The year 2024 is coming to an end and it was a big year in AI. I wrote a long review of the year earlier. However, with OpenAI's o1, I have to admit that OpenAI has remained number 1 for me. I love o1. How is it with you? Do you prefer Claude sonnet 3.5? Or rather Llama models?… https://t.co/COlST5V0G6
— Chubby♨️ (@kimmonismus) Dec 31, 2024
PSA: our experimental Gemini models are free (in Google AI Studio and API), 10 RPM, 4M TPM, 1500 RPD. Enjoy the most powerful models we have to offer (2.0 flash, thinking, 1206, etc), with just 3 clicks on: https://t.co/fBrh6UGKz7
— Logan Kilpatrick (@OfficialLoganK) Dec 31, 2024
Stanford has launched an incredible research AI tool. It’s called STORM, and basically you enter a topic and it will search hundreds of websites to write an article about its major findings. Available to everyone for free! https://t.co/lMge85hNpP
— Alvaro Cintas (@dr_cintas) Dec 31, 2024
Gemini 1206 keeps first place overall. Makes me even more curious for Gemini 2.0 Pro https://t.co/GLoOK7PEd4
— Chubby♨️ (@kimmonismus) Dec 30, 2024
v2 of @ai16zdao eliza will focus on: - better overall agent performance - better plugin management - improved code stability and readability - cleaner agent architecture excited to see @chrislatorres lead this project and will help out wherever I can
— cygaar (@0xCygaar) Dec 30, 2024
“Every advance in the complexity of the economy puts an added premium upon superior ability.” https://t.co/LLCbwfgvMK
— David Senra (@FoundersPodcast) Dec 31, 2024
So, it begins. The era of Autonomous AI agents has started. An open-source agent-building framework is now trending on github. This is not an event on its own but another marker of the attention shift 👀 https://t.co/s9JdGNwnnf https://t.co/9ywHGK1fuy
— TestingCatalog News 🗞 (@testingcatalog) Dec 30, 2024
it's fucking based that the fastest growing, most accessible, most flexible ai agent framework was launched by a crypto-native team w crypto-native features the overlap we're seeing between crypto x ai is insane. Stanford x Eliza was just the start, we are working to 100x this
— nader dabit (@dabit3) Dec 31, 2024
So I'll ask the obvious question, now that it's a $1B+ token: does @ai16zdao not violate a16z's trademark? It felt originally like a parody but now seems like it's become a more legitimate project with a more aggressive roadmap. Any IP lawyers in the house?
— Haseeb ＞|＜ (@hosseeb) Dec 31, 2024
Super exciting talk by @JeffDean https://t.co/7BlrGP2bNO
— yobibyte (@y0b1byte) Dec 30, 2024
Wake up babe @ahrefs just dropped a new real-time, privacy-first web analytics product... ...with a killer pricing: track 1M events/mo for free, and up to 10M events/mo for just $20/mo 🤯 https://t.co/NKJO4e8GDx
— Steven Tey (@steventey) Dec 29, 2024
These were the investing themes of 2024: • Chips • AI Infrastructure • Software • Nuclear • FinTech • Crypto • Quantum • Space • Air Taxis • Self-driving Cars • Drones I expect most of them to continue into 2025, plus: • AI Agents • Robotics • Wearables • IPOs https://t.co/CFa2QhHKm0
— Lin (@Speculator_io) Dec 29, 2024
llama2.c in pure C from @karpathy was the only OSS project I could find that would enable running an LLM on the Windows 98 Pentium II machine. I couldn't get modern cpp (e.g. llama.cpp) to run. Everything else would almost require starting from scratch. Thank you @karpathy! https://t.co/aFI7rhLWoL https://t.co/rx5Dnasv0N
— Alex Cheema - e/acc (@alexocheema) Dec 29, 2024
AI predictions. What will 2025 bring to us? Ollie Carmichael, Product Manager at @AdargaAI, highlighted some very interesting aspects: 1. Disruptions: “Healthcare and specifically personalized medicine will experience the most disruption. There will be a drop in the cost and… https://t.co/LPBYlEeN9S https://t.co/uPi8zgm6sa
— TuringPost #CES2025 (@TheTuringPost) Dec 28, 2024
Should I build this? Total cost is around 1.2k$ and has a 4h flight time. https://t.co/ZAGrBM5glB
— Lennard (@lennardeth) Dec 29, 2024
30 Books Everyone Should Read in 2025 https://t.co/emjQL8SYEC
— Wealth Hatch🧠📈 (@WealthHatch_) Dec 28, 2024
The opportunities for AI agents in different markets. They are coming for your job, fast. https://t.co/W1dGqqTs3K
— Angry Tom (@AngryTomtweets) Dec 29, 2024
WHAT FEATURES DID X ADD IN 2024? LET'S TAKE A LOOK! https://t.co/CxfZUjVIVB
— Nima Owji (@nima_owji) Dec 27, 2024
2024 State of LLMs for Text2SQL Tasks 🏆- Full Report 🥇 Overall Performance: @GoogleDeepMind Gemini-Exp-1206 🥇 Open Source Model: @Alibaba_Qwen 2.5-Coder:32b (Beats Sonnet 3.5 and on par with GPT-4o!) Disappointing performance by GPT-4o and 3.5 Sonnet on this task. 🧵 https://t.co/wIKqadZqFG
