Agentforce, code named Atlas, achieves state-of-the-art results in answering users' queries for our top customers: > 95% effectiveness in preventing hallucinations and > 90% answer relevance — bakes offs show it… https://t.co/L8JAcc21Hk https://t.co/VRbHyfNLeR
— Marc Benioff (@Benioff) Aug 19, 2024
Margin of Safety" by Seth Klarman is an incredible investing book. But a used copy costs more than $2,000! Here are 6 lessons I learned from him: https://t.co/Yn4Y9OR38U
— Compounding Quality (@QCompounding) Aug 19, 2024
this was a very deep and technical dive into all things agents and evaluation! was a lot of fun. I've been a listener of twiml for a while, so was an honor to be on https://t.co/soMu13LYXF
— Harrison Chase (@hwchase17) Aug 19, 2024
15 books recommended by Peter Thiel: 1) The Sovereign Individual by James Dale Davidson and William Rees-Mogg https://t.co/w7m82Ar4Yi https://t.co/vMZw1RhS8E
— Reads with Ravi (@readswithravi) Aug 19, 2024
My article on the State of the GenAI market is out. Covering: - OpenAI's edge vs $GOOGL, $META & Anthropic. - Custom silicon development by $AMZN, $MSFT, $GOOGL - $META and social media GenAI effects - $ORCL the 4th player - Databricks, $PLTR & more. https://t.co/1M715p1Bv9
— Rihard Jarc (@RihardJarc) Aug 19, 2024
NEW: Roblox is Already the Biggest Game In The World. Why Can't It Make a Profit (& How Can It)? --- At 380M, Roblox probably has more MAUs than (de-duped) PS/Xbox/Switch/Steam combined, and is 2x as played as Disney+. Can its biz ever match its scale? https://t.co/SUizDaJDfO
— Matthew Ball (@ballmatthew) Aug 18, 2024
Japan has been collecting data on the date of peak cherry tree blossom since 812 See if you can spot climate change... https://t.co/kXDJBGnjZi https://t.co/3jSv137cXo
— Alec Stapp (@AlecStapp) Aug 18, 2024
Best bits of @GaryMarcus keynote at #AGI24 - we uploaded the full video on MLST. https://t.co/LRkXr26MLJ
— Machine Learning Street Talk (@MLStreetTalk) Aug 18, 2024
I’ve heard many people say “Concept Artists are screwed” over the last couple years. People will likely be surprised to find concept artists to be the group most equipped to build their personal franchise empires in early Ai World Engines. Tools: Procreate Magnific (style… https://t.co/vUWq1Isa23 https://t.co/d6qBI6qlPA
— Jon Finger (@mrjonfinger) Aug 18, 2024
HN user in 2022 wrote: "The opportunity in AI all seems to be on the front end" 2 yers later, this turned out to be very true: The foundational models are free to use and keep getting updated and it's about what you built on top of it for your users that brings value https://t.co/uVKpZHsPLh
— @levelsio (@levelsio) Aug 18, 2024
In @lennysan 's latest podcast, Tanguy Crusson (Head of Product Atlassian) throws truth bombs one after another on aspects of building new businesses and products. First 60 minutes are a gem. 1/ Know the Decision Makers better than your competition Intimately understand who… https://t.co/ODiovGDdTP
— Bandan (@bandanjot) Aug 18, 2024
Unpopular opinion I agree with- learning should not be fun. This blurb from @karpathy summarises it very well. https://t.co/G0xVPsiwZl
— Udayan Walvekar (@udayan_w) Aug 17, 2024
The best interactive visualization of Transformer Architecture ever made 🤯 https://t.co/jXoZEoeR3u https://t.co/Aq0tBM9f2e
— Victor M (@victormustar) Aug 17, 2024
You can completely skip university and just watch Karpathy’s 10 lectures on building GPT-2 from scratch. This is not an exaggeration.
