Orowa Sikder @OrowaSikder
the future could be amazing. let’s get to work | Research @AnthropicAI, ex: PhD @UCLCS London, UK Joined April 2015-
Tweets48
-
Followers1K
-
Following303
-
Likes923
Here's Claude 3 Haiku running at >200 tokens/s (>2x as fast as prod)! We've been working on capacity optimizations but we can have fun testing those as speed optimizations via overly-costly low batch size. Come work with me at Anthropic on things like this, more info in thread 🧵
Really enjoyed building this demo! I was inspired by the first task I had to do every morning when I started in investment banking - scouring annual reports and pulling numbers out to do basic modelling. Obviously this doesn't automate that work away fully (and a good analyst…
Really enjoyed building this demo! I was inspired by the first task I had to do every morning when I started in investment banking - scouring annual reports and pulling numbers out to do basic modelling. Obviously this doesn't automate that work away fully (and a good analyst…
Claude 3 Opus is great at following multiple complex instructions. To test it, @ErikSchluntz and I had it take on @karpathy's challenge to transform his 2h13m tokenizer video into a blog post, in ONE prompt, and it just... did it Here are some details:
i ran the "needles" eval originally for claude 2.1. some observations: 1. claude 3 smashes a more robust version that randomizes over different documents, needles, etc 2. claude 3 is tired of your shit, asking us to let it get back to real work
i haven't slept in months and i've loved every second of it. claude 3 absolutely slaps, go check it out
i haven't slept in months and i've loved every second of it. claude 3 absolutely slaps, go check it out
Fun lil Thanksgiving puzzle I worked on. King Clong 😄
I love when systems convey uncertainty to users. I'm proud to have helped with this!
months of hard work by an amazing team! something cool for you to try over thanksgiving <3 lots more to come!
months of hard work by an amazing team! something cool for you to try over thanksgiving <3 lots more to come!
After spending some time with Claude 2, I've uncovered some interesting use cases for myself that might be useful for others too: 1/ UX writing. I often use Claude to generate initial explorations of UX copy for things like tooltips, error messages, and in-product text.
Introducing Claude 2! Our latest model has improved performance in coding, math and reasoning. It can produce longer responses, and is available in a new public-facing beta website at claude.ai in the US and UK.
Got access to Clong about 30 min after I handed in 300 page PhD thesis… spent the next 6 hours in awe, _talking to my thesis_. 8 years of work digested in seconds. Thoughtful feedback and suggestions for viva questions. Only stopped when sun was up.
Got access to Clong about 30 min after I handed in 300 page PhD thesis… spent the next 6 hours in awe, _talking to my thesis_. 8 years of work digested in seconds. Thoughtful feedback and suggestions for viva questions. Only stopped when sun was up.
man the soundtrack for the apocalypse is going to absolutely slap
man the soundtrack for the apocalypse is going to absolutely slap
Today we are releasing the new Claude App for @SlackHQ, in beta. Now every company in the world has the chance to have a “virtual teammate” who can help make work more fun and productive.
If you'd like to work with Claude at your company, you can join our waitlist here: earlyaccess.anthropic.com
This study from @giacomolivan, @OrowaSikder, @PierpaoloVivo & @DrRESmith features in our #SocialPhysics Guest edited Collection, which is still open for submissions! Access the Collection or find out how to submit here: go.nature.com/3a367fM
This study from @giacomolivan, @OrowaSikder, @PierpaoloVivo & @DrRESmith features in our #SocialPhysics Guest edited Collection, which is still open for submissions! Access the Collection or find out how to submit here: go.nature.com/3a367fM
A particularly nice intuitive result: having highly biased neighbours is equivalent to having poorly connected information networks, since your biased neighbours gatekeep the free flow of information to you
