EXAM FILES

Tesla Optimus Thread the Needle Test

Elon: "One of the test goals I have is can it pick up a needle and a thread and thread the needle, just by looking." Lex Fridman: "How far away are we from that?" Elon: "Maybe a year."

#ElonMusk | Source | @ByteShare

Channel logo changed 🤓 🫡 Thank you all for your support!

LLM OS. Bear with me I'm still cooking. - Andrej Karpathy

Specs: - LLM: OpenAI GPT-4 Turbo 256 core (batch size) processor @ 20Hz (tok/s) - RAM: 128Ktok - Filesystem: Ada002

#OpenAI | Source | @ByteShare

You can now operate robots by just thinking about it. With your brain signals. WOW. This robot system from Stanford has so much sci-fi vibe and wild implications that I don't even know where to start. NOIR decodes the EEG signal from your head into a library of robot skills. It is demonstrated on 20 household activities, such as cooking Sukiyaki, ironing clothes, grating cheese, playing Tic-Tac-Toe, and even petting a robot dog! NOIR learns to predict your intended goals in advance, so that your thinking efforts (literally) can be reduced to a minimum. It works with both adults and children as young as five years old. Imagine scaling up this system to Neuralink Humanoid robots like Tesla Optimus Bot in the future! There's now a decent probability that we can see the movie Avatar become a reality in our lifetime.

Website: noir-corl.github.io Paper: arxiv.org/abs/2311.01454

This work is from Stanford Vision Lab. Congrats on the phenomenal idea! #Robotics #AI | @ByteShare

Elon Musk replied to Grōk's question.

Human-AI symbiosis, for those who want it, is the optimal outcome - Elon Musk

#Grok #ElonMusk | Source | @ByteShare

Omegle 2009-2023 🪦

Omegle shuts down after 14 years visit : omegle.com

#Omegle | Source | @ByteShare

Most popular social platforms by monthly active users: 🇺🇲 Facebook: 3 Billion 🇺🇲 YouTube: 2.49 Billion 🇺🇲 WhatsApp: 2 Billion 🇺🇲 Instagram: 2 Billion 🇨🇳 WeChat: 1.32 Billion 🇨🇳 TikTok: 1.21 Billion 🇺🇲 Facebook Messenger: 1.03 Billion 🇺🇲 LinkedIn: 930 Million 🇦🇪 Telegram: 800 Million 🇺🇲 Snapchat: 750 Million 🇨🇳 Douyin: 743 Million 🇨🇳 Kuaishou: 673 Million 🇺🇲 X/Twitter: 666 Million 🇨🇳 Sina Weibo: 599 Million 🇨🇳 QQ: 571 Million 🇺🇲 Pinterest: 465 Million 🇺🇲 Reddit: 430 Million 🇺🇲 Quora: 300 Million 🇺🇲 Skype: 300 Million 🇨🇳 Baidu Teiba: 300 Million 🇺🇲 Teams: 300 Million 🇱🇺 Viber: 260 Million 🇯🇵 Line: 230 Million 🇺🇲 imo: 200 Million 🇺🇲 Discord: 196 Million 🇮🇳 Josh: 153 Million 🇸🇬 Likee: 150 Million 🇺🇲 Twitch: 140 Million 🇷🇺 VK: 100 Million

Note: according to Kepios analysis of company announcements of monthly active users; platforms self-service advertising resources, analysis. As of October, 2023

#SocialMedia #Users | @ByteShare

GPTs are a new way for anyone to create a tailored version of ChatGPT to be more helpful in their daily life, at specific tasks, at work, or at home — and then share that creation with others. No code required. Plus and Enterprise users can start creating GPTs this week at chatgpt.com/create. Later this month, OpenAI will launch the GPT Store, so people can feature and make money from their GPTs. They plan to offer GPTs to more users soon. #OpenAI #GPTs | @ByteShare

Grok now available on 𝕏 @grok handle #xAI #Grok | Source | @ByteShare

OpenAI Developer Day Announcement and Implications for Open-Source AI development