— James Campbell (@jam3scampbell) Aug 17, 2024
The core concept of knowledge distillation from Gemma-2 paper. Knowledge distillation is becoming very crucial. This week we saw the great result from @nvidia - that a 4B Llama 3.1 model from Llama 3.1 8B combining weight pruning and knowledge distillation 🧬 Knowledge… https://t.co/CDx5pLgTNx https://t.co/l1hYpr11fm
— Rohan Paul (@rohanpaul_ai) Aug 17, 2024
The brilliant AI pioneer @demishassabis explains how machines will have to conquer “planning” in order to move from the current models of artificial intelligence to become true artificial general intelligence. https://t.co/HsvDHvL68e
— Walter Isaacson (@WalterIsaacson) Aug 16, 2024
Just had a *mind blown* moment with Lindy AI I accidentally enabled the AI agent I built on Friday that validates inbound leads, sends hyper-personalized emails, and negotiates on my behalf... ... and it WORKED For context, I'm new to the platform and was tinkering with it… https://t.co/AFhtvK1UrA https://t.co/OXoA3iJ93b
— Rowan Cheung (@rowancheung) Aug 17, 2024
RAGChecker: a fine-grained evaluation framework for diagnosing retrieval and generation modules in RAG. Shows that RAGChecker has better correlations with human judgment. Reports several revealing insightful patterns and trade-offs in design choices of RAG architectures.… https://t.co/IoqBMLzyRN https://t.co/ZgwCJQszVM
— elvis (@omarsar0) Aug 16, 2024
“How do I build my social presence from scratch?” is the most common question I get from startups. So we (@a16zcrypto) put together an easy to follow guide. It includes hyper-specific tips, tools, frameworks, & more. Let's dig in. 🧵 https://t.co/MaSu6gmLwJ https://t.co/wKFYelyHpd
— Ish Verduzco (@ishverduzco) Aug 16, 2024
Wow. Movies and commercial video production is going to get 1000x cheaper. Midjourney, Flux, Gen-3, Luma, Kling https://t.co/7Ykp8ROdxd
— Rohan Paul (@rohanpaul_ai) Aug 16, 2024
DeepMind's Diffusion Augmented Agents (DAAG) Explained Imagine you're learning to cook. You wouldn't expect to become a master chef by only tasting successful dishes, right? You need to experiment, make mistakes, and understand why something didn't work out. But what if you… https://t.co/KyDAt84kQv https://t.co/Ra9o79mmCM
— Carlos E. Perez (@IntuitMachine) Aug 16, 2024
🤯 This 100% AI generated ad for eToro aired during the Paris Olympics Opening Ceremony. But did it work? Is this the future of advertising? More importantly what do you think? https://t.co/NtVbiB5Xm3
— Linus ●ᴗ● Ekenstam (@LinusEkenstam) Aug 16, 2024
What Walmart has done in digital is kind of amazing. (via B of A) $WMT https://t.co/dPrxGoGJqD
— Carl Quintanilla (@carlquintanilla) Aug 15, 2024
This is wild. Virtual try-on will change e-commerce for ever. All these images are 100% AI This is how: https://t.co/VUQUMC9CmS
— Linus ●ᴗ● Ekenstam (@LinusEkenstam) Aug 14, 2024
Cursor Composer + Control Panel is by far the best AI codegen tool in the world. This workflow is ridiculous. https://t.co/gPf85yMGan
— Mckay Wrigley (@mckaywrigley) Aug 14, 2024
The 10 most liked paper (that I've mentioned) for the month of July. (generated by Grok) https://t.co/U4qAI6Qfx5
— Carlos E. Perez (@IntuitMachine) Aug 15, 2024
Automatic knowledge graph construction can be slow and expensive. Also I find there's a lack of resources on how to build something principled (do you just stuff text into an LLM prompt?) That's why I love this blog by @tb_tomaz which not only outlines the step-by-uses Relik, a… https://t.co/IUSrPofgAK https://t.co/2ycK5duydv https://t.co/Yl0tBXEkno
— Jerry Liu (@jerryjliu0) Aug 16, 2024
11x is adding $700K+ ARR/month building AI employees like Alice (AI SDR) Today they're launching Jordan, an AI voice agent for Sales/CS phone operations. Here's why we're excited about the launch: https://t.co/EsinCAz9HQ
— Chief AI Officer (@chiefaioffice) Aug 15, 2024
