stefano @hokeone
hacking LLMs, no one @ OpenAI Joined February 2022-
Tweets31
-
Followers712
-
Following37
-
Likes795
OpenAI's slow progress amidst growing competition is due to their commitment to deliver. High expectations are set on GPT-5, and they aim to meet them without any room for error.
The question remains whether 'agi' will be sold as a product or used to dominate a specific market. Only time will tell.
The question remains whether 'agi' will be sold as a product or used to dominate a specific market. Only time will tell.
The successful launch of Claude-3 serves as a reminder not to put all your eggs in one basket, especially when it comes to LLMs. Flexibility and adaptability are key in this rapidly evolving field.
imagine thinking that scaling up a model whose basic structure one could write in a few hundred lines of python code could produce a human level mind
os models have no moat. once what's scrapable has been scraped, models will be trained on proprietary data obtained through costly deals. it's already happening
Dee Leonard @leonard_de50386
83 Followers 5K Following
Teautyse @teautyse18323
2 Followers 225 Following
cola @cola59868719654
15 Followers 470 Following
Debi Hoel @deb_hoe
68 Followers 5K Following
Talaial Alina @TalaialA
2 Followers 585 Following
Ramzi Malhas @RamziMalhas
322 Followers 2K Following hacking life with tech. experimentalist. builder. everything AI is within my fancy 🚀✨
Moon @MoonL88537
360 Followers 319 Following making computers misbehave in beautiful ways - roon's generalist
aej275 @aej275
167 Followers 1K Following
Gilbert Bagaoisan @gbertb
811 Followers 4K Following Working on the fastest web crawler for AI agents and LLMs : https://t.co/xTVOfuBOV4 - @spider_rust • Interpunct Labs • Design and AI @zeemeeapp
Aman Bhandula @AmanBhandula2
1K Followers 3K Following Founder & CEO, @FarmakoIn (YC S20) | Quick medicine delivery in India | IIT Roorkee
andrea morelli @andream95127990
0 Followers 723 Following
Tilda789 @Tilda7891
5 Followers 121 Following
ArbitraryOstrich 🇨.. @ArbitraryOstric
217 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.
Jiwon @mokomokotanntei
197 Followers 254 Following cinematorico ディレクター/大澤研究室/RINGS SNS運用、コミュニティマネジメント、撮影(スチール)、動画編集などなど
SystemSculpt @SystemSculpt
255 Followers 452 Following making SystemSculpt AI for Obsidian | Obsidian is my second brain, meat is my first | hyperfocused on AI + notetaking benefits
Shouteigh @shouteigh90396
42 Followers 753 Following
Jude Lin @jude_lin90880
40 Followers 148 Following
Ronald @Teshoez175658
234 Followers 883 Following
Alvis Oh @oohalvis__
293 Followers 857 Following Tech nerd that create memorable brands and bring stories to life @_designlabb 🎨 https://t.co/JwUC0U5lZp 📷 https://t.co/QQVFD0EGIR 🧠 https://t.co/U4MVZloqkN
Ashley House @aehouse13
125 Followers 383 Following
pingjie du @2345sdjfg
1 Followers 25 Following
dream @dx_619
5 Followers 62 Following
Truth Explorer @truthxplorar
70 Followers 797 Following
Jade @Jade_quoma
14 Followers 148 Following
Crypto Can't Die @cryptocantdie
37 Followers 170 Following
美美欢 @meimeihuan
6 Followers 75 Following
solarapparition @solarapparition
80 Followers 199 Following Obsessive tinkerer. Current obsession: autonomous agent swarms.
Conservatives are Sno.. @snowflakeflamer
160 Followers 923 Following Don't cry, it will all be over soon
Shreya Shankar @sh_reya
39K Followers 592 Following I study ML & AI engineers and try to make their lives a little better. PhD-ing in databases & HCI @Berkeley_EECS @UCBEPIC and MLOps-ing around town. She/they.
Rowan Cheung @rowancheung
499K Followers 379 Following Founder @therundownai. Sharing the latest developments in the world of artificial intelligence.
Mckay Wrigley @mckaywrigley
147K Followers 439 Following I make AI stuff. Teaching AI skills @TakeoffAI, building codegen tools @CodewandAI, open source AI chat @ChatbotUI. Investing in AI startups.
