#AIChatter #PythonInExcel #meta #nvidia #llmmodels #llama #DataScience #AI #TechInsights
Welcome back to AI Chatter!
Join us as we follow alongside Ken, Marty and Canon as they share their insight into the latest AI trends and insights. In Episode 14 of AI Chatter, the hosts discuss Meta's significant increase in computing power requirements for training Llama 3.1, highlighting the company's move from 8 billion to 450 billion parameters. The conversation also covers the integration of Python into Excel, offering new possibilities for users to perform complex data analysis within their familiar environment.
Don't miss out – subscribe now and join the conversation!
Watch Our Previous Episode: https://youtu.be/LhUun40PiyY
Watch AI Chatter Series Playlist: https://www.youtube.com/watch?v=0Wkwfl1x_CY&list=PL3dqJmcJrJs1ZlDwr4WB3hnb2Hl5zSG79&pp=iAQB
=============================
Subscribe and Listen to our Archetypical Podcast:
YT: ➡︎ https://www.youtube.com/channel/ https://www.youtube.com/watch?v=21lCUGKTSaw
Apple: ➡︎ https://podcasts.apple.com/us/podcast/archetypical/id1665285587
Spotify: ➡︎ https://open.spotify.com/show/00RoJcYYD3jlOi26AN7Fns
Question - Would uses would you have for Python in Excel?
📒 Show Notes, Links and Resources 📒
Zuckerberg says Meta will need 10x more computing power to train Llama 4 than Llama 3: https://techcrunch.com/2024/08/01/zuckerberg-says-meta-will-need-10x-more-computing-power-to-train-llama-4-than-llama-3/
Anaconda Debuts New Solution to Run Python Locally in Microsoft Excel: https://www.datanami.com/this-just-in/anaconda-debuts-new-solution-to-run-python-locally-in-microsoft-excel/
Disclaimer: This video is not sponsored.
Timecodes ⏰
0:00 - AI Chatter Intro
0:36 - Introductions: Ken, Marty & Canon
0:57 - Canon Pattillo
01:05 - Marty Gaupp
01:19 - Meta 10x Power for Llama 3.1
02:00 - Why all this power?
02:09 - 405 Billion Parameters
03:20 - Nvidia Blackwell chips encounters delays
03:42 - What's the cost on this Canon?
04:12 - Pascal's wager
04:55 - Where are they going to deploy these models?
05:36 - Where else can Facebook use this?
06:06 - Is it hubris?
07:02 - Nvidia laughing to the bank selling more machines
08:28 - Marty's excitement for Python for Excel
09:09 - Is this a useful tool?
10:26 - Practical application of business technology
11:50 - Does this open up for a new user base?
12:55 - Is this a mini app deployment?
15:23 - Final thoughts
15:48 - Outro
=============================
Connect with us!
X: ➡︎ https://twitter.com/ArchetypeConsul
IG: ➡︎ https://www.instagram.com/archetypeconsulting/
Archetype LinkedIn: ➡︎ https://www.linkedin.com/company/archetypeconsulting
=============================
About:
At Archetype we build a foundation that leverages data as a competitive advantage.
===========================
Connect with us on Social Media:
➡︎ Kenneth Sanford - https://www.linkedin.com/in/ekenomics/
➡︎ Marty Gaupp - https://www.linkedin.com/in/martygaupp/
➡︎ Canon Pattillo - https://www.linkedin.com/in/canonpattillo/
➡︎ Archetype - https://www.linkedin.com/company/archetypeconsulting
===========================
What We Use For Video, Image, & Music Content: https://www.storyblocks.com