Hi, I am Louis-François (or just Louis, loo·ee, French pronunciation), from Montreal, Canada, also known as "What's AI".
Making AI accessible. I aim to demystify the AI “black box” for everyone and sensitize people about its potential and risks.
Other than RU-vid and my podcast, I am the co-founder of Towards AI, and ex-PhD at Mila.
I also publish articles on my blog every week if you enjoy short reads: www.louisbouchard.ai/, and have my updates on the newsletter www.louisbouchard.ai/newsletter/.
Follow @whats_ai on Instagram, Twitter, and connect on LinkedIn! Come exchange with me and follow my projects!
►Join our discord server with many other AI enthusiasts: www.louisbouchard.ai/learn-ai-together/ ►If you'd like to support me, I have a Patreon where you can do that: www.patreon.com/whatsai
AI algorithms facilitate better decision-making in business by providing actionable insights from data analysis.This enhances strategic planning and operational efficiency.
They can do many things in RAG like using various tools to do different actions as in other pipelines or like here having a router (could be called an agent) determining which type of dataset to query (graph or vector db) or to use what kind of keywords or complement user queries before retrieval, etc!
Haha check other comments! I did in fact. It just has a very different name in French! I still watched it again after the episode (in English) and really loved it!
Hmmm so now some AI needs to be pushed to the database search to make it useful, don't think a vector search can do the trick well enough. Has anybody used it in real escenarios?
Hello. Can you inbox me. I have a small start up data Annotation team in Uganda East Africa. Let's talk business at the best cost effective manner that saves you alot.
Actually most approaches uses an LLM first to understand the query, improve it for retrieval or reject it before doing any kind of retrieval. Plus, this query needs to be embedded which here is simplified as within the LLM part too :)
Great touch into model's inside view. This is what I was looking into. People often go for directly applying LLM's into finetuning for their usecase. But a few thinks about how this transformer performs the pattern understanding . I am looking into understanding the patterns
Haha sounds good! We also have a free course series called Gen ai 360 if you want more free content before taking the chance :) And tons of articles and videos of course
Why do I feel 100% charged I mean my eyes automatically open after 9 hours of sleep. Sometimes I go to bed at 12 Am and wake up at 12 PM😂 and still feel dazy and sleepy.
Has anyone tried to build a diffusion LLM yet? Proly a silly idea but just curious. I see no reason why it wouldn't turn out just as nuts as the image models (which are getting insanely good).
It depends on the code implementation of the system! Most will put in place a system to detect it and summarize or extract key points to make it shorter.
My latest experience with RAG was using it in RU-vidr's AI Telegram bot. I prepared a lot of AI processed data of him, with fine-grained segments. The prompt itself is also crafted according to the latest Claude prompt discovery. He is a life coach and it's the most amazing AI experience for me yet. JulienHimselfBot
It'll be interesting to see how far the models can go in extrapolating beyond the current state of human knowledge. Will they be genius' or a delusional crackpots? (Likely a bit of both given the source of the original data set... humans.) 🖖😎👍
Oh wow, I completely forgot to share any table and results. Sorry about that. Since you can already use it I assumed (like GPT4) that most people would try it or heard of it. An error on my end, but you can see it there: build.nvidia.com/nvidia/nemotron-4-340b-instruct
References: ► Watch the full video: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-88Pl9zD1Z78.html ►Read the full article: www.louisbouchard.ai/neroic/ ►Kuang, Z., Olszewski, K., Chai, M., Huang, Z., Achlioptas, P. and Tulyakov, S., 2022. NeROIC: Neural Rendering of Objects from Online Image Collections. arxiv.org/pdf/2201.02533.pdf ►Project link with great video demo: formyfamily.github.io/NeROIC/ ►Code: github.com/snap-research/NeROIC ►My Newsletter (A new AI application explained weekly to your emails!): www.louisbouchard.ai/newsletter/
There are many explanations of the theory in AL online, but there is a lack of code to show practical implementation. Please provide an explanation of the code in the future if possible
The guy in the bottom of the screen contributed so much to me understanding what RAG is. Thank you so much. It must be difficult being burdened with all that talent and just being surrounded by all these people that don't do anything and just stitch themselves onto your content while nodding and not even being from that video. Just literally paste clip edited on to hundreds of clips because you know that you're so untalented and your content is so bland and copy paste that the only way you stand the foggiest chance in hell of earning more than $2 a month off this platform is to machine gun blast a quadrillion videos in the hopes that the algorithm makes a mistake and accidentally recommends your video to people en masse Don't you hate those people guy, at the bottom of the screen?
This is a clip created from a podcast I’ve recorded with Jerry. Not sure what you mean with stitching faces into content. As one would say on Reddit, “username checks out”.
