15% of YC's Demo Day were for AI/ML companies. These are a few that caught my eye:
— Ben Tossell (@bentossell) Nov 30, 2022
from Twitter https://twitter.com/bentossell
November 30, 2022 at 06:55AM
via IFTTT
15% of YC's Demo Day were for AI/ML companies. These are a few that caught my eye:
— Ben Tossell (@bentossell) Nov 30, 2022
"Scale is all you need" (for advancing AI capabilities, alignment is a different story) narrative is being refined (as expected) by scaling getting "smart". It is not just size that matters. https://t.co/FODsxh3nGg
— Irina Rish (@irinarish) Nov 30, 2022
I’m often asked about best prompting strategies - OpenAI has released an extensive Cookbook which contains examples, best practices and other prompting advice https://t.co/QDuUY4ST6C https://t.co/KtHR2k9AxR
— Boris Power (@BorisMPower) Nov 29, 2022
The result is a breakthrough for a technique known as imitation learning, in which neural networks are trained to perform tasks by watching humans do them. https://t.co/KmXSF2ZdM1
— MIT Technology Review (@techreview) Nov 29, 2022
The new version of AlphaZero discovered a faster way to do matrix multiplication, a core problem in computing that affects thousands of everyday computer tasks. https://t.co/YnjC3tjM63
— MIT Technology Review (@techreview) Nov 29, 2022
Now awaiting Memories of Ice. Steven Erikson is a blessing. The gun and the crossbow are under the bed. https://t.co/ab5rLPhZdA
— Cassandra B.C. (@michaeljburry) Nov 29, 2022
The result is a breakthrough for a technique known as imitation learning, in which neural networks are trained to perform tasks by watching humans do them. https://t.co/GH9uVmS6Bb
— MIT Technology Review (@techreview) Nov 28, 2022
Meta AI releases Galactica, an LLM trained on 48M+ scientific papers. Functions like "Copilot for researchers", demonstrates a wide base of scientific knowledge. See thread below for paper summary & discussion! Paper: https://t.co/yU9zYQ6bvI Site: https://t.co/LaF5r3MkYh 7/14 https://t.co/RuekCb5yoh
— AI Pub (@ai__pub) Nov 28, 2022
Last two weeks in AI (Nov 13-27): - MinD-Vis: reading mental imagery from brain recordings, - Meta AI's CICERO plays human-level Diplomacy, - Stable Diffusion 2.0, - Galactica: science-generation text model, - Action-driven LLMs, - The future of data moats, ... and more: 1/14 https://t.co/W4uGNbELTL
— AI Pub (@ai__pub) Nov 28, 2022
@patrick_oshag Read this: https://t.co/ZsDXF8pZwV Then this: https://t.co/1wpZHT8rHh
— Tren Griffin (@trengriffin) Nov 28, 2022
I'm publishing a call to action: The AI War and How to Win It. AI for national security will define the future of our world. Either the USA wins, or our authoritarian adversaries do. I walk through The AI War, The China Threat, and How to Win It. ๐งต https://t.co/7by2iroWWD
— Alexandr Wang (@alexandr_wang) Nov 27, 2022
Great line from Steve Jobs’s biography: https://t.co/5JhDQV5dDw
— David Senra (@FoundersPodcast) Nov 26, 2022
$SHOP Shopify breaks Black Friday record. • Sales grew +17% Y/Y to $3.4 billion. • Sales grew +19% Y/Y fx neutral. • Shopify POS grew +27% Y/Y. It turns out e-commerce is not dead? https://t.co/Hf21Cw1o3v
— App Economy Insights (@EconomyApp) Nov 26, 2022
Holiday book gifts? 1. Influence by Robert Cialdini 2. Shoe Dog by Phil Knight 3. The Outsiders by William Thorndike. 4. Competitive Strategy by Michael Porter 5. The Misbehavior of Markets by Benoit Mandelbrot (harder) 6. Incerto by Nassim Taleb https://t.co/2sh5RmbNeb
— Tren Griffin (@trengriffin) Nov 26, 2022
Imagine an AI agent teaching you a new skill in the metaverse. Or a non-player character helping you find your way to a win in a video game. How ever developers responsibly build on CICERO’s code, there’s potential for AI to help humans reach more of their goals. #CICERObyMetaAI