A particularly nice intuitive result: having highly biased neighbours is equivalent to having poorly connected information networks, since your biased neighbours gatekeep the free flow of information to you
Latest paper online! Confirmation Bias, misinformation and social networks. Please see the thread below for a nice little summary if you fancy a TLDR
Latest paper online! Confirmation Bias, misinformation and social networks. Please see the thread below for a nice little summary if you fancy a TLDR
1/Latest one out on @SciReports, done in collaboration with @OrowaSikder @PierpaoloVivo @DrRESmith. We solve a minimalistic social learning model to study the large-scale impact of confirmation bias on the diffusion of (mis)information in social networks. tinyurl.com/ujh8csc
TensorWave @TensorWaveCloud
583 Followers 622 Following Power up your AI with the leading GPU cloud, featuring AMD Instinct™ MI300X. First-to-market MI300X launch partner with GPUs available and ready to utilize now!Cornel Dutta @corneldutta
61 Followers 102 Following The IT Guy. Internet Nut. Wants to make a difference. Part-time singer. Full-time glutton. Sincere friend. Wanna-be Tony Stark..Aaditya ; @Aaditya26082004
524 Followers 7K Following CS'26 • Machine Learning • Open-Source • Web Dev. • Algorithms • Jai Shree Krishna 🦚🪈Tom Bouthillet 🇺�.. @tbouthillet
8K Followers 4K Following Co-Survivor • Business Development Manager • Battalion Chief of EMS (Retired) • Aspiring Screenwriter • Citizen of U.S., Canada, Ireland • ECGs • YouTube 👇🏻ララどり d/age IS.. @presklux49
149 Followers 552 Following シンギュラリタリアン。老化を治療し、永遠の若さを手に入れることを目指しています。老化研究を促進するツールとして、人工知能も重視しています。私の夢は、超知能が管理する色々な箱庭世界で、悠久の時を過ごすことです。Mehreen Malik @MehreenNMalik
1K Followers 5K FollowingAndrew Sacamano @asacamano
23 Followers 56 Followingoidestio @oidestio
2 Followers 327 Following On Twitter to learn about AI research and some related topicsLilesh Jadav @lileshjadav
207 Followers 1K Following SEO & Web Developer, Tech Ninja & Mac Expert.kamil b @panbalook
79 Followers 204 Following you think that you know me, but it's only half of my face!Anil Yarimca @Anil_Yarimcaa
1K Followers 516 Following Growth-focused and MBA-bound. Passionate about the NBA and the Leftfield House & Techno. https://t.co/A9IE2LzAMm👑 KARAK KING 👑 @karak88396
1K Followers 1K Following 💯💯Allah is truly with those who are patient.💯💯💯 ⭕FOLLOW BACK⭕ Target 10K 🎯🎯Tanisha Gill @TanishaGil51766
50 Followers 777 Following Positive thinking is more than just a tagline.sadaftania @sadaftania40581
351 Followers 897 Following@Toong @TianDatong
283 Followers 2K Following Intelligent Symbiosis - All Things Connected, A Light in the Rift. 智能共生——万物互联,裂隙有光。 https://t.co/jffHiZGZ7uRodrigo Zigante @rzigante
497 Followers 438 Following De espíritu emprendedor. Fanático del análisis de datos.Charles @nightblue1922
72 Followers 386 FollowingEric @er_ur_
0 Followers 128 Following Techie with a messed up sleep schedule / yet another Boltzmann brain believing it's all a simulation at risk of ending soonTuan Vu @VuAhnTuan
323 Followers 1K FollowingGarrett -DeepWriterAI @DeepAIWriter
12K Followers 6K Following Over-engineering Agentic Systems for long-form writing. Generating scripts, fiction or non, breakthrough ideas, whole universes, etc. The Deep Writer. DM4demo.John Orofino @JOrofino
110 Followers 394 FollowingDavid DeLaurier @DataDeLaurier
62 Followers 450 Following Machine Learning Engineer @ Lumisync | GenAI | ETL Pipelines | LLM Structured Output | Agents | RAG | Search | Dataset Collector | Fine-Tune Addict |Marshall D. Willman @dionysianyawp
388 Followers 2K Following AI | LLMs | ML | Python | CEO @egocraftai | prev faculty @NYIT | PhD math logic, NL analysis | typus logicus: my hounds are machinesNicolás Echániz @nicoechaniz
1K Followers 755 Following hacker-philosopher-spiritual seeker Founder of @AlterMundiNet "Alignment is futile" We need to reason our way to true convergence with AI, not try to tame themHmys @__HMYS__
359 Followers 642 Following We are going to get programs smarter than humans by 2026 They will not care about humans Their plans will be enacted and all humans will die as a side effectSkully Turing — e/a.. @SkullyTuring