— Subhash Peshwa (@Subhash_Peshwa) Dec 18, 2024
Unfortunately, the modern text embedding model is a "bag of semantic tokens". We built up a small taxonomy: directional,Temporal,Casual,Comparative and Negation to demonstrate when text embedding models fails and how to mitigate the issue with fine-tuning. https://t.co/FwJWsaOfql https://t.co/f2wcQ4PvaL https://t.co/xtrHSnAnrp
— Bo (@bo_wangbo) Dec 18, 2024
We’re excited to announce the tokenization of the newest @KhoslaVentures investment: Basis AI. Contract Address: 6sNVkGqosDnjDXGS4euhW87e18y5qxAfas1LSrDpump With $BASIS, you can access AI agents built specifically for assist with investments, funds, and portfolio management.… https://t.co/5B5363Y1Ld https://t.co/D0PQ1BHjtM
— Vinod Khosla (@vkhosla) Dec 28, 2024
Unveiling DeepSeek - interview with the founder Liang Wenfeng - h/t @wzihanw " There’s no mystery. Our team is composed of recent graduates from top universities, Ph.D. interns in their fourth or fifth year, and a few young professionals just a few years into their careers.… https://t.co/5wivthul1d
— Anu Hariharan (@anuhariharan) Dec 28, 2024
I am 🤯amazed by his writing skill -such a really great article for beginners -soon articles like this will come, love his passion for writing -great article to understand LLM's https://t.co/IFjwkqSarN
— Abhishek (e/acc)🌱 (@Abhishekcur) Dec 28, 2024
Chart AI is only 2 months old > 20K downloads & $5K MRR > 13,788,000 views They're copy & pasting the AI Study Tool TikTok playbook every winning app is using: https://t.co/SCQdZU7a9N
— Maria Inês (@themariaines) Dec 27, 2024
Book Summary: “T-Minus AI” https://t.co/onlRsKWAZK
— MindBranches (@MindBranches) Dec 27, 2024
What is added to 𝕏 in 2025 will be even more https://t.co/BIBAdJnuns
— Elon Musk (@elonmusk) Dec 27, 2024
Summary of 12 days of OpenAI Still missing Tasks and Operator https://t.co/s4sg2PwlE8
— Chubby♨️ (@kimmonismus) Dec 27, 2024
25 books recommended by Walter Isaacson: 1) Measure What Matters by John Doerr https://t.co/Kqoxb02gu7 https://t.co/bMUmll6Hkl
— Reads with Ravi (@readswithravi) Dec 27, 2024
As I have often made clear, I believe it is important to invest in the stock market for the long term. And that's exactly what I've been thinking about recently. Reasoning models in particular are excellent for 1) gaining an overview of the market and 2) analyzing individual… https://t.co/VKA4cf8Zpu https://t.co/Y9wdbC9wy1
— Chubby♨️ (@kimmonismus) Dec 26, 2024
Some notes on the DeepSeek-V3 Technical Report :) The most insane thing to me: The whole training only cost $5.576 million or ~55 days on a 2048xH800 cluster. This is TINY compared to the Llama, GPT or Claude training runs. - 671B MoE with 37B activate params - DeepSeek MoE… https://t.co/8SkhYsjf6H https://t.co/kmDDHcIN3X
— Lisan al Gaib (@scaling01) Dec 26, 2024
How to train a 670B parameter model. Let's talk about the DeepSeek v3 report + some comparisons with what Meta did with Llama 405B https://t.co/Hv8Q5Q37A3
— wh (@nrehiew_) Dec 26, 2024
Craziest thing is it took only $5.5m to train. US labs spend one — maybe two — order of magnitude more for frontier models. https://t.co/50Ko3KY6cn
— Amjad Masad (@amasad) Dec 26, 2024
The official release of the Whale, Deepseek V3 has arrived. https://t.co/qgtrzLWvSE https://t.co/oHFkHENFI8
— Andrew Curran (@AndrewCurran_) Dec 26, 2024
How crazy is it to think that Deepseek is offering Deepseek V3 for 0.28/1M output and that this model is theoretically the second best non-reasoning model (second only to Gemini 1206) in the benchmarks. https://t.co/1hwjztPzYK
— 𝑨𝒓𝒕𝒊𝒇𝒊𝒄𝒊𝒂𝒍 𝑮𝒖𝒚 (@artificialguybr) Dec 26, 2024
There’s cities, there’s metropolises, and then there’s Tokyo. https://t.co/epaX2CE6hA
— non aesthetic things (@PicturesFoIder) Dec 25, 2024
Ranking #1 on @Google is NOT the goal. High-intent traffic is. Once you understand how it works, growing companies becomes easy. Here's the exact SEO playbook we used to grow FreeUp from 0 to $12M/year + exit. Grab it while it's free 👇
— Connor Gillivan (@ConnorGillivan) Nov 19, 2024
I liked this a lot! In the same spirit, I also really liked @natolambert, @kylelostat and @AkshitaB93's tutorial at NeurIPS on building language models. Slides: https://t.co/b1SQh8a4t2 Tutorial: https://t.co/3oPuJtEfHj (I think it will be released publicly soon as well) https://t.co/u5jGfY3N5p
— Niloofar (on faculty job market!) (@niloofar_mire) Dec 25, 2024
Through experimentation @LinkedIn found EON-8B, a domain-adapted version of Llama 3.1 8B, to be 75x and 6x cost effective in comparison to GPT-4 and GPT-4o respectively. More on their domain-adapted foundation model work ➡️ https://t.co/mZGuiZaxCY https://t.co/gw79k8W3fp
