Global Retail Startups Lead | RETHINK Retail Top Retail Expert 2024 | RTIH Top 100 Retail Technology Influencers | Advisory Board GMU Center for Retail Transformation | Mentor at Unlimited Awesome
Because they use less computing power, small models can answer questions for as little as ๐จ๐ง๐-๐ฌ๐ข๐ฑ๐ญ๐ก ๐ญ๐ก๐ ๐๐จ๐ฌ๐ญ ๐จ๐ ๐ฅ๐๐ซ๐ ๐ ๐ฅ๐๐ง๐ ๐ฎ๐๐ ๐ ๐ฆ๐จ๐๐๐ฅ๐ฌ in many cases. Experian shifted from large models to small for their #AI chatbots for financial advice and customer service. Once trained on the companyโs internal data, the smaller models performed as well as large ones at a fraction of the cost: https://lnkd.in/gVN_nvXz (thank you Eric Best for the pointer & for the amazing conversation last week. Your high energy level & contagious enthusiasm continues to be the same as it was 20+ yrs ago ๐) The rise of SLMs represents an exciting development in the field of #AI, and it will be interesting to see how they shape the future of #retail and other industries: https://lnkd.in/gMBp8ehM SLMs consume significantly less computational power, leading to lower energy usage and reduced carbon emissions. This makes them more environmentally friendly compared to LLMs ๐โป๏ธ๐๏ธ๐๐ฑ Microsoft for Startups Pegasus partner Snorkel AI focuses on optimizing and customizing language models for enterprise use. One of their key offerings is the ability to distill large language models (LLMs) into specialized โsmall language modelsโ (SLMs).ย These SLMs are designed to improve task-specific accuracy while significantly reducing cost. #startups #AI #GenAI #LLM #SLM #msft4startups Tom Davis, Panteha Healey, Cathy Zhu, Sally Ann Frank, Ted Roduner, Heena Purohit
Good stuff ShiSh S.! Seems like Reactor โข Intelligent Data Pipeline needs a native Snorkel AI enterprise data orchestration (Alexander) to get that platform into the hands of our customers like https://reactordata.com/hearst-customer-success/ ๐
The shift towards small language models is quite intriguing. I find the balance between cost efficiency and environmental benefits particularly compelling. It's great to see how these advancements are shaping the future of AI and retail. Thanks for sharing these insights! ๐ฑ
Global Retail Startups Lead | RETHINK Retail Top Retail Expert 2024 | RTIH Top 100 Retail Technology Influencers | Advisory Board GMU Center for Retail Transformation | Mentor at Unlimited Awesome
1wAn interesting case study from SLB (also known as Schlumberger): The world leading offshore drilling services company used #AI to extract 70+ years of text data to enhance proactive well management. Using Snorkel AI's solution, they built an AI application (in just 3 days) to process PDFs of well report and extract relevant information to perform analysis for client reports. ๐๐๐ ๐ผ๐ ๐๐ฅ๐ฅ๐ก๐๐๐๐ฉ๐๐ค๐ฃ ๐๐ช๐๐ก๐ฉ ๐ฌ๐๐ฉ๐ ๐๐ฃ๐ค๐ง๐ ๐๐ก ๐๐ก๐ค๐ฌ ๐ง๐๐๐ช๐๐๐ ๐ฉ๐๐ ๐ฅ๐ง๐ค๐๐๐จ๐จ๐๐ฃ๐ ๐ฉ๐๐ข๐ ๐ค๐ ๐ง๐๐ฅ๐ค๐ง๐ฉ๐จ ๐๐ง๐ค๐ข ๐๐๐ฉ๐ฌ๐๐๐ฃ 1 ๐ฉ๐ค 3 ๐๐ค๐ช๐ง๐จ ๐ฅ๐๐ง ๐ง๐๐ฅ๐ค๐ง๐ฉ ๐ฉ๐ค ๐๐ช๐จ๐ฉ ๐ ๐๐๐ฌ ๐จ๐๐๐ค๐ฃ๐๐จ ๐ฅ๐๐ง ๐ง๐๐ฅ๐ค๐ง๐ฉ: https://snorkel.ai/how-schlumberger-uses-snorkel-flow-to-enhance-proactive-well-management/