— Meta AI (@MetaAI) Nov 25, 2022
The secret prompt words are: "bismuth neon chrome copper patina verdigris". Use wisely. https://t.co/FeaqTBP4ux
— surea.i (@sureailabs) Nov 25, 2022
https://t.co/6OeoiIW358 https://t.co/39LBvOHKjp
— Diagram (@diagram) Nov 25, 2022
266-page FREE PDF — #Probability & #Statistics content used to prepare first-year #mathematics students at Cambridge University: https://t.co/KGEJpK5CBB ———— #StatisticalLiteracy #BigData #DataScience #AI #DataScientists #DataLiteracy #MachineLearning https://t.co/nFgv8wvsRA
— Kirk Borne (@KirkDBorne) Nov 25, 2022
Most interesting telling of the story of the Pilgrims and the Mayflower: https://t.co/L5XV8TCdkQ @bhu_srinivasan https://t.co/jt4YyGKA5z
— Marc Andreessen (@pmarca) Nov 24, 2022
⚡️ New Tutorial: @StableDiffusion v2.0 on Colab https://t.co/6W8tQL1Jmd ⭐️Use #StableDiffusion2 on Google Colab with @huggingface {diffusers} ⭐️ Also, learn to build a super fast @Gradio web UI (2 lines of ๐ code ๐คฏ) https://t.co/S6XY2A4Vsn
— 1LittleCoder๐ป (@1littlecoder) Nov 24, 2022
https://t.co/EdLv2f9sbD
— jack (@jack) Nov 24, 2022
I'm thankful today that we've moved past 1- trillion parameters. “The industry is moving past 1-trillion-parameter models, and we are extending that boundary by two orders of magnitude, enabling brain-scale neural networks with 120 trillion parameters." https://t.co/ybmrggyMgl
— Tren Griffin (@trengriffin) Nov 24, 2022
I repeat: OpenRAIL licenses are explicitly anti open source, moralistic, narrow-minded, and they force anyone who re-uses or fine-tunes this model to implement a binding contract mechanism just to share what they have done ๐ https://t.co/3hYqV9RtHV
— Yannic Kilcher, Tech Sister (@ykilcher) Nov 24, 2022
Stable Diffusion 2.0 has officially been released! - image quality improved - base 512x512 and 768x768 image generation - up to 4x upscaling for 2048x2048+ images - depth->image model released as well blogpost: https://t.co/9oCVvwUfbK HF: https://t.co/ej9JmDJOUA https://t.co/FKIumbCaMY
— nearcyan (@nearcyan) Nov 24, 2022
Good read about how fighting crime is done in crypto. The main law enforcement characters, Tigran Gambaryan and Matt Price both work in the cybercrime team @binance now. https://t.co/LvlAKQ8Jkj
— CZ ๐ถ Binance (@cz_binance) Nov 23, 2022
GPT3 is powerful but blind. The future of Foundation Models will be embodied agents that proactively take actions, endlessly explore the world, and continuously self-improve. What does it take? In our NeurIPS Outstanding Paper “MineDojo”, we provide a blueprint for this future:๐งต https://t.co/YZps22n9pA
— Jim (Linxi) Fan (@DrJimFan) Nov 23, 2022
$BLK is king https://t.co/YTPna6qoXN
— The Transcript (@TheTranscript_) Nov 23, 2022
"So it’s a lot of hard work and it’s a lot of worrying constantly, and if you don’t love it, you’re going to fail. So you’ve got to love it and you’ve got to have passion. I think that’s the high-order bit." - Steve Jobs, D5 conference, 2007
— Steve Jobs Stories (@SJobs_Stories) Nov 22, 2022
Nothing to see here. Just an AI system learning to strategically plan world conquest, negotiate with humans in English and then betray them via the game of Diplomacy. https://t.co/zkN1EuJnAe
— Erik Brynjolfsson (@erikbryn) Nov 22, 2022
Ooof. Ten. Billion. Dollars. https://t.co/ruxy8inXyI
— Grady Booch (@Grady_Booch) Nov 22, 2022
Freezing of soap bubble.. ❄️ https://t.co/zV03WKhbjt
— Buitengebieden (@buitengebieden) Nov 21, 2022