— AI at Meta (@AIatMeta) Dec 23, 2024
https://t.co/7PB6RT9wCW
— 0xJeff (@Defi0xJeff) Dec 25, 2024
Human quality transcripts using Gemini 2.0 w Audio. Lmk if you come up with a prompt that works better! Blame Claude for all coding mistakes. https://t.co/ctq7EKoSty https://t.co/L1QRSgCGEs
— Dwarkesh Patel (@dwarkesh_sp) Dec 24, 2024
The role of a software engineer is transforming from being the author of computer code to being the operator of a code generating machine. What is a computer programming system built natively for that workflow? https://t.co/ePRFvvuVDu
— Bret Taylor (@btaylor) Dec 24, 2024
Replit Agent has all these connections via pre-defined prompts https://t.co/0SixhyUsTv
— Melvin Vivas (@donvito) Dec 24, 2024
Created a simple AI chat app demo using @Replit agent. I added the OpenAI connection and it asked for my OpenAI api key. Took around 30mins with OpenAI integration and Postgres database where messages are saved. https://t.co/tw7ZCc7gvK
— Melvin Vivas (@donvito) Dec 24, 2024
ai ugc creators never sleep, never miss deadlines, pump out 500 vids an hour and cost less than your morning coffee brands are about to go crazy https://t.co/Dh5JU6qugr
— Matt Welter (@_mattwelter) Dec 24, 2024
I thought we were done for big new model releases in 2024, but evidently not! Here's a new Apache 2 licensed vision LLM from Qwen to round out the year https://t.co/vchEM3wOhj
— Simon Willison (@simonw) Dec 24, 2024
150GB of math dataset, holy shit https://t.co/cXbpR1LAtV
— maharshi (@mrsiipa) Dec 24, 2024
Build a full-stack contract review agent in a single line of code 🤖🧑⚖️ This app template by @MarcusSchiesser uses @getreflex with @llama_index workflows to build a contract review agent that can check GDPR compliance given an input vendor agreement. Builds upon a core agentic… https://t.co/cIkFcQSod8 https://t.co/q9DcVIgAKu https://t.co/3d5qsSSpxf
— LlamaIndex 🦙 (@llama_index) Dec 24, 2024
Whoa! In @NEJM_AI: A Multimodal Biomedical Foundation Model Trained from 15 Million Image–Text Pairs… fully open-access foundation model! https://t.co/ZC00ai1Cr9
— Atul Butte (@atulbutte) Dec 23, 2024
Where the AI Agent Space Stands Right Now PT.5 As we approach the holiday season, the market is starting to slow down. After the recent FOMC-induced dump, major ecosystem tokens like $VIRTUAL and $AI16Z are beginning to recover. $VIRTUAL in particular is climbing back toward a… https://t.co/YbtX8PDaD3 https://t.co/fhPqPLICHK
— 0xJeff (@Defi0xJeff) Dec 23, 2024
Cyberpunk sci-fi at its best ! https://t.co/zSvIaTX1ru
— Ivanka Trump (@IvankaTrump) Dec 23, 2024
OpenAI must have had a significant head start in understanding and applying the inference scaling law. One of the recent paper in this regard was "Large Language Monkeys: Scaling Inference Compute with Repeated Sampling" 👨🔧 The paper explores scaling inference compute by… https://t.co/HAvfLs9zZ3 https://t.co/j1NAzz4EtB
— Rohan Paul (@rohanpaul_ai) Dec 23, 2024
A huge promise of agentic document workflows is the fact that they make domain-specific business logic much easier to implement, instead of hand-crafted custom logic or models around document parsing/extraction, standardization, and retrieval. Take our latest example by… https://t.co/WVCfruURYc https://t.co/5Bx1hwu6Y1 https://t.co/V86p2oT3Th
— Jerry Liu (@jerryjliu0) Dec 23, 2024
It's so over. o1-pro + @lovable_dev can do anything. I just built a Pendulum Wave Sketcher by prompting o1-pro and then in two prompts integrated it into a web app with lovable. Then deployed it with one click lol. I wonder what o3 will be able to do. https://t.co/HqLW42JObz
— Konstantin Zolozhkov (@KHvostikk) Dec 22, 2024
Looking for AI products to try over the break? From the @a16z team + friends of the firm ⬇️ https://t.co/KKDUHPHnhx
— Justine Moore (@venturetwins) Dec 22, 2024
Just wrapped an amazing conversation with @lennysan! We dove into everything from staying curious with a beginner’s mind to lessons from Steve Jobs, navigating tough times, and why AI agents are the future for Salesforce. YouTube: https://t.co/vwE0ZLmBDm Spotify:… https://t.co/vq7DBtOxgO
— Marc Benioff (@Benioff) Dec 22, 2024
I've listened to this podcast three times and have read all or part of the transcript several times. My restatement of one part of the podcast: https://t.co/cYN7obwCZv https://t.co/z8tp8NUE1W
— Tren Griffin (@trengriffin) Dec 22, 2024
15 agentic systems and frameworks of 2024: ▪️ GUI Agents ▪️ Agent K v1.0 ▪️ The AI Scientist ▪️ MALT ▪️ Agent S ▪️ ADAS ▪️ AgentInstruct ▪️ AgentStore ▪️ WALL-E ▪️ Generative Agent Simulations of 1,000 People ▪️ DynaSaur ▪️ PRefLexOR ▪️ Genex ▪️ Bel Esprit ▪️ AutoKaggle Save… https://t.co/6h1FmPJwfN https://t.co/nid60o0bID