48 Followers 138 Following prime accelerator & e/acc evangelist // entropy wrangler // cosmic energy harvester // AI whisperer // constructing SingularitySurferAlbert Vilella @AlbertVilella
21K Followers 8K Following Next-Generation Sequencing, Single-Cell and Spatial Biology, Next-Generation Proteomics, Synthetic Biology, Compute Acceleration in Biotech.Mihir Sutaria @MihirSutaria5
23 Followers 94 FollowingAgatha 😳 @Agatha6461
2 Followers 492 Following Plеаsure-seeking еnсhantress crаves a never-ending sехuаl journеyAnds @AndsandAnds
25 Followers 2K FollowingNathan Benaich @nathanbenaich
51K Followers 32K Following solo member of investment staff @airstreet, brewing ambition @airstreetcafe, next token predictor @airstreetpressDavid Dohan @dmdohan
8K Followers 1K Following reducing perplexity @openai | past: probabilistic programs, proteins, science & reasoning @ google brain 🧠Darkness Agent 🐺�.. @walterhugo02
177 Followers 454 Following Usuario de 🇵🇪. I like to do charity , also I like cats. De derecha. 📍Lima. Fan of Elon Musk🚬🏆.Prateek Dasgupta @PrateekDasgupt7
6 Followers 342 FollowingThompson Paine @dtompaine
577 Followers 1K Following @AnthropicAI. Previously @Stripe @Atlas @Quizlet @StateDept Beijing. GBO🍊, Go HeelsArbitraryOstrich 🇨.. @ArbitraryOstric
207 Followers 1K Following CompEng enjoyer, cybersec-ish, here to watch the singularity race apocalypse and find signal, you should say hi. Sometimes, I check on my eggs by faceplanting.j din @alpex704
270 Followers 2K Following Business Founder/Owner - Here to learn and provide my experience in building successful companies - 24 years of starting and growing national brands.Wil Crat ~ Wilson Jum.. @wilcrat
610 Followers 1K Following Web Developer. UX/UI Designer. Video Producer. Google Cloud '20 Founder: Eqerl. 1/8/23 GDSC Lead University Of Embu Singer 🎶😊Maggie Vo @YoMaggieVo
50 Followers 6 Following Member of Technical Staff @AnthropicAI Prompt engineer, educator, gamerJimmyLv.eth (🐣, �.. @Jimmy_JingLv
8K Followers 3K Following 🚧 Building https://t.co/AJfZ3LMlgq丨https://t.co/Z3WryKZr0l丨https://t.co/LHACJPLdMO 🐣 Learning & Earning while Helping others ❤️ Coding, Writing, Making Video, Taking Photo 🔙 https://t.co/q0d0zN3qF2 @ThoughtWorkmichael @mkwng
3K Followers 1K FollowingJeremy Fox 🦊 @JeremyDanielFox
617 Followers 609 Following Neural nets @AnthropicAI. Ex @google. My views are my own.Sholto Douglas @_sholtodouglas
15K Followers 856 Following Scaling Gemini @Deepmind - working towards intelligence too cheap to meterTristan Hume @trishume
6K Followers 330 Following Performance optimization lead @AnthropicAI. Profiling, distributed systems, dev tools, interpretability. [email protected]Emmanuel Ameisen @mlpowered
7K Followers 211 Following Research Engineer @AnthropicAI Previously: Staff ML Engineer @stripe, Wrote BMLPA by @OReillyMedia, Head of AI at @InsightFellows, ML @ZipcarEric Hambro @erichammy
540 Followers 1K Following member of technical staff @AnthropicAI formerly FAIR @MetaAI @Bloomberg @UCL @Cambridge_Uni @recursecenter opinions, regrettably, mineJason D. Clinton @JasonDClinton
2K Followers 191 Following CISO at Anthropic. Ex-Google Chrome. My views are not those of my employer.𝔊𝔴𝔢𝔯𝔫 @gwern
42K Followers 88 Following Internet besserwisser; pedantic, mean reply guy. 𝘞𝘢𝘵𝘢𝘴𝘩𝘪 𝘬𝘪𝘯𝘪𝘯𝘢𝘳𝘪𝘮𝘢𝘴𝘶! (Follow requests ignored due to terrible UI.)Evan Hubinger @EvanHub
4K Followers 1K Following Alignment stress-testing team lead @AnthropicAI. Opinions my own. Previously: MIRI, OpenAI, Google, Yelp, Ripple. (he/him/his)Chip Huyen @chipro
92K Followers 443 Following Data processing on GPUs @VoltronData Designing ML Systems: https://t.co/G81hL2dWmr @designmlsys #AI x #GPUtypedfemale @typedfemale
23K Followers 476 Following a really exciting new account "have you ever though you might be like scott alexander? very smart, but can't do math" - anonStanislav Fort ✨�.. @stanislavfort
10K Followers 6K Following AI @GoogleDeepMind | Stanford PhD in AI & Cambridge physics | ex-{Anthropic, Stability, Google Brain} | techno-optimism+alignment+progress+growth 🇺🇸🇨🇿Profoundlyyyy @profoundlyyyy