About time. UK's financial conduct authority sounds the warning on gamification of stock trading apps. Trading apps everywhere should take note. A big influence of social media on day-to-day lives is gamifying everything for greater engagement! https://t.co/BhobU8UYxO https://t.co/sJ5f3GRLmf
— Subrahmanyam KVJ (@SuB8u) Nov 21, 2022
Bedroom with a view! ๐ https://t.co/1Z1so3jEm6
— INTERIOR VILLA (@Interiorvilla1) Nov 21, 2022
Why did Renรฉ Magritte paint a man with his face hidden by an apple in 1946? And why did he call it The Son of Man? Welcome to the uncanny world of Surrealism... https://t.co/Y46BtRXpkL
— The Cultural Tutor (@culturaltutor) Nov 20, 2022
https://t.co/ht8Tbj1xXs
— Gopi Vikranth (@GopiVikranth) Nov 20, 2022
Our paper on understanding AlphaZero♟️is now published at PNAS! https://t.co/R2Cpt1XsaW The paper "studies" AZ's internals and its behaviors in collaborations with @DeepMind and world chess champion @32gcfhkmm. What did we learn? ๐งต
— Been Kim (@_beenkim) Nov 20, 2022
@BlancheMinerva Not the paper you're looking for, but this paper perhaps might cite it https://t.co/wD5PCL4Sa9
— Jack Clark (@jackclarkSF) Nov 19, 2022
David Walsh is one of the most successful professional gamblers ever. His syndicate places $10 billion in bets per year. This is a fascinating rare interview. https://t.co/E1lOXJUG4V https://t.co/YjXmevewM1
— Ian Cassel (@iancassel) Nov 19, 2022
@cz_binance https://t.co/SzGg7rTyjQ
— Sal the Agorist (@SallyMayweather) Nov 19, 2022
In 1992 (revised in 2009), Tweedy Browne published a paper titled ‘What has worked in investing’ The paper covers over 50 studies of investment approaches associated with exceptional returns. It’s a must-read for all investors, here is the link: https://t.co/jhCEtx0EO6 https://t.co/A0qRZLm3Gi
— Frank - Lone Wolves (@lonewolvesfrank) Nov 19, 2022
US TTM E-commerce Sales. It is not dead, just crossed $1T. It is still very much above trend-line, +25%. Total Retail was $7T. E-commerce penetration ~14.1%, ex-categories that don’t compete (i.e. restaurants, car dealers, gas station), the penetration was higher at ~21%. https://t.co/JRO62HIJZ0
— Eugene Ng (@EugeneNg_VCap) Nov 18, 2022
This week in enterprise software: Top 10 #SaaS #Cloud multiples as of today's market close https://t.co/P8N7j1zymJ
— Jamin Ball (@jaminball) Nov 18, 2022
Peter Thiel on the most important lesson to learn from Steve Jobs: https://t.co/FAMKpiGzhj
— David Senra (@FoundersPodcast) Nov 18, 2022
Meet MultiRay, Meta’s new platform for efficiently running large-scale, state-of-the-art AI models. By converting input to an embedding that’s cached & shared between users, MultiRay greatly reduces complexity & energy use. Here’s how it's enabling us to build & scale ML work ⬇️
— Meta AI (@MetaAI) Nov 18, 2022
Visualizing Major Layoffs at U.S. Corporations ๐ Hiring freezes and layoffs are becoming more common in 2022, as U.S. businesses look to slash costs ahead of a possible recession. Learn more: https://t.co/9gq2S8747E https://t.co/ELmldkmX0s
— Visual Capitalist (@VisualCap) Nov 18, 2022
If you'd like to take the new Merlin Dataloader library for a spin ๐ I created a @kaggle kernel that trains a matrix factorization model to generate candidates for a live RecSys competition ๐ Please find it here: https://t.co/KA8LtgcfMh https://t.co/74ZSoUERJ3
— @radek@sigmoid.social (Mastodon) ๐บ๐ฆ (@radekosmulski) Nov 18, 2022
Foundation models are everywhere! Except enterprise production environments. Why? They need to be adapted to your specific use case, and their massive size makes them difficult (and $$$) to deploy. Today we're introducing Data-centric Foundation Model Development in Snorkel Flow! https://t.co/SnqK8aazpi