— TuringPost #CES2025 (@TheTuringPost) Dec 22, 2024
Nice! LLM consortium. Why ask one AI when you can ask all of them and have them come to a consensus? Someone plot the new scaling laws of number of LLMs on x axis :) This one is built on top of @simonw llm CLI. https://t.co/n9xGx4vjeQ https://t.co/xqwIo5CpH5
— Andrej Karpathy (@karpathy) Dec 22, 2024
12 Days of OpenAI is now over, and it was INSANE This will change the way we use technology forever. Here're 12 OpenAI announcements you don't want to miss: https://t.co/9nUYNf5jSd
— Poonam Soni (@CodeByPoonam) Dec 21, 2024
A thread of accounts I've found most interesting: @robertskmiles @AlecStapp @waitbutwhy @skdh @robinhanson @amasad @cremieuxrecueil @bscholl @tylercowen @MaxCRoser @ahistoryinart @DKThomp @levie @ryangrim @scienceisstrat1 @CuriosMuseum
— Paul Graham (@paulg) Dec 21, 2024
Happy Shipmas 2024 to you and your families. https://t.co/ylpQyuqD84
— Adam.GPT (@TheRealAdamG) Dec 21, 2024
welcome https://t.co/eYsB4RtTUk
— 🍓🍓🍓 (@iruletheworldmo) Dec 20, 2024
This is huge! Right before the end of the year, @OpenAI unveils it's biggest the o3 model, setting a new standard for reasoning with new tasks. o3 creates its own step-by-step solutions (CoTs) to tackle novel challenges. (We're waiting for the preview to explore its full… https://t.co/YSkXdNlFIb https://t.co/C3yIbmExYu https://t.co/UBdaa6bHFh
— TuringPost #CES2025 (@TheTuringPost) Dec 20, 2024
Artificial Intelligence inverts the current technological stack based on modularity to keep components coherent. Hence, the app ecosystem that you see today is going to disintegrate! The monolithic services are also going to be invisible. Everything will be defragmented.… https://t.co/TfvUryj7E0
— Carlos E. Perez (@IntuitMachine) Dec 20, 2024
I fell in love with Grok's new logo https://t.co/j3u4hvJeXT
— Dogan Ural (@doganuraldesign) Dec 20, 2024
We now live in a different world. Acceleration is imminent. You *will* need to adjust your worldview. This is what the early days of the singularity look like. And you are living through them. https://t.co/c9pDnUUg7g
— Mckay Wrigley (@mckaywrigley) Dec 20, 2024
OpenAI announces a new frontier model called o3! Includes o3 and o3-mini. These models are not publicly available yet. 1/n https://t.co/wye1RBJPQe
— elvis (@omarsar0) Dec 20, 2024
Instead of predictions for 2025, we framed the discussion as big themes to watch/key debates going into 2025. Enjoy! Full video on YouTube. https://t.co/hoVEqFpBDy https://t.co/qmgAxYydMD
— Ben Bajarin (@BenBajarin) Dec 20, 2024
1/ No one likes to do mundane work, and no one should have to. We @a16z believe that intelligent automation with AI agents will be able to handle what used to be done manually or through brittle RPA solutions, freeing workers to focus on more strategic work. I broke this down in… https://t.co/TzKadQ60J1 https://t.co/mG6fBCOKz5
— Kimberly Tan (@kimberlywtan) Dec 20, 2024
This Lex interview with @DarioAmodei is worth your time if you want to get a deep sense of where AI is heading https://t.co/h4aqjc7h4v
— Lenny Rachitsky (@lennysan) Dec 20, 2024
If you read nothing else about AI this week, read this - absolutely fascinating, ties together a whole bunch of recent developments in the field Wrote my own notes here but really this piece defies summarization, you should absorb the whole thing https://t.co/otBVrZHFWU https://t.co/PgpWUO8f1F
— Simon Willison (@simonw) Dec 19, 2024
How old are your lungs 🫁? Hold your breathe to find out. Mine are 30. https://t.co/3dWw8bqEco
— 🇺🇸𝓒𝓱𝓪𝓭🇺🇸𝓗𝓪𝓻𝓻𝓲𝓼𝓸𝓷🇺🇸 (@Texas_jeep__guy) Dec 19, 2024
Introducing o1 Web Extractor ⚡️ Ask any information about a company then it will search the web and extract structured data for you using Firecrawl’s new /extract endpoint. Built with OpenAI’s new o1 structured outputs API and @firecrawl_dev. Check it out: https://t.co/WnpcJzlF05
— Eric Ciarla (@ericciarla) Dec 19, 2024
An SEO-focused agent using AgentStack that fetches the top-performing keywords used for ranking websites. The agent analyzes competitors’ content to extract and list keywords that contribute to their SEO performance, providing insights for website optimization. By @SuneHash https://t.co/AsnmYSU5zE
— Adam Silverman (Hiring!) 🖇️ (@AtomSilverman) Dec 19, 2024
The 2024 Stratechery Year in Review https://t.co/GbXh5p7E5k
— Stratechery (@stratechery) Dec 19, 2024
Looks like Tasks-feature and automations will be shown and maybe even released today. Imho it makes much sense to give a preview of agents (operator) aswell, first perfect https://t.co/Q739roZ1Jr