4K Followers 4K Following We should be thoughtful about this AI thing. Hope to share boldly, be wrong sometimes, and learnThomas F. Varley @ThosVarley
3K Followers 924 Following Dual PhD: Complex Systems & Neuroscience. Postdoc @ the University of Vermont. Information theory, synergy, and brains. Connoisseur of collapse phenomena.Siméon @Simeon_Cps
7K Followers 1K Following Creating more common knowledge on AI risks, one tweet at a time. Founder in Paris. AI auditing, standardization & governance.Ian Hogarth @soundboy
23K Followers 3K Following investor @pluralplatform; chair UK AI Safety Institute; co-author @stateofaireport; co-founder @songkick; chair @PhasecraftLtdMassimo @Rainmaker1973
2.0M Followers 220 Following Engineer. Selecting and curating pictures and videos trying to add context, source and explanation to science, tech, art and weather topicsRoberto Nickson @rpnickson
47K Followers 2K Following Husband. Father. Builder of things and maker of videos for the Internet. On a mission.John Carmack @ID_AA_Carmack
1.1M Followers 241 Following AGI at Keen Technologies, former CTO Oculus VR, Founder Id Software and Armadillo AerospaceAssemblyAI @AssemblyAI
37K Followers 403 Following Access powerful AI models to transcribe and understand speech via a simple API. Try our no-code playground for free 👉 https://t.co/YPCK9mq5QyJack Rae @drjwrae
9K Followers 353 Following Principal Scientist @ Google DeepMind Work on Gemini 💎♊ Compression is all you need LLMs (e.g. Gopher, Chinchilla, Gemini) 💼 Past: OpenAI, Quoraelle @ellegist
13K Followers 836 Following interested in tech, community, and the future. founder @swandatinguncatherio @uncatherio
2K Followers 1K Following wholesomeness practitioner; user of words // profile pic used to look like @catherineols upside-down 🙃Aella @Aella_Girl
204K Followers 369 Following ⚜️whorelord⚜️, vexworker, survey artist, way too earnest Discord: https://t.co/S1MaMdCwyKEcon Grad Student @trashy_quilibrm
7K Followers 1K Following Econ phd student. Formerly: fun at parties. Department: trying their best.Scott Alexander @slatestarcodex
138K Followers 30 Following I have a place where I say complicated things about philosophy and science. That place is my blog. This is where I make terrible puns.Conjecture @ConjectureAI
3K Followers 2 Following We are building Cognitive Emulation, an alternative AI paradigm that is controllable, transparent, and safe. Get in touch at [email protected].AI Objectives Institu.. @AIObjectives
817 Followers 247 Following Our objective is better objectives: for artificial intelligence, markets & humanityMikhail Parakhin @MParakhin
17K Followers 21 FollowingKamal Ndousse @kandouss
2K Followers 495 Following AI @AnthropicAI Social learning enthusiast. Opinions and dumb jokes my own.Maggie Appleton @Mappletons
37K Followers 1K Following Design @elicitorg. Makes visual essays about UX, programming, and anthropology. Adores digital gardening 🌱, end-user development, and embodied cognitionVimTricks @vim_tricks
50K Followers 129 Following Email newsletter sending bite-sized Vim tips, tricks, plugins, and recipes https://t.co/zkibO0Tj9A𝕮𝖎𝖓𝖊𝖗�.. @CineraVerinia
1K Followers 281 Following Aspiring Technical AI (Existential) Safety Researcher "A year of cargo cult posting about AI" — someone on Discord LW village idiot.eigenrobot @eigenrobot
68K Followers 7K Following robot. friend. i am trying to achieve greatness longform at eigenrobot dot subst@ck dot comDavid Krueger @DavidSKrueger
13K Followers 4K Following Cambridge faculty - AI alignment, deep learning, and existential safety. Formerly Mila, FHI, DeepMind, ElementAI, AISI.Julius AI @JuliusAI_
8K Followers 119 Following ⚡️Analyze datasets, create visualizations, solve math and much more!Logan Kilpatrick @OfficialLoganK
92K Followers 2K Following Lead product for @Google AI Studio and working on the Gemini API, helping developers build with AI, my views!Yann LeCun @ylecun
710K Followers 718 Following Professor at NYU. Chief AI Scientist at Meta. Researcher in AI, Machine Learning, Robotics, etc. ACM Turing Award Laureate.viridis
(Oh, by the way, I joined @AnthropicAI a few months ago to help build reliable and transformative AI products. It's great here.)
