dida Machine Learning’s Post

"* Language is low bandwidth: less than 12 bytes/second. A person can read 270 words/minutes, or 4.5 words/second, which is 12 bytes/s (assuming 2 bytes per token and 0.75 words per token). A modern LLM is typically trained with 1x10^13 two-byte tokens, which is 2x10^13 bytes. This would take about 100,000 years for a person to read (at 12 hours a day). * Vision is much higher bandwidth: about 20MB/s. Each of the two optical nerves has 1 million nerve fibers, each carrying about 10 bytes per second. A 4 year-old child has been awake a total 16,000 hours, which translates into 1x10^15 bytes. In other words: - The data bandwidth of visual perception is roughly 16 million times higher than the data bandwidth of written (or spoken) language. - In a mere 4 years, a child has seen 50 times more data than the biggest LLMs trained on all the text publicly available on the internet. This tells us three things: 1. Yes, text is redundant, and visual signals in the optical nerves are even more redundant (despite being 100x compressed versions of the photoreceptor outputs in the retina). But redundancy in data is *precisely* what we need for Self-Supervised Learning to capture the structure of the data. The more redundancy, the better for SSL. 2. Most of human knowledge (and almost all of animal knowledge) comes from our sensory experience of the physical world. Language is the icing on the cake. We need the cake to support the icing. 3. There is *absolutely no way in hell* we will ever reach human-level AI without getting machines to learn from high-bandwidth sensory inputs, such as vision. Yes, humans can get smart without vision, even pretty smart without vision and audition. But not without touch. Touch is pretty high bandwidth, too." https://lnkd.in/emaUTrjr

Let's jump into Part 3 of our TEKnically Speaking series topic of Retrieval-Augmented Generation! How do you train these language models without putting your proprietary data at risk? Check it out here: https://hubs.la/Q02tdxz30 Mark Campbell, Chief Innovation Officer, EVOTEK Ned Engelke, Chief Technology Officer, EVOTEK EVOTEK Labs #artificialintelligence #innovation #infrastructure #ai

Let's jump into Part 3 of our TEKnically Speaking series topic of Retrieval-Augmented Generation! How do you train these language models without putting your proprietary data at risk? Check it out here: https://hubs.la/Q02tdPkQ0 Mark Campbell, Chief Innovation Officer, EVOTEK Ned Engelke, Chief Technology Officer, EVOTEK EVOTEK Labs #artificialintelligence #innovation #infrastructure #ai

The new thing with the new thing - RAG is adding some much needed features to GenAI. #emergingtechnologies #artificialintelligence

In a world fuelled by AI advancements, leveraging Language Models like ChatGPT, BERT, and LLaMA is inevitable. But with great power comes great responsibility! Discover the nuances of using LLM responses in our latest blogpost. Learn how to navigate misinformation, bias, and privacy concerns while maximising efficiency. Dive in now!

Let's jump into Part 3 of our TEKnically Speaking series topic of Retrieval-Augmented Generation! How do you train these language models without putting your proprietary data at risk? Check it out here: https://hubs.la/Q02tdC-S0 Mark Campbell, Chief Innovation Officer, EVOTEK Ned Engelke, Chief Technology Officer, EVOTEK EVOTEK Labs #artificialintelligence #innovation #infrastructure #ai

Hey peeps!🎉 Just share something about LLM and its Applications. Because Pathway x GTech MuLearn conducting the BootCamp of LLM. So, dive into it! A Large Language Model (LLM) is an advanced artificial intelligence algorithm that employs neural networks with a vast number of parameters to perform a variety of natural language processing tasks. Applications:- 📌 Text Generation: Creating coherent and contextually relevant content. 📌 Machine Translation: Translating text between different languages. 📌 Summary Writing: Condensing long pieces of text into shorter summaries. 📌 Image Generation from Texts: Visualizing descriptions as images. 📌 Machine Coding: Generating and understanding code. 📌 Language Translation: Facilitating communication across language barriers. 📌 Chatbots and Conversational AI: Engaging in dialogue with users. These are simple descriptions of LLM and its applications. Want to know more about LLM? Visit here : https://lnkd.in/gSHs6aPM #Pathway #GTechMuLearn

RAFT: A new way to teach LLMs to be better at RAG “Retrieval-Augmented Fine-Tuning” combines the benefits of Retrieval-Augmented Generation and Fine-Tuning for better domain adaptation

Are you ready to dive back into our topic of Retrieval-Augmented Generation (RAG)? This week's episode of TEKnically Speaking focuses on how our customers can utilize a private language model. Check it out here: https://hubs.la/Q02szTG10 Mark Campbell, Chief Innovation Officer, EVOTEK Ned Engelke, Chief Technology Officer, EVOTEK #artificialintelligence #innovation #infrastructure #ai

Find out how new AI architectures like RAG and RAG 2.0 will impact your underlying infrastructure. As my dad used to say, "There's a lot of friggin' in the riggin'" #emergingtechnologies #artificialintelligence