OpenAI's developer day was filled with a bunch of annoncements. Most notable amongst them was a 3x price cut on GPT-4-turbo, Custom GPTs and Assistants APIs Sadly and predictably, there was no new GPT-5 announced. Open AI is rumored to have tried to train a new version of GPT-4, Arrakis, earlier this year but the model didn't run as efficiently and therefore had to scrap it Rumors are they are training, Gobi, a multimodal modal (GPT-5). that will be way superior to GPT-4. It's not clear when Gobi will be available but I suspect it won't be until next year. So let's review the announcements made by them and understand how it impacts the eco-system Bigger context length and price-cuts on GPT-4-Turbo - While the 128K context length is somewhat interesting, LLMs are notorious for forgetting stuff in long context inputs! It's a huge pain to make long contexts work efficiently and you have to repeat the same instructions multiple times. Fundamentally, this just seems like they are one-upping Claude which was superior to OpenAI when it came to summarizing large PDFs. The price cut while somewhat interesting is pretty much a nohup. Open source fine-tunes that match GPT-4 performance were available for 60x cheaper than GPT-4. Now they are 20x cheaper!! 20x is still a very substantial number! For simple use-cases, open-source will MATCH GPT-4 performance out of the box - i.e. no fine tuning required! I don't see a case for anyone to use OpenAI at scale for those use-cases. As we innovate and iterate on open-source models, we may not even need the "fine-tuning" to match GPT-4 performance even for the most complex use-cases. Every day, a new open-source model is on top of the charts and right now open-source is just behind Claude-2 and GPT-4 beating everyone else and for a fraction of the price. Custom GPT store - TBH, you need 10s millions of subscribers before you start worrying about a paid developer eco-system on top of ChatGPT. Like plugins, it seems pre-mature. I also suspect this is to compete with Bard that promises to unify information from all your Google products. If I were OpenAI, I wouldn't worry, no one seems to be using Bard at the moment and unless Gemini is launched and is really good, no one will continue to use it NLP based custom GPTs - This sounds interesting in theory, but won't work in practice, for anything except the most basic of use-cases. GPT-4 makes a lot of mistakes in code generation. We have tried it extensively on the Data Science co-pilot that we use on Abacus AI and our data-scientists are tired of correct it's mistakes. We simply need a much better model! We have had to working on implementing purpose built models for this use-case. Assistants API - This is interesting and somewhat useful. In fact, we are implementing it as part of Abacus AI to compare it against a more modular vector store+orchestrator solution. I suspect it does a good job for simple use-cases. Not so sure how it performs on more complex use-cases. FWIW, it does take aim against all the AI module start-ups including open-source vector store and orchestrator start-ups. Yet to be seen, how powerful the Assistant API is and if it works well on some hard use-cases. We are experimenting with it as we speak and will report back in a week. IMHO, OpenAI is focussing on the wrong things - if they had launched GPT-5 and cut GPT-4 prices by 20-50x, they would still be the far and away leaders in the space - but that may have been too tall an order Plus, if you are really building AGI, what in the world are you doing implementing these features that seem to be directly competitive to Azure 🤯🤯

#OpenAI #GPT4 | @ByteShare

OpenAI is launching GPT-4 Turbo

- 128,000 token context (300 page book). - April 2023 knowledge cutoff. - Higher rate limits (doubling for current customers) - Price: 1 cent for prompt tokens, 3 cents for completion tokens.

#OpenAI #GPT4 | @ByteShare

OpenAI is launching GPTs, custom ChatGPT agents you can create and share with others.

Later in November OpenAI is launching a GPT Store featuring revenue share for Custom GPT Creators.

#OpenAI #GPTs | @ByteShare

"xAI developer tool for improving and understanding large AI models" - Elon Musk

#PromptIDE | Source | @ByteShare

PromptIDE highlights: The user_input() function asks the user to enter a string. You can use that to write a chatbot in four lines of Python. While executing the prompt, you see the tokenization, sampling probabilities, alternative tokens, and aggregated attention masks. #PromptIDE | Source | @ByteShare

PromptIDE, one of the internal tools built to accelerate our work on Grok. It allows you to develop and run complex prompts in the browser using an async based Python library.

#PromptIDE | @ByteShare

Giveaways in Channels and Free Premium

Telegram Channel owners can now launch Giveaways to randomly distribute prizes among their followers. This is great if you want to promote your channel, get new subscribers or reward existing ones.

#Telegram #NewFeature | @ByteShare

Announcing the xAI's PromptIDE

Announcing the xAI PromptIDE The xAI PromptIDE is an integrated development environment for prompt engineering and interpretability research. It accelerates prompt engineering through an SDK that allows implementing complex prompting techniques and rich analytics that visualize the network's outputs. We originally created PromptIDE to accelerate our development of Grok. It has helped us iterate quickly over different prompts and prompting techniques. We are now making the IDE available to members of our Grok early access program. x.ai/prompt-ide

#PromptIDE | Source | @ByteShare

The biggest players in the field of artificial intelligence (AI) have expressed their opposition to paying for copyrighted material as training data for generative AI models. The US Copyright Office has been seeking public input on potential new rules regarding the use of copyrighted content in AI model training, and several major AI companies have submitted their arguments against the proposed changes.

#AI #Copyright | @ByteShare

"Knock, knock, singing telegram!" - Elon Musk

#Telegram | Source | @ByteShare