— Braden Hancock (@bradenjhancock) Nov 17, 2022
"It is clearly a person that spent his life on one thing. He found his great subject, early in life, and he pursued it over decades. And he's made a deep deep deep contribution that will last well beyond his physical years." — Steve Jobs, about Dr Juran, 1991
— Steve Jobs Stories (@SJobs_Stories) Nov 16, 2022
3 papers to understand Time-Series Forecasting ⏳ better. 1. Time-series Extreme Event Forecasting @UberEng 2. AutoML for Time-Series Forecasting @GoogleAI 3. AR-Net @MetaAI A Thread ๐งต๐ https://t.co/ul6KRFhOrX
— TheSequence (@TheSequenceAI) Nov 16, 2022
New post just dropped: The future of AI is action-driven https://t.co/h4k7LAHBMx
— John McDonnell (@johnvmcdonnell) Nov 15, 2022
Each chip delivers up to 12.1 TB/s of internal bandwidth (96.8 Terabits) to the AI cores, but the data is fed to the CS-2 processors via 100 GbE networking spread across 124 server nodes in 16 racks, which is nice. https://t.co/m4JX2kJFLm
— Tren Griffin (@trengriffin) Nov 15, 2022
"I don't always run large language models, but when I do, I like to use a CS-2 Wafer-Scale Engine, with 850,000 AI optimized compute cores, 40 Gigabytes of on chip high performance memory and 20 Petabytes/s of memory." https://t.co/KctCXaQepf
— Tren Griffin (@trengriffin) Nov 15, 2022
We are glad to announce our new work collaborating with https://ift.tt/JMQDnec a large-scale text-audio contrastive learning model with a large-scale text-audio dataset! Paper: https://t.co/JvBdgHqO18 Code & Model: https://t.co/3ZdicperXg Dataset: https://t.co/0EHQkxKsgW 1/6 https://t.co/iweDsYVe3v
— Yusong Wu (@wuyusongwys) Nov 15, 2022
Narratives by @stratechery @benthompson looks at (reflects on) role of narratives in why there are big misses. IMO this is much larger issue than looking at extremes of FTX/Twitter. Narratives have come to replace reporting, analysis, and even facts. 1/ https://t.co/aMUOFJnICd
— Steven Sinofsky (@stevesi) Nov 14, 2022
Matt Levine goes 15/10: a MUST-read post. “[The FTX balance sheet is] an Excel file full of the howling of ghosts and the shrieking of tortured souls. If you look too long at that spreadsheet, you will go insane.” https://t.co/TqWaWj2ANu
— Bruce Kleinman (@brucedkleinman) Nov 14, 2022
๐ค Your AI assistant that creates graphics for presentations. Simply tell https://t.co/L3xQ4xoOSe what your visualization should show and we'll find the data / create it. Let us know your thoughts or questions below! #ai #gpt3 https://t.co/3npknuIdMq
— Sean Doherty ๐ซ (@seandoher1y) Nov 14, 2022
Nice blog post on tackling the ML tech stack. Actually, keeping up via books is also the more calm & structured way to learn (compared to sifting through preprints & GitHub repos ๐ ) https://t.co/DqpjwpITJx
— Sebastian Raschka ๐ (@rasbt) Nov 13, 2022
Unifying Diffusion Models' Latent Space, with Applications to CycleDiffusion and Guidance + Diffusers and Gradio Demo https://t.co/Yibq1EERpw
— /MachineLearning (@slashML) Nov 13, 2022
"The Vision Fund segment lost 1.02 trillion yen in July-September, following a 2.33 trillion yen loss in the June quarter." Do you know what a "prepaid forward" contract is? https://t.co/yh3GFFvPAe
— Tren Griffin (@trengriffin) Nov 13, 2022
If you're trying to wrap your head around FTX, I bookmarked 100+ tweets over the past week including: the timeline, SBF's story and major red flags. And the memes (so many memes). Please read and share: https://t.co/ZsBs4JY4KB
— Trung Phan (@TrungTPhan) Nov 12, 2022
@patrick_oshag Just saw this https://t.co/qcQ3oIDkT4 interested in those PDF things especially with respect to how it will disrupt traditional media and video games.