We've been working on this for quite some time, and I'm thrilled to share a preview of Arcee-Swarm. Instead of relying on a single large generalist model, Swarm utilizes multiple domain-specialized models working together to deliver exceptional results with both speed and nuance. https://t.co/ZyypiDd4rO https://t.co/CwYP10teAV
— Lucas Atkins (@LucasAtkins7) Aug 14, 2024
A ramen stall in Japan https://t.co/SFOiRKFBa4
— Science girl (@gunsnrosesgirl3) Aug 14, 2024
Eric Schmidt throws down the gauntlet: Google is losing to startups because of life-work balance nonsense/wfh. 🔥 https://t.co/CV4FL7g5H9
— @jason (@Jason) Aug 14, 2024
old yc interviews are a goldmine https://t.co/4gZq1Li3Kx
— keshav (@keshavchan) Aug 14, 2024
This is such a big deal. Imagine caching entire books worth of content, any coding documentation you want, and retrieving it with multiple API calls. Imagine the sophisticated agents you can now build with this. https://t.co/dwgS6MrPkv
— Pietro Schirano (@skirano) Aug 14, 2024
✨Integrate Microsoft’s GraphRAG into Neo4j and implement retrievers using @LangChainAI and @llama_index ✨ This is the second part of the blog about Integrating Microsoft #GraphRAG into #Neo4j by @tb_tomaz 👇 https://t.co/hPABso2dVq
— Neo4j (@neo4j) Aug 14, 2024
Next gen of LLMs for E-Commerce arrived! 🛒 🛍 Xinyi Ling presents eCeLLM which comes in 3 sizes (S, M, L). As you guessed it can be used to boost Search, Recommendation, and product-based Question-Answering https://t.co/uZFc4Km2e3 #icml #icml24 #icml2024 https://t.co/Fr8LPiWrld
— Florian Hönicke (@florianhoenicke) Aug 13, 2024
A short explainer video for “The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery” created by @omarsar0 YouTube: https://t.co/BinZrthJRf Paper: https://t.co/cVRR95cPap https://t.co/24zmoL8LkX https://t.co/pWClXcyyrd https://t.co/HOkbqNCssW
— Sakana AI (@SakanaAILabs) Aug 14, 2024
Freepik Flux Realism Mode has been released. The most realistic AI Image model ever, and the results are absolutely incredible. Video created with Runway GEN-3. https://t.co/m8hX0x8sQA https://t.co/cCYofT8zii
— Jerrod Lew (@OnwardsProject) Aug 13, 2024
In a Stanford talk posted today, Eric Schmidt says the reason why Google is losing to @OpenAI and other startups is because Google only has people coming in 1 day per week 👀 https://t.co/XPxr3kdNaC
— Alex Kehr (@alexkehr) Aug 13, 2024
It's a $100B+ opportunity to create agentSREs. Success requires expertise in ML, LLMs, and distributed systems. The field is wide open for multiple decacorns. Building in this space? Email me at jgupta@foundationcap.com https://t.co/AuOsekkphQ @FoundationCap @ashugarg
— Jaya Gupta (@JayaGup10) Aug 13, 2024
This is one of the cleanest implementations of a complex multi-agent system that I've seen. Props to @seldo. All the "multi-agent" code can be implemented in a single Python class as a set of decomposable steps. You get ✅ all the benefits of an event-driven architecture (high… https://t.co/FNMUaq9PRy https://t.co/aPcb56IBbT https://t.co/Kx2x3jpJOj
— Jerry Liu (@jerryjliu0) Aug 13, 2024
8-12-2024 ($) •Olympic Ratings •Peacock’s Abundance •Comcast Earnings https://t.co/luD0YYowQQ
— Stratechery (@stratechery) Aug 12, 2024
Finetune Mistral Large 2 or Codestral on your data in minutes without writing a single line of Python Code (100% Nocode): https://t.co/qXZbQbrYb5
— Shubham Saboo (@Saboo_Shubham_) Aug 13, 2024
@simonw Is this what you're looking for? Thierry from @OctoAICloud used Weave to show how a finetuned LLama was better for PII reduction than GPT-4 in a recent Webinar https://t.co/1NdjbszUfY
— Alex Volkov (Thursd/AI) (@altryne) Aug 12, 2024
Do any of the companies that are selling LLM fine-tuning as a service have a compelling live interactive demo of a fine-tuned model I can try out, which demonstrates how that model is “better” for a specific use-case than the foundation model it builds on top of?