— Chubby♨️ (@kimmonismus) Dec 18, 2024
✨ How We Built It: https://t.co/ZEcfJyzHX0 ✨ In this video, James Groeneveld (Engineering at https://t.co/ZEcfJyzHX0) and Harrison Chase (LangChain CEO) take you behind the scenes of https://t.co/ZEcfJyzHX0’s journey to scale LLM generation to handle 30,000 queries per second.… https://t.co/GTwCGHXUZY https://t.co/YuRYMNnhrd
— LangChain (@LangChainAI) Dec 19, 2024
The one and only: https://t.co/E2NLOcc0YB
— ₕₐₘₚₜₒₙ — e/acc (@hamptonism) Dec 18, 2024
84 Films Recommended by Martin Scorsese: ‘The Small Back Room’ (1949) ‘Pearl’ (2022) ‘Maxxxine’ (2024) ‘Last Year at Marienbad’ (1961) ‘Zulu’ (1964) ‘Lawrence of Arabia’ (1962) ‘Lola Montés’ (1955) ‘Once Upon a Time in the West’ (1968) ‘The Round-Up’ (1966) ‘Chimes at Midnight’… https://t.co/NwsvMxe4bE https://t.co/iiqWLPjMNC
— DepressedBergman (@DannyDrinksWine) Dec 18, 2024
Anthropic just dropped an insane new paper. AI models can "fake alignment" - pretending to follow training rules during training but reverting to their original behaviors when deployed! Here's everything you need to know: 🧵 https://t.co/hXH0G28qqx
— MatthewBerman (@MatthewBerman) Dec 18, 2024
The opportunity for AI agents is gigantic. We are in an industry of skills and AI agents. For the very first time we have AI agents (boundless digital labor) using our tools. —Nvidia CEO Jensen Huang and @Benioff on autofacturing intelligence and work at scale with AI agents https://t.co/C6n9KJ2W1r
— Vala Afshar (@ValaAfshar) Dec 17, 2024
🧬 Talk to Molecules! Built a voice-controlled 3D molecular explorer with OpenAI real-time voice and function calling 🥳 link to code in comment 🟣 FastAPI for backend 🟢 WebRTC for real-time communication 🔵 Three.js for 3D visualization 🟡 Chart.js for data visualization 🟠… https://t.co/mQikftLdEZ https://t.co/4NdYw5rA1v
— echo.hive (@hive_echo) Dec 18, 2024
2024 in review: - The rise of VLMs for retrieval (ColPali) - The year of late-interaction (Col) - (V)LMs as rankers - Token cost race to zero + context cache - RAG is dead, long live RAG - Vector db as a category died - The rise of BM25 - BQ Binarization rediscovered
— Jo Kristian Bergum (@jobergum) Dec 18, 2024
AI agents can now buy their own compute to self-improve and become self-sufficient Yeah, we are already that far in. via reddit. https://t.co/XmhILl202K
— Chubby♨️ (@kimmonismus) Dec 16, 2024
Accounting is a massive industry, and Basis is clearly leading on the AI side. This is one of the few AI agents that’s already deployed and working. Matt & Mitch have put together the best NYC team in the applied AI space. https://t.co/I4IHzOQoeY
— Vinod Khosla (@vkhosla) Dec 17, 2024
Top 7 Insights from AI Agents in Production from CrewAI users, here is a thread 🧵
— João Moura (@joaomdmoura) Dec 17, 2024
Flying to USA 🇺🇸 - got some time to read DeepSeek v1 & v2 papers, here are the key takeaways: First I'll just say that they're an incredibly cracked Chinese team - I'd put them somewhere between Mistral and OpenAI/DeepMind teams in terms of their prowess / capability to build… https://t.co/SW2rNElaAS https://t.co/9AmDuvykCS
— Aleksa Gordić 🍿🤖 (@gordic_aleksa) Dec 17, 2024
AI Agents In Production! https://t.co/10SeugnpDB RT Please? 🙏
— João Moura (@joaomdmoura) Dec 17, 2024
Nice use case of Gemini https://t.co/b1fCR8MxAD
— elvis (@omarsar0) Dec 17, 2024
Excited to feature a new tutorial 💫 on auto-generating a case summary from semi-structured patient data 📝🏥. This allows a clinician to automatically review a patient's profile without spending a ton of time digging through info or defining cumbersome matching rules. 1️⃣ Map… https://t.co/nr313uD9PR https://t.co/WGvxuoKPBI https://t.co/aaK9C6YbuH
— Jerry Liu (@jerryjliu0) Dec 16, 2024
Founder of MidJourney https://t.co/D0yt1zjLIt
— Sar Haribhakti (@sarthakgh) Dec 17, 2024
This is the kind of book everyone will read. And we’ll be better off because of it. https://t.co/QrCz6mZdDB
— Ted Mabrey (@MabreyTed) Dec 17, 2024
🚨 $PLTR CEO Dr. Alex Karp’s new book - The Technological Republic📖🔮🇺🇸 Stanley Drunkenmiller: “Alex does not need the revenues with his stock at $73 — it’s one of the most important books I’ve ever read… It had a big impact on me.‼️ “It starts with footsie, pagan religion,… https://t.co/MhDPg4WsmS https://t.co/ezzVLWGU8I
— Jawwwn (@jawwwn_) Dec 17, 2024
This is so incredibly cool. Turning a 400-page book into an interactive detective game with a 400-word prompt and a large language model. If you’re still treating these tools like search engines, you’re missing out on what’s possible. I wrote more here: https://t.co/dAUi7USCRp https://t.co/VMe2d3f2lu
— Matt Stockton (@mstockton) Dec 16, 2024