— Emad (@EMostaque) Nov 12, 2022
This circuit will light your abs on fire and tone the backside of your body. Do each exercise until failure, one after the other. Rest 90s, 3-5 sets. Enjoy the burn, and have a wonderful weekend. https://t.co/NxZv2d1THT
— Alex Bernier (@mythoughtfood) Nov 11, 2022
The Sam Bankman-Fried empire crumbled. What happened? https://t.co/p5T54lasD8 (https://t.co/h57nHeymJx)
— Hacker News 200 (@newsyc200) Nov 10, 2022
.@NVIDIA has just announced and published the latest set of performance benchmarks for various Deep Learning training and inference tasks on A100 and H100 GPUs. 1/3 https://t.co/S0Nc6KNZUa
— Bojan Tunguz (@tunguz) Nov 10, 2022
Winston Churchill on seizing the moment: https://t.co/M0o1eFGsbd
— FS (@farnamstreet) Nov 10, 2022
BLOOM: A 176B-Parameter Open-Access Multilingual Language Model abs: https://t.co/cEYC0zSd51 https://t.co/VUbLKOStpi
— AK (@_akhaliq) Nov 10, 2022
Practical #MLOps — Operationalizing #MachineLearning Models: https://t.co/VmqhPIVAzr ————— #BigData #DataScience #AI #EnterpriseAI #AutoML #ModelOps https://t.co/mBq8NAwCEO
— Kirk Borne (@KirkDBorne) Nov 9, 2022
“Who doesn’t want to bet on a young wunderkind? That was a part Bankman-Fried played so well.” https://t.co/6KX7fRYPxQ
— Kate Clark (@KateClarkTweets) Nov 9, 2022
FTX raised nearly $2 billion from investors including Sequoia Capital and Paradigm. A forced sale risks slashing the value of their stakes to zero. https://t.co/ppVfmuFoOb
— The Information (@theinformation) Nov 8, 2022
DigitalOcean with a 3.5 month gross-margin adjusted payback in Q3. Added $97m of net new ARR this quarter off of $18m of fully burdened S&M spend last quarter. 15% FCF margin as well $DOCN
— Jamin Ball (@jaminball) Nov 7, 2022
As of @elonmusk's purchase of Twitter, there's a whole new crowd of users trying to coordinate their exit. It's 2022 tho, & we have a place we can put our social graph where we can own & control it and any social platform can use it: the blockchain. https://t.co/3VHkdUrqQP
— jonstokes.(eth|com) (@jonst0kes) Nov 7, 2022
Great overview blog post on why transformers need certain optimization tricks that aren't needed by other architectures https://t.co/TnKJlZfMCj
— Eric Jang (@ericjang11) Nov 7, 2022
PhotoRoom raised $19 million for its AI photo studio, which is already powering millions of resellers and small businesses. Today, we are launching a new generative AI feature, magic studio, to create stunning marketing images from any product photo A thread ๐งต https://t.co/39CI8ePx8d
— Matthieu Rouif (@matthieurouif) Nov 7, 2022
I guess this makes me an expert? https://t.co/Ty6MmMZXHq
— Jordan Schneider (@jordanschnyc) Nov 7, 2022
An incredible remote controlled F-22 https://t.co/dtLpH3NLQx
— How Things Are Manufactured (@fastworkers6) Nov 6, 2022
Commentary from $NFLX Netflix Q3 FY22 Shareholder letter worth revisiting while digesting the $WBD report. https://t.co/Fwm8mLUvKP
— App Economy Insights (@EconomyApp) Nov 6, 2022
Article: GARCH-Like Multivariate Volatility Models for Stock Prices - using MVARCH https://t.co/IOrhrFMvYi Python GitHub for MVARCH: https://t.co/mW3mTMYVhO https://t.co/1NLG91ZkTJ
— Carl Carrie (@๐ ) (@carlcarrie) Nov 6, 2022
Good read. Givers, fake givers, matchers and takers. Why givers win. https://t.co/2A9vConQ3d https://t.co/Sj2yFwm5Yz
— CZ ๐ถ Binance (@cz_binance) Nov 6, 2022
Peter Thiel: Distribution should be part of product design https://t.co/qzwcdJ3xDs
— David Senra (@FoundersPodcast) Nov 5, 2022
Steve Jobs on coming up with product ideas: https://t.co/xALZ4C1E8k
— David Senra (@FoundersPodcast) Nov 4, 2022
Stripe is cutting 14% of staff, which will return it to a headcount of about 7,000, and is “firmly reining in all other sources of cost.” https://t.co/gmgG308PSS
— Kate Clark (@KateClarkTweets) Nov 3, 2022
Paul Graham on why ambitious people need to be around other ambitious people: https://t.co/1HKekIandV
— David Senra (@FoundersPodcast) Nov 2, 2022
startup versus big company https://t.co/EmZc3RnUvm
— clem ๐ค (@ClementDelangue) Nov 2, 2022
Effective medicines from nature! https://t.co/3M08NXHWsx
— How Things Are Manufactured (@fastworkers6) Nov 2, 2022
OpenAI was able to strongly relax our policies by: 1. aligning our models to be great at following instructions https://t.co/dXb2ZFQ5H3 2. releasing a moderation endpoint, to help developers https://t.co/eFPbWB5UtS Almost anything net-good is allowed now! https://t.co/vLNLIYPmSn https://t.co/p8nGd5gjlM
— Boris Power (@BorisMPower) Nov 1, 2022
A lot has been speculated about TikTok's recommendations. This is the first paper I've read by the team, and it has many interesting details: expirable embeddings, parameter server, online training... Good #recsys stuff https://t.co/r8UBCrnZi8
— Xavier Amatriain (@xamat) Oct 31, 2022