Sharon Zhou, PhD

Thanks to #genai360 for the shoutout here for Consilience AI and our AlphaIQ product. Davit Buniatyan Joseph Byrum, Ph.D., MBA

10

Adrienne Murphy discusses an outstanding example of data leadership by the Massachusetts Department of Early Education and Care. Transcript: Adrienne Murphy (Massachusetts Department of Early Education and Care): First, in terms of understanding the pre-K landscape, the CUSP network has really helped us understand, in this three and four-year-old space, who's being served across the mixed delivery system, which has been especially important when our governor, the Healey Administration, announced a new Gateway to Pre-K Initiative, which included a proposal to deliver universal, high-quality preschool access for four-year-olds in all of our Gateway cities, which are our sort of mid-sized urban centers that are sort of the regional economic anchors across the state. And the cost model has been able to help us to benchmark sort of where are we now, who's being served across our licensed and funded early education care providers, across our public K-12 system, across our private school system, to be able to understand sort of the full landscape of who we're currently serving. And, by virtue of who we're currently serving, also, who are we not serving? Sort of what's the delta between where we are now in pre-K access for four-year-olds across the state, but really particularly in these gateway cities? And then sort of, what's the delta between where we are now and the universal access that is proposed in the gateway to Pre-K initiative? So, this is, you know, really one instance in which being able to know who are we currently serving, who are we not serving, and where are we not serving children has become really key for us and for Massachusetts as sort of the policy design process. Click here to learn more: https://team3si.com/ #communityofpractice #dataforgood #earlychildhooddata #acf #ecids

1

This new little product demo is awesome! Our new Rev AI Transcript Assistant will save you TONS of time. Easily pull out nuggets of info from your transcripts, making valuable content more accessible than ever. https://lnkd.in/gzzFpcDa #transcription #asr #accessibility #transcript #summarization #ai #assistant

31

2 Comments

Heading to Southern California! Join me and Garrett Springer at a free public meetup on July 25th where I will talk all about LLMs. (Don't tell corporate, but it's not really about Snowflake) 😃 What I want to talk about: 3 Steps for training a LLM - these have consequences around hallucination, agentic workflows, and expectations of LLMs Improving LLM Performance - I get a lot of questions about training your own model, fine-tuning, prompting, long context, function calling, so lets address them Using LLMs - Importance of evaluation and the latest lessons, Risks around LLMs (security, legal, ethical), and having a strategy to using LLMs If you are in the area, swing by and say hi. I will also make a recorded version of this talk at some point before the end of summer. Meetup Info: https://lnkd.in/g9qmgvtx Check out my talk from March: https://lnkd.in/g6nhnKF7

30

2 Comments

Lin Guan and Yifan Zhou have led a very interesting study investigating the effectiveness of Vision Language Models in critiquing robot behaviors in terms of the desirability of their "style." If you are wondering how our work on LLM's inability to verify correctness squares with this study on using them as style critics, see my earlier post https://lnkd.in/gBYAr5qq ! This is the reason why being a style critic is one of the roles LLMs can play in LLM-Modulo framework: https://lnkd.in/gyAyKx4E Check out the details of the work @ https://lnkd.in/gfVZ4EYW ( Lin Guan also has a nice twitter thread on the work here: https://lnkd.in/gnieU-qM )

89

2 Comments

Powerful look at NVIDIA NIM in action! Therapyside is using NIM to enhance their therapy platform with virtual assistants that can take notes in sessions, letting therapists focus on connecting with their clients. Read more: https://lnkd.in/efeKKABC #AIagents #AIinaction #generativeAI

39

4 Comments

Domain expertise has tremendous value in #explainable machine learning. Causality is the bridge to aligning this domain knowledge and understanding why models are guided to a particular response. In episode 12 of the #aiportfolio podcast Serg Masís and I discuss explainable ai and #llms. If you have watched Jon Krohn's show, Super Data Science Podcast, lots of the great technical questions come from Serg Masís, so his insights into #machinelearning in highly regulated environments can expand your knowledge on the topic. One thing that stuck out to me is that if you have an exploration mindset, you can leverage models to help you do better Exploratory Data Analysis vs just visualizing statistics on the data. Links to show are in the comments

85

7 Comments

There are a lot of people in the Edu space struggling to get rolling with AI. There are lots of myths and confusion going around in #