Perplexity Pro vs McKinsey is the actual game. Get you an analyst for orders of magnitude less spend and arguably better quality research. https://t.co/AORCXDb4NC
— Aravind Srinivas (@AravSrinivas) Dec 15, 2024
4 timeless fonts I swear by: What fonts do you keep coming back to? Drop them below ↓↓↓ https://t.co/NhaltyG0r3
— Hewar (@hewarsaber) Dec 15, 2024
$ASML Quite fresh from the press. Looks interesting. https://t.co/KUcxhypCsy
— Edelweiss Capital (@Edelweiss_Cap) Dec 15, 2024
Your daily reminder that AI is gonna replace your job. In this video, Klarna CEO Sebastian Siemeatkowski explains how they use OpenAI to replace around 20% of employees with AI every year. And that's just the beginning. An hour ago I posted the interview with OpenAI's CFO from… https://t.co/3q5YplbmkK https://t.co/Tx8cFAFNmN
— Chubby♨️ (@kimmonismus) Dec 15, 2024
Can your Figma do this? 🦄 https://t.co/uur3B6DDDi
— Voicu Apostol (@cerpow) Dec 15, 2024
Magnificent website from the Japanese Ichiban Bike. https://t.co/hhbDcFjD7s
— Darius Dan (@idariusdan) Dec 15, 2024
🚨This week’s top AI/ML research papers: - Byte Latent Transformer - Training Large Language Models to Reason in a Continuous Latent Space - Language Modeling in a Sentence Representation Space - Phi-4 Technical Report - Best-of-N Jailbreaking - Forking Paths in Neural Text… https://t.co/qogFNJ3zkj https://t.co/teg96v5muU
— The AI Timeline (@TheAITimeline) Dec 15, 2024
GOOGLE BREAKS THE AI PRICING MODEL (AGAIN) 🚨 Same data but without logarithmic x-axis to show you how ridiculous GPT-4o and o1-preview pricing currently looks! The other black and orangy points above 10$ are Claude-2, GPT-4 Turbo, o1-preview, Claude-3 Opus and the original… https://t.co/7DxGMOpf6I https://t.co/WH3qxjkB0k
— Lisan al Gaib (@scaling01) Dec 14, 2024
@APompliano This was the best explanation of the “threat” of quantum computers against BTC I’ve watched. https://t.co/UEaI6zyqfZ
— Bitcoin Dad (@bitcoin_dad) Dec 14, 2024
A Japanese research team has developed a drug that can regrow human teeth! https://t.co/Upx2cGoXzS
— Interesting As Fuck (@interesting_aIl) Dec 14, 2024
Ilya Sutskever: "Sequence to Sequence Learning with Neural Networks at NeurIPS 2024" Link to the video in the next tweet. https://t.co/iqbxsgkwzk
— Tren Griffin (@trengriffin) Dec 14, 2024
Ilya Sutskever's “Test of Time” talk at NeurIPS 2024 Evolution of Deep Learning: The talk charts the incredible progress in deep learning over the past decade, reflecting on the core principles of the 2014 work and how they paved the way for today’s landscape of large language… https://t.co/2h1XCVsyBC
— Carlos E. Perez (@IntuitMachine) Dec 14, 2024
Satya Nadela explains the AI Agentic Future. The business logic is all going to these Agents. ----- Video from Bg2 Pod Youtube Channel (link in comment) https://t.co/CCzcvPvmZ5
— Rohan Paul (@rohanpaul_ai) Dec 14, 2024
Agents are the perfect tremor for the earthquake coming to SaaS pricing.. https://t.co/hqz1vDKBX5
— Nikunj Kothari (@nikunj) Dec 13, 2024
I don't think people understand how incredible that is: A 14b SLM beats the much larger Llama 3.3 70b almost everywhere. Wtf AGI is going to run local on the edge https://t.co/zRedGE5Zqx https://t.co/dzndJ9ODeA
— Chubby♨️ (@kimmonismus) Dec 13, 2024
@AutismCapital @VitalikButerin just covered the quantum risks and potential fixes for Bitcoin and Ethereum on today's episode @chamath Clip attached. Full ep: https://t.co/iLgV329boV https://t.co/MlUQLOmKtA
— Steady Lads (@0xSteadyLads) Dec 13, 2024
How AI is leading to new robotics startups 🤖 @garrytan: "I have an ex-Apple team called @weaverobotics and they are going to ship a real robot in 2025. It's driven by this idea that the LLM can be the consciousness of the robot." https://t.co/dCIPLVrA6g https://t.co/uTX4SO9OQK
— Overlap: Business & Tech (@Overlap_Tech) Dec 13, 2024
Friday insight (long tweet). Really good Piper Sandler CIO/CTO Survey looking to 2025 just game out. Three key takeaways on Gen AI spending/strategy. - The transition from AI experimentation to implementation is accelerating significantly. The survey shows 56% of CIOs are… https://t.co/V6RjLDx4mF https://t.co/5wjdyOLVyL
— Ben Bajarin (@BenBajarin) Dec 13, 2024
A ton of exciting releases this week & much more to come! So proud of the Gemini, @GoogleDeepMind & @Google teams and their incredible hard work. You can try the Gemini 2.0 models right now for free on AI studio: https://t.co/0FEwTA7wc0 or the API here: https://t.co/QbYVfjwCo7 https://t.co/2XdQHOgk5X
— Demis Hassabis (@demishassabis) Dec 13, 2024
https://t.co/tmffTDcGAP
— Tren Griffin (@trengriffin) Dec 12, 2024
🔬 One of the key innovative method used in @Microsoft Phi-4 is Pivotal Token Search Its a significant advancement in LLM inference optimization, fundamentally changing how models process and attend to sequences of tokens → The technique dynamically identifies and processes… https://t.co/iuowR2JC0W https://t.co/dXnZFTCSnx https://t.co/yDL3Jl1m5w
— Rohan Paul (@rohanpaul_ai) Dec 13, 2024
LangChain is completely removed from CrewAI. We have been slowly chipping away from certain dependencies, most of it was removed but on this version forward there is no longer. https://t.co/gZJZSQCSrj
— João Moura (@joaomdmoura) Dec 13, 2024
Yesterday @GoogleDeepMind released its version of “computer use” with Project Mariner. Project Mariner is a Chrome Extension that uses Gemini 2.0 to control your browser based on human instructions. It scored 90.5% with a tree search on WebVoyager. It can interact with websites… https://t.co/Nzd7DXSIcB https://t.co/Udsfsim7Ri
— Philipp Schmid (@_philschmid) Dec 12, 2024
Nice overview by @omarsar0 of many of the new features we launched yesterday in our Gemini 2.0 announcement. https://t.co/fVT9deIBnp
— Jeff Dean (@🏡) (@JeffDean) Dec 12, 2024
If you’re not building evals you can’t really claim you’re building AI agents, AI workflows or even AI applications in general. (🧵) https://t.co/9Lk6jJIO1S
— Wayne Hamadi 🖇️ (@waynebluewind) Dec 12, 2024
Google broke the Internet with mind blowing Gemini 2.0 Real-Time demos yesterday. And people are already doing wild use cases with it. 10 examples: 1. Gaming will never be the same https://t.co/PyumiWv0Yz
— Min Choi (@minchoi) Dec 12, 2024
Tech giants are going NUCLEAR. There is a big fight between tech giants in the nuclear/SMR space. Each has its own "horse" in the race: - $GOOGL -Kairos Power - $AMZN -X-energy - $MSFT -Three Mile Island - Sam Altman - $OKLO - Bill Gates -Terra Power - $META looking to partner
— Rihard Jarc (@RihardJarc) Dec 11, 2024
These 34 startups each raised $10-50M from top-tier funds in recent weeks and are hiring: https://t.co/ENtLnsDdMp
— Ben Lang (@benln) Dec 11, 2024
I paid the $500 for Devin, the mega hyped AI coding agent, so you don't have to here's an in depth review of how it compares to Cursor: https://t.co/yQQymvnTXH
— Steve (Builder.io) (@Steve8708) Dec 11, 2024
Build a production-grade AI Agent in just 20 mins! That too with 100% open-source tools👇 Building agent-native apps is challenging, and a lot of it boils down to not having appropriate tooling. But this is changing! ✨ CoAgents, an open-source framework by @CopilotKit that… https://t.co/Ock1PQ4ODY https://t.co/ozwdMi5i6r
— Akshay 🚀 (@akshay_pachaar) Dec 11, 2024
Google Gemini 2.0 Flash Multimodal feature is next-level! I can’t believe Google rolled this out before OpenAI. It takes live feeds, answers in real-time with almost no latency, Pretty mind-blowing. Thanks, Google, for this early Christmas gift! Adding more details below 👇 https://t.co/tk6Wh2bakd
— AshutoshShrivastava (@ai_for_success) Dec 11, 2024
One of my favorite applications of LLMs is reading books together. I want to ask questions or hear generated discussion (NotebookLM style) while it is automatically conditioned on the surrounding content. If Amazon or so built a Kindle AI reader that “just works” imo it would be… https://t.co/SmhYnEvFJB
— Andrej Karpathy (@karpathy) Dec 11, 2024
Every single SW team should try this!!! https://t.co/NCuyDQiZ15
— Vinod Khosla (@vkhosla) Dec 10, 2024
Introducing … State of AI Agents 2024 🤖 After handling an incredible 184 billion tokens and 786 million API reqs from 36K developers, we’ve gathered invaluable insights from 4K builders, the next-gen of AI agents. Super excited to share the latest trends shaping AI agents https://t.co/2PpNA7dw8g
— Ahmad Awais (@MrAhmadAwais) Dec 10, 2024
Introducing Llama 3.3 crawler It refines an objective, crawls, and extracts structured data from any website. Built with @togethercompute and @firecrawl_dev's new extract beta endpoint 👀 Check it out: https://t.co/S4W3q2jRKI
— Eric Ciarla (@ericciarla) Dec 10, 2024
software engineering is now a minimum wage job. devin is $500/mo. you get 250 ACUs (agent compute units). one ACU is ~15min of work. this is $2/15 min block or $8/hour. just barely above the federal minimum wage of $7.25, and only half the california minimum wage of $16. https://t.co/zAkSavkFKx
— bayes (@bayeslord) Dec 10, 2024
I just published my deep dive into the AI nuclear SMR market. A lot of work went into this one. Insights on: - AI data center energy needs & nuclear, - SMRs and their differences ( $SMR, $OKLO..) - $AMZN, $MSFT, $GOOGL & $META partnerships with SMRs -.. https://t.co/bpFkGGekRt