— Simon Willison (@simonw) Aug 12, 2024
Introducing The AI Scientist: The world’s first AI system for automating scientific research and open-ended discovery! https://t.co/8wVqIXVpZJ From ideation, writing code, running experiments and summarizing results, to writing entire papers and conducting peer-review, The AI… https://t.co/e3whOgPRQC https://t.co/SJuat9a2Uw
— Sakana AI (@SakanaAILabs) Aug 13, 2024
1/n The Next Generation of AI in Software Engineering: Moving Beyond LLMs to Intelligent Agents The field of software engineering is on the cusp of a revolution, driven by the remarkable capabilities of large language models (LLMs). These AI systems, trained on massive datasets… https://t.co/EdXF8JgXn4 https://t.co/LNZNWKOxgr
— Carlos E. Perez (@IntuitMachine) Aug 12, 2024
Star gazing in Norway 🇳🇴 https://t.co/2znwEA1m63
— OThingstodo (@othingstodo_com) Aug 12, 2024
Introducing Einstein Sales Agent on Salesforce's new Agentforce Platform! Scale your sales force easily with autonomous sales agents! Sales agents can turn deals over to humans directly in Salesforce Sales Cloud! Humans with AI drive customer success, together. 🔹 24/7… https://t.co/nC04dRDoTz https://t.co/k6acYOfe7X
— Marc Benioff (@Benioff) Aug 12, 2024
What are your favorite towns and villages in Italy? https://t.co/vsIE8T0vu9
— Architecture & Tradition (@archi_tradition) Aug 12, 2024
I'm excited to share that we've built the world's most capable AI software engineer, achieving 30.08% on SWE-Bench – ahead of Amazon and Cognition. This model is so much more than a benchmark score: it was trained from the start to think and behave like a human SWE. https://t.co/OyvqKLxcGV
— Alistair (@AlistairPullen) Aug 12, 2024
Yes, but actually building multi-agent with LLMs is only one of the contributions in the 🐫 CAMEL paper. The arxiv paper has 77 pages: https://t.co/nQMjkVv2AR. We proposed or discussed many other things that are worth to revisit such as: 🔹Data generation with multi-agent… https://t.co/sCN9EclqPh https://t.co/7gGWDv8cRI
— Guohao Li (@guohao_li) Aug 12, 2024
Flux is all over the internet with its amazing photorealism! Not only can you create images with it, but you can now also make up stories using the Story Illustrator app! Here are the 4 panels I just created, providing a story! Important Tip: Select 'Photorealistic' for the… https://t.co/w52iFTj5jf https://t.co/NtygukAghj
— Umesh (@umesh_ai) Aug 11, 2024
✨ Flux's new Realism LoRa now works on https://t.co/1vEawpIDkJ You can run it with your own trained characters and also with generic public characters (to make AI influencers for ex) I asked GPT4 to make this prompt based on a pic by @AndrewCurran_ and here's the result:… https://t.co/s6cr55DYR7 https://t.co/c1FapyFfTl
— @levelsio (@levelsio) Aug 12, 2024
✨FLUX: Cosmetic Ad 8-Step workflow and tutorial below: https://t.co/eEh0HXolhh
— Heather Cooper (@HBCoop_) Aug 11, 2024
Final blog on agent UXs (for now) 🧮Spreadsheet (agents in batch) 🖼️Generative UI (agent generates UI) 👩👧Collaborative (agent/human) https://t.co/8bX5lmHzyT If you're FE/fullstack/design, in SF, and like thinking and iterating on these ideas - please reach out we're hiring! https://t.co/p0Hie0cwvv
— Harrison Chase (@hwchase17) Aug 10, 2024
Jeff Bezos' heuristic for hiring: https://t.co/lJtPRAZKnm
— Kpaxs (@Kpaxs) Aug 10, 2024
Hattori Hanzo out now on Youtube https://t.co/fjG1o9dKWS
— Solar Heavy (@SolarHeavy) Mar 1, 2024
