Good overview of Generative AI providers' growing challenges to scrape the Web for publicly available training data. While courts historically adopted the approach that scraping public sites is generally allowed, complex nature of…lnkd.in/gnpQwySw lnkd.in/gDqpU-9m

Check out my latest article: Generative AI: Entertainment industry - an early adoption primer linkedin.com/pulse/generativ… via @LinkedIn

Good overview of emerging LLM data stack (lnkd.in/gbh9WH6n) from Matt Bornstein @nolefp from Andreessen Horowitz. In my opinion, one of most critical dependencies is not sufficiently called out in this stack: Robust information retrieval (based on…lnkd.in/g_Yqcz7M

Observe.AI (Scale Venture Partners portfolio company developing conversation intelligence AI solutions), unveiled an LLM specifically designed for contact center use case, along with a suite of generative AI products d…lnkd.in/gWCtBz7P lnkd.in/ggaYjFka

Good summary of different techniques for information retrieval and summarization to power LLM-based applications. Robust information retrieval / Knowledge Graph capabilities is the most critical dependency for scalable and useful G…lnkd.in/g44YRfBs lnkd.in/gyJ8Nmfy

Very promising approach of combining LLMs (strong in fluency) with" Knowledge Graphs" (strong in structured data - > increased accuracy), to create a "better together" Geverative AI applications (eg natural language recommendation system). The critical de…lnkd.in/g-C_cU-f

Meta's LLaMA is one of the best open source Large Language Models that rivals top proprietary models like OpenAI's #chatgbt. The main limitation of those OS models is that they can't be used for commercial purposes. Meta's potentia…lnkd.in/g6MHmMqJ lnkd.in/gtYxY5HS

Thanks @itsbobjohnston for inviting me to talk about tremendous promise and risk of #artificialintelligence! lnkd.in/gTX3eV7r

Really enjoyed sharing my insights on disruptive potential of AI in the Enterprise on Executive Council podcast with @itsbobjohnston! #generativeai #largelanguagemodels #artificialintelligence #airevolution lnkd.in/gGgTfPuX

Really enjoyed sharing my insights on disruptive potential of AI in the Enterprise with @itsbobjohnston, CEO of Executive Council! lnkd.in/gUXHpuSw

Great insights from @digishopgirl on Instagram and Tiktok launch of search ads lnkd.in/g6mijR7T

Interesting approach from UC Berkeley and Microsoft Research: LLM focused on suggesting relevant APIs (eg actions) and prompts to call those, to make foundational LLMs like GPT4 work better with external tools and data sources #ge…lnkd.in/g-ivrEK9 lnkd.in/gG2X-ET2

Interesting approach of using (lightweight) LLMs to create reusable tools for more complex tasks. In addition to better automation ratio and increased productivity, this approach is expected to significantly reduce costs of LLM inf…lnkd.in/gFkdJKqg lnkd.in/g5WTTDyK

Interesting use cases of using Generative AI - improving creator (eg writing Sci-fi books) or white collar (eg writing customer facing email/chats etc) are the most practical applications of this disruptive technology (lnkd.…lnkd.in/gRd7MVz5 lnkd.in/grQcCWs3

Meta recently released LIMA LLM model is at least ~3X smaller than GPT-3.5/GPT-4 (65B vs 175B paramenters). Moreover, it is trained on very small set (1,000) of supervised samples. However, per Meta's @ylecun, LIMA's performance…lnkd.in/gjMaTCZF lnkd.in/guFP3ZyQ

Great insights from @digishopgirl Mets's AI-based Promotional Ads tool lnkd.in/gsHaefs6