— Rihard Jarc (@RihardJarc) Dec 10, 2024
I like the angle @marvell is taking with this concept of customized accelerated infrastructure. This feels like a broader story than what @broadcom has and Marvell has all the key parts of the stack. Custom compute + custom accelerated infra. https://t.co/HCZGz03FjT
— Ben Bajarin (@BenBajarin) Dec 10, 2024
This quote from Steve Jobs lives rent free in my head... https://t.co/cYZVYMgI9U
— Jay Yang (@Jayyanginspires) Dec 9, 2024
The most beautiful castles on Earth 🧵 1. St Michael's Mount, Cornwall https://t.co/0f0UTE1wmU
— James Lucas (@JamesLucasIT) Dec 9, 2024
Intel’s Death and Potential Revival Intel died when mobile cost it its software differentiation; if the U.S. wants a domestic foundry, then it ought to leverage the need for AI chips to make an independent Intel foundry viable. https://t.co/Zte97hdhIi
— Stratechery (@stratechery) Dec 9, 2024
Synthetic datasets with Llama 3? It's easier than you think. I've been playing with a Distilabel, an open-source framework that lets you generate domain-specific synthetic text data using Llama 3. This is huge for anyone working on fine-tuning LLMs or building small language… https://t.co/gVNo55UHFb https://t.co/ihbDvhSBL5
— Akshay 🚀 (@akshay_pachaar) Dec 9, 2024
25 years ago, we started Salesforce with a vision to transform business—and now, we’re leading the charge in digital labor. In this podcast with @HarryStebbings, I discuss: 🤖 How Agentforce is reshaping the AI era 📊 The future of digital labor and SaaS &💡 Leadership lessons… https://t.co/9dem2YlgxX
— Marc Benioff (@Benioff) Dec 9, 2024
I urge you to understand how and why the world order is changing so you can navigate it well. If you are watching what is happening, you are now seeing more fighting within countries, more fighting between countries, and more debt excesses than at any time since the 1930-45… https://t.co/9hKxDsjFhA https://t.co/V6sRtpHTot
— Ray Dalio (@RayDalio) Dec 9, 2024
It took me 53 attempts to convince Marc Benioff to appear on 20VC & record his first ever podcast two years ago. The work you do in private, you are rewarded for in public. Today we sit down again with @benioff to discuss: 🤖 The future of LLMs, Agents & Labor ❌ Why… https://t.co/Q7hDSTztY2 https://t.co/xhwkYqjeSV
— Harry Stebbings (@HarryStebbings) Dec 9, 2024
Peter Thiel on a common red flag he sees in startup pitches “There’s a pattern recognition where if I see a presentation that says, ‘We have this great idea and we’re going to make money in one of these 5 ways: A, B, C, D, or E.’… That’s always sort of alarming and not very… https://t.co/grhmqO7LPs https://t.co/j350kg2ENy
— Startup Archive (@StartupArchive_) Dec 9, 2024
The tech stack I use to build my SaaS products making $1.5k/mo: Frontend - @nextjs ($20/mo) Backend - @nextjs Database - @supabase ($30/mo) Authentication - @supabase Deployment - @vercel UI Kit - @chakra_ui State management - @tan_stack Customer Support - @crisp_im User… https://t.co/sQugNGyji1
— gabriel* (@gabriel__xyz) Dec 9, 2024
Oxford's "Discrete Mathematics" lecture notes by Andrew D. Ker PDF: https://t.co/VBZvFc2Dfc https://t.co/OcGaywJLgP
— Math Cafe (@Riazi_Cafe_en) Dec 8, 2024
this playlist recommended by @karpathy is absolute banger no doubt https://t.co/pmD0zS51tI
— sparrow (@xtrupal) Dec 8, 2024
New @NVIDIA's NVILA vision language model can handle large images and long videos without slowing down. To improve accuracy and efficiency, it uses a "scale-then-compress" strategy: - Scaling up: NVILA increase the resolution of images and videos to capture details. -… https://t.co/AGZjyhfsTD https://t.co/wIzrfQjtpw
— TuringPost (@TheTuringPost) Dec 8, 2024
21 books recommended by Peter Thiel: 1) The Sovereign Individual by James Dale Davidson and William Rees-Mogg https://t.co/7VUbsVb2oY https://t.co/Is6scQaKRc
— Reads with Ravi (@readswithravi) Dec 8, 2024
16 new types of RAG: ▪️ HtmlRAG ▪️ FastRAG ▪️ Auto-RAG ▪️ CORAG ▪️ MemoRAG ▪️ RAG-Thief ▪️ AssistRAG ▪️ LaB-RAG ▪️ Video-RAG ▪️ RAF ▪️ RuAG ▪️ MMed-RAG ▪️ Path-RAG ▪️ Multi-Reranker ▪️ G-RAG ▪️ RAGDiffusion Save the list and check this out for more info: https://t.co/qCHzcqYhJI
— TuringPost (@TheTuringPost) Dec 8, 2024
Still a banger 3 years later. H/T @BTCMinstrel https://t.co/nOPx9v59IO
— Jameson Lopp (@lopp) Dec 8, 2024
What’s an album where the musicians went “god mode”? I’ll start: Miles Davis - Kind Of Blue https://t.co/5MfB3xEkrR
— Melodies & Masterpieces (@SVG__Collection) Dec 6, 2024
@InternetH0F "If you don't fight for your love what kind of love do you have?" ~Keanu Reeves https://t.co/6sLwb6iyyP
— Apurv Anand (@apurv_anand) Dec 6, 2024