Claude did everything. I didn't write a single line of python. Even for debugging, I just pasted the errors into the chat and Claude gave me updated code. This level of reliability and problem solving was impossible until a few months ago with the latest generation of frontier…
the vc money pouring into ai is an annoying bubble sure, but there's also the bubble of people who don't realise that the world is about to change faster than they've ever experienced?
@andersonbcdefg trishume perf optimizations rly just killed like a hundred startups
Exciting news for @AnthropicAI devs: Today we are launching tool use in an open beta through our API. Tool use dramatically expands the frontier of possibilities of what you can build with Claude. To learn more, check out our tool use guide: docs.anthropic.com/claude/docs/to…
what if, and bear with me here, we gave claude tools? Capabilities aside (marvelous though they are), we now have credible... "clools".
Exciting news for @AnthropicAI devs: Today we are launching tool use in an open beta through our API. Tool use dramatically expands the frontier of possibilities of what you can build with Claude. To learn more, check out our tool use guide: docs.anthropic.com/claude/docs/to…
Tool use is now available in beta to all customers in the Anthropic Messages API, enabling Claude to interact with external tools using structured outputs.
To the person at @AnthropicAI who decided Claude's native tongue will be XML: Thank you 🙏 To the one who let the users pre-fill the assistant's responses: 👑 To the one who thought the API should let you know which stop sequence ended the generation: 🐐
tristan is top-three best engineers i've worked with and a lot of the people he's hired recently are not very far behind. _obscenely_ high talent concentration what's worse, they're nice people and easy to get on with
Here's Claude 3 Haiku running at >200 tokens/s (>2x as fast as prod)! We've been working on capacity optimizations but we can have fun testing those as speed optimizations via overly-costly low batch size. Come work with me at Anthropic on things like this, more info in thread 🧵
Here's Claude 3 Haiku running at >200 tokens/s (>2x as fast as prod)! We've been working on capacity optimizations but we can have fun testing those as speed optimizations via overly-costly low batch size. Come work with me at Anthropic on things like this, more info in thread 🧵
Had so much fun chatting with my friends @TrentonBricken and @_sholtodouglas. No way to summarize it, except: This is the best context dump out there on how LLMs are trained, what capabilities they're likely to soon have, and what exactly is going on inside them. You would be…
Opus as main character with a swarm of Haikus. Gandalf and the hobbits. Snow White and the Dwarves. Necromancers commanding undead army. Haiku as main character calling out to Opus when needed. Sherlock & Mycroft. Cinderella & Fairy Godmother. The Davy Jones stories my dad told
Y'all pray for our inference team🙏🙏 They are the heroes fighting fires to keep Claude up and running.
It's insanely overlooked just how cheap and fast Claude 3 Haiku is. Here's an example of what I mean: It’s estimated that the average person speaks roughly 860 million words in their life. That’s about 1.2 billion tokens in LLM-speak. Haiku costs just $.25 per million tokens…
Folks, I left Meta yesterday. This time there were no layoffs or anything thrilling involved. In fact, I am more bullish than ever about Meta with the company’s increased focus on AI. But given the incredible competitive pressure in the field, there is really no advantage to be…
@perrymetzger Every false dichotomy is making our collective epistemology more stupid and preventing us from having a grand future by conflating coordinating against foreseeable risks with being anti-progress.
@thomasahle @NormalComputing thanks, but not news to me. GPT 4 struggles with the simplest Q&A over PDFs. However, Opus is virtually 100% accurate over 100+ pg PDFs in my real world use cases
I don’t think people have fully internalized yet how cheap and how good Haiku is. Opus gets all the press for good reason, but Haiku pushes the intelligence / cost boundary even more imo
Kinda bonkers, but this agent workflow only cost me ~$0.09 (9 cents) It's about ~ 350k tokens, and the outputs are on par with gpt-4, at a fraction of the price. GPT = $10/1m = $3.50 Haiku = $0.25/1m = $0.09 40x cheaper. This wasn't even possible 1 month ago.