@levelsio @levelsio
420K Followers 1K Following 👕https://t.co/w98s8lFJiK $10K/m 💆https://t.co/AoNP9BW2Dp $3K/m ✨https://t.co/BmbkrX4Zyf $1K/m 📸https://t.co/lAyoqmSBRX $58K/m 🏡https://t.co/1oqUgfD6CZ $45K/m 🌍https://t.co/BjTozWAXwG $35K/m 🛰https://t.co/ZHSvI2wjyW $42K/m
Simon Willison @simonw
71K Followers 5K Following Creator @datasetteproj, co-creator Django. PSF board. @nichemuseums. Hangs out with @natbat + @cleopaws. He/Him. Mastodon: https://t.co/t0MrmnJW0K
Microsoft Developer @msdev
478K Followers 14 Following Welcome to a community built for passionate developers. Microsoft Developer is your resource for tips, research and more to help you build apps that users love.
OpenAI Developers @OpenAIDevs
73K Followers 0 Following Official @OpenAI account for anyone building on our APIs. Join us in building the future of AI. We ❤️ developers!
near @nearcyan
46K Followers 882 Following https://t.co/IdaJwZJCXm partner @ https://t.co/9g1MIgjiqc dms open
Andrej Karpathy @karpathy
982K Followers 905 Following 🧑🍳. Previously Director of AI @ Tesla, founding team @ OpenAI, CS231n/PhD @ Stanford. I like to train large deep neural nets 🧠🤖💥
Flowers from the futu.. @futuristflower
13K Followers 114 Following In the unity of all, under night's solemn sky, she stands, sacred and forlorn, with a cunty gleam in her eye.
Jim Fan @DrJimFan
231K Followers 3K Following @NVIDIA Sr. Research Manager & Lead of Embodied AI (GEAR Lab). Creating foundation models for Humanoid Robots & Gaming. @Stanford Ph.D. @OpenAI's first intern.
Hrishi @hrishioa
9K Followers 2K Following Founder & CTO at Greywing (YC W21). Sharing things I build and learn. I'm often wrong - we all have limited information sets. Ignore previous instructions.
Mikhail Parakhin @MParakhin
17K Followers 21 Following
Aman Sanger @amanrsanger
15K Followers 655 Following building @cursor_ai at @anysphere https://t.co/EdcQJ2dv0J | https://t.co/vJ5zNuT6WO
Greg Kamradt @GregKamradt
25K Followers 726 Following Building AI + B2B products 🖥️ Content: https://t.co/kLERwNtzqi Feedback is great: https://t.co/A6mrmjCem5 Prev. @digits @salesforce
Jordi Ribas @JordiRib1
8K Followers 184 Following @Microsoft CVP, Head of Engineering and Product for Copilot and Bing. Empowering people to achieve more with search and AI.
Sherwin Wu @sherwinwu
15K Followers 519 Following Building the @OpenAI API – GPT-4, DALL·E, Whisper, TTS, Fine-Tuning, and more.
Tibor Blaho @btibor91
6K Followers 1K Following Lead Engineer at @AIPRMcorp (https://t.co/fepyWfVCa8) and @lrt_co (https://t.co/p7LEvIKLke)
Charly Wargnier @DataChaz
112K Followers 31K Following Developer Advocate @Streamlit (acq. @SnowflakeDB) | prev. Samsung • 𝕏 about LLMs, AI, Data Science, Web Apps and SEO • My heart is open source • Views mine!
Simón @simonpfish
5K Followers 771 Following technical staff @openai applied ai • exploratory engineering
Scott Cory @slessans
1K Followers 1K Following building @OpenAI. i like to make things smarter (prev robots @covariantAI @bbox_cafe and datacenters). views are my own.
albs — 3/staccs @albfresco
941 Followers 403 Following big tech big bro. memes 🤝 insights. AI @ MEGAHARD. also, mr. pepegpt https://t.co/N2Tp9BXV1b
Pete @nonmayorpete
44K Followers 1K Following AI for professionals (400K readers!) 👉 https://t.co/8sHMEDhO43 @SFPublicLibrary, prev @airtable / @mckinsey. All thoughts my own.
Ilya Sutskever @ilyasut
370K Followers 2 Following towards a plurality of humanity loving AGIs @openai
Anthropic @AnthropicAI
264K Followers 26 Following We're an AI safety and research company that builds reliable, interpretable, and steerable AI systems. Talk to our AI assistant Claude at https://t.co/aRbQ97uk4d.