Paper from Ocr-2023 but very interesting. Program-of-Thought (PoT) Prompting Outperforms Chain-of-Thought (CoT) by 12% 📌 Unlike Chain-of-Thought (CoT) prompting, which uses LLMs for reasoning and computing the final answer, program-of-thoughts (PoT) prompting, delegates… https://t.co/7vrrPH4UZ3 https://t.co/1lGiW7dEI3
— Rohan Paul (@rohanpaul_ai) Aug 10, 2024
1/n An Agentic AI for Rewriting AI Captioning Imagine a world where AI can watch a video and describe it to you with the same detail and nuance as a human friend. Not just listing objects, but narrating the unfolding story, capturing the subtle interactions, and even conveying… https://t.co/JWv29aE4yO https://t.co/uHv0ErQ4CU
— Carlos E. Perez (@IntuitMachine) Aug 10, 2024
Multimodal RAG over legal, insurance, product documents 🔥 One of my favorite things about LlamaParse is that it can effortlessly crunch through both text documents and complex diagrams with different options trading off between parsing fidelity and cost - use text mode to parse… https://t.co/aXPXEgVG5L https://t.co/y9DMOaJe4Y https://t.co/IDrI2DLFri
— Jerry Liu (@jerryjliu0) Aug 10, 2024
An interactive explainer of the revolutionary Transformer https://t.co/dWiWgLTzdr
— Carlos E. Perez (@IntuitMachine) Aug 9, 2024
Transformers (LLMs) clearly explained with visuals: https://t.co/GvFcRTCX46
— Shubham Saboo (@Saboo_Shubham_) Aug 10, 2024
Larry Ellison on time: https://t.co/xN5odMu35x
— David Senra (@FoundersPodcast) Aug 8, 2024
#1 trending github repo right now looks INSANE Single image to live stream deep fake. Look at that quality!! It's called Deep-Live-Cam (link in replies) https://t.co/VgtnJ1LKRM
— MatthewBerman (@MatthewBerman) Aug 9, 2024
After almost indistinguishable photorealistic FLUX images, checkout these lip-sync demos from ReSyncer. We've truly arrived in the post-truth era. https://t.co/rqe5bo8Saw https://t.co/RMinlGoSrs
— Dreaming Tulpa 🥓👑 (@dreamingtulpa) Aug 9, 2024
wow. an amazing 325 page google strategy document quietly unsealed buried in google antitrust docket. It's gonna take a long thread but I have pulled out the gems. It's from 2017 planning, no doubt Google will just say these were only ideas but many will look very familiar. /1 https://t.co/aQzSlKQQlU
— Jason Kint (@jason_kint) Aug 8, 2024
Tried to get a similar vibe out of Midjourney with the prompt: a young woman smiling while speaking onstage from google, white background with corporate logos blurred out, tech conference --style raw --v 6.1 Result is here https://t.co/qkStf6wWTC https://t.co/jZWVyM2dLu
— Kiri (@Kyrannio) Aug 8, 2024
42 page PDF (it says 30k tokens) that contains both images and text thrown into gemini-flash 1.5, gets every answer correct... it's so over
— anton (@abacaj) Aug 8, 2024
First time an AI image really struck me as uncanny. Not my prompt, this is an AI image generated using Flux and a Lora combo, I believe by Leo Kadieff. (He had posted it, at least) https://t.co/XL7SlVNiD9
— Kiri (@Kyrannio) Aug 8, 2024
So many AI agent startups to keep track of. These raised capital recently: Emergence - $97M (multi) Ema - $36M (multi) Orby AI - $30M (web) Dust - $16M (web) Tezi - $9M (recruitment) Norm Ai - $27M (compliance) Leya - $25M (legal) LegalFly - $16M (legal) Didero - $7M… https://t.co/ncKP6M5z7H
— Chief AI Officer (@chiefaioffice) Aug 7, 2024