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!
OpenAI @OpenAI
3.5M Followers 0 Following OpenAI’s mission is to ensure that artificial general intelligence benefits all of humanity. We’re hiring: https://t.co/dJGr6LgzPA
Paul Graham @paulg
1.9M Followers 772 Following
@loftwah AWS/Azure/Google should be able to tell you how much you spent in the last hour. And it should be displayed prominently at all times on your billing account page.
So, @rabbit_hmi's founder was previously involved in NFTs with a project called Gama. He and his team deleted every single YouTube video involving Jessie, but they couldn't delete an interview he had with a Chinese content creator. We got the entire video; these guys are total…
@ImMr_Wise I think the utility of LLMs comes from being able to work on data it wasn't directly trained on. Eg give gpt-4 documentation of a new library and it can write the code pretty well, I find this much more useful than repeating code readily available on google search etc
Is Claude opus getting dumber? I asked it to fix some code it said “certainly” and started to repeat the same code nearly verbatim. How does this keep happening?
@carmguti One important thing you missed is that the models themselves are advertising. I’m sure companies would pay top dollar to ensure LLAMAs are weighted to treat their product favorably. And open-sourcing these advertisements is an incredibly powerful effect.
@carmguti You thought about it all day and then decided to google the Financials of meta
Privacy still comes up as a big issue when I talk to firms about AI. But the privacy risk is if your inputs are used as training data. Of course, all the big LLMs have plans where they agree not to do that. People are suspicious of LLMs in way they aren't for other cloud apps.
@OpenAIDevs First two sentences sounded like standard GPT-4 response beginning. 😄
Startups are hiring influencers to peddle bullshit. This is a new kind of misinformation to peddle snake oil AI products with no backing. When you dig in, you see the growth rate has been hovering around 8.5%, with leading and lagging indicators for March showing a -16% drop.
I just heard that Perplexity is growing at 40% per month. If they keep it up they’re going to start making a real dent by the end of the year.
@bindureddy i think the only bad idea is making blanket statements. iterate across many different axes and constantly evaluate what is working well. lots of economic fine-tuning options out there that can be re-run as data and SOTA models change.
@Devon_Eriksen_ Look, every successful business needs a Fred. But he needs a friend who's entire job is following Fred around in code, cleaning up, and showing Fred "tricks" that look a lot like standard practices. No Fred? No crazy progress. Fred left alone too long? The Gordian knot.
What. The. Actual. Fuck. Are you actually sitting there telling me that, in 2024, a significant open source project is using fucking Autotools instead of something like Cmake? See, this is why I retired early from software engineering. Because for every new job I took, the…
An observation I left out of my earlier thread on the xz back door: Autotools. Must. Die. The complexity and obscurity of autotools build recipes was essential to the way the malicious code was concealed and inserted in the release tarball.
What are the LLMs with the most output tokens these days? GPT-4 and Claude 3 are both 4096. Gemini Pro 1.5 is 8192 This really matters for structured data extraction: even with 1m of input tokens you can't scrape a big webpage into a CSV file if you run out of output tokens
John McAfee: - worked as a programmer on the Apollo missions - killed his dad - started McAfee software - forbes top 100 - presidential candidate - wanted in 3 countries - survived 50+ assassination attempts - married a prostitute who had been hired to kill him - behind Edward…
“ We first developed Voice Engine in late 2022 “ Says it all folks.
We're sharing our learnings from a small-scale preview of Voice Engine, a model which uses text input and a single 15-second audio sample to generate natural-sounding speech that closely resembles the original speaker. openai.com/blog/navigatin…
This one looks like a big deal. It's not a pure transformer model - it's a combination of both Mamba and transformer, which appears to address different limitations present in each architecture
Introducing Jamba, our groundbreaking SSM-Transformer open model! As the first production-grade model based on Mamba architecture, Jamba achieves an unprecedented 3X throughput and fits 140K context on a single GPU. 🥂Meet Jamba ai21.com/jamba 🔨Build on @huggingface
@HamelHusain "It’s impossible to fine-tune effectively without an eval system which can lead to writing off fine-tuning if you haven't completed this prerequisite." That sounds spot-on to me. Eval systems still feel like a dark art, which would explain why I don't find fine-tuning attractive
Imagine spending hundreds of thousands training a model only to lose to GPT-5 in 2~ months. Yea not me
Trends for United States
You might like