"When agents can do a task, they do so at ~1/30th of the cost of the median hourly wage of a US bachelor’s degree holder" AI agents will eat the world task by task THEN job by job. What's next? PREDICTION: Skeptics, suffering from Exponential Slope Blindness, will dismiss AI… https://t.co/4Pk1xmNgju https://t.co/BuVmvfDv0F https://t.co/7BI6BLbbBY
— AI Notkilleveryoneism Memes ⏸️ (@AISafetyMemes) Aug 7, 2024
How do you get your genAI prototype enterprise ready? @Pinecone CEO @EdoLiberty and @LangChainAI CEO @hwchase17 discuss what to build or buy, improving off-the-shelf solutions with great data, and how to pick the right partners to scale: https://t.co/42gHrSLuH1
— a16z (@a16z) Aug 7, 2024
Web scrapers are insane. $249 for 100k results. I wrote a 578 line node script and used 50 dedicated proxies to grab 40 million results in 24 hours. The IP addresses are $2/month/proxy. The data is $0. Y'all should learn to code. https://t.co/oFtWM13yhL
— Jordan O'Connor (@jdnoc) Aug 6, 2024
A Primer to Building Multi-agents as a Service 📖🤖 A lot of our users have asked for more comprehensive documentation around llama-agents - our core repo around building multi-agents as a service. It’s advanced significantly in the past few weeks thanks to @_nerdai_ . This… https://t.co/k0TEeMi3C5 https://t.co/8uGgjnMcps
— LlamaIndex 🦙 (@llama_index) Aug 7, 2024
🚨 CHINA UNVEILS ‘AMPHIBIOUS’ UNDERWATER DRONE This innovative drone can be launched underwater, has fixed wings for vertical takeoff and landing, and can land on its tail. Source: RT https://t.co/CZ6iRcPdh7 https://t.co/rZuuuEhBGC
— Mario Nawfal (@MarioNawfal) Aug 7, 2024
How are modern AI platforms architected? "Building A Generative AI Platform” outlines the common components of building a generative AI platform and explains their functions and implementation.👀 TL;DR: 🧠 Context enhancement: RAG, query rewriting, and agentic systems 🛡️… https://t.co/NnNHFRVvD9 https://t.co/HU5Pdmo5n9
— Philipp Schmid (@_philschmid) Aug 7, 2024
Friendly Google and Enemy Remedies The DOJ brought the right kind of case against an Aggregator, which stagnates by being too nice; the goal is for companies to act like they actually have enemies. https://t.co/cJuZKl8GM9
— Stratechery (@stratechery) Aug 6, 2024
The Prestige Limitless Black Swan Jiro dreams of sushi The Novice The Founder Kill Bill I + II Foxcatcher Oppenheimer The Aviator Steve Jobs Social Network Molly's Game Pi Adaptation Burnt Ferrari + Ford vs Ferrari Fight Club Batman - Nolan trilogy Mad Max - all films https://t.co/U75c3cXLCo
— Zach 🏴 (@zachpogrob) Aug 5, 2024
AI is changing the ads game, fast. Put these tools on your radar ASAP: https://t.co/S15YJiOC2z
— Boring Ads (@imboringads) Aug 5, 2024
BIG AI News this week, here's everything you need to know: 0:00 - Figure 02 Robot 1:51 - Groq Raises Series D 4:12 - ChatGPT Cheating Detection 6:20 - Google Acquires https://t.co/VwHeq2pgue 8:24 - Runway "Stealing" Videos 9:57 - Ollama Tool Usage 10:55 - Apple Intelligence… https://t.co/OIay0Qu6D8 https://t.co/mVWUxFzC32
— MatthewBerman (@MatthewBerman) Aug 6, 2024
Observe, evaluate and optimize your RAG stack! Observability and evaluation are always hot topics in Llama-land, and this step-by-step tutorial will show you how to use @ArizePhoenix and LlamaTrace as well as @Azure AI Search to build a RAG app, debug and optimize it.… https://t.co/XdUzFRygoK https://t.co/XHyF1q3FNt
— LlamaIndex 🦙 (@llama_index) Aug 5, 2024
CodeGPT AI extension in Visual Studio can quickly fix and refactor your Python Code using Llama-3 running locally on your computer (100% free and without internet): https://t.co/9u2TTYkKGU
— Shubham Saboo (@Saboo_Shubham_) Aug 6, 2024
$500M+ raised by AI startups last week Here's 7 of them VCs backed you should know about:
— Chief AI Officer (@chiefaioffice) Aug 5, 2024
This is the best AI-generated movie trailer I have personally done! I am happy to share my latest AI trailer, "BloodSpace," based on a feature film I am currently developing. Make sure you watch til the end! I had to sneak in at least one shot with @runwayml's new last-frame… https://t.co/qqVTnYiaLC https://t.co/gewxu3ey0p
— Dave Clark (@Diesol) Aug 5, 2024
Flux's Architecture diagram :) Don't think there's a paper so had a quick look through their code. Might be useful for understanding current Diffusion architectures https://t.co/eO2e5tYbpl
— wh (@nrehiew_) Aug 5, 2024
Has AI killed SaaS? ☠️ @zebulgar “People love software because the marginal distribution costs are zero. What people need to realize is that the marginal returns are zero too. 2 Stanford kids with today's AI can replicate your SaaS company in a weekend.” Love to hear your… https://t.co/mxNJboIgqj https://t.co/ZLIG0HVGDe
— Harry Stebbings (@HarryStebbings) Aug 4, 2024
there are businesswomen making 8 figures on etsy selling gold plated jewelry and you think you're coming out on top with your 700k quant cash salary at jane street 😂🫵😳
— kache (@yacineMTB) Aug 3, 2024
Data products are what is going to elevate data engineering into the stratosphere! They power everything you could imagine in the big tech companies! - At Airbnb, I worked on a data product that helped detect "bad hosts" to increase guest satisfaction - At Netflix, I worked… https://t.co/yvXVqazYgu https://t.co/Enb6mk9JMX
— Zach Wilson (@EcZachly) Aug 3, 2024
I'm blown away by @gumloop_ai. Someone on reddit wanted help finding general emails from a text list of hotel names. I'm used to using Make, so I thought this would take a minute. But I guess I was training on hard more, because in gumloop I built it in seconds! Sick product! https://t.co/EGc068XC5P
— Akili (@akilikajunju) Aug 3, 2024
I stopped writing code entirely in 2024. I only copy-paste code generated by AI ✌️🤓 How do I do that? 10 hints: Hint 1: if you have a creative task such as code architecture, you want to use so called chain of thoughts. You add "Think step-by-step" to your prompt and enjoy a… https://t.co/u89MlbCdgP https://t.co/cVTTGBKFWs
— Alexander Isora 👽 (@alexanderisorax) Aug 3, 2024
LangChain Introduces LangGraph Studio: The First Agent IDE for Visualizing, Interacting with, and Debugging Complex Agentic Applications Read our take on this: https://t.co/mdTtk7jZsv Try out LangGraph Studio for free here: https://t.co/gJmnDiqzCZ @LangChainAI https://t.co/2G55fJFxDf
— Marktechpost AI Research News ⚡ (@Marktechpost) Aug 4, 2024
Drag & Drop to build AI Agents. LangGraph Studio is a visual graph IDE to build LLM apps with AI agents. It lets you visualize, interact, and debug complex AI agentic workflows to build LLM & RAG apps. https://t.co/oTVMpUxbmR
— Unwind AI (@_unwind_ai) Aug 4, 2024
Just when you thought it wasn’t possible to make yet another major advance — the latest SpaceX Raptor iteration. Additive manufacturing internally integrates so many formerly discrete components… Taking “no part is the best part” to the extreme. 🌎 🚀 🔴 https://t.co/AC5DZIw8iB
— Steve Jurvetson (@FutureJurvetson) Aug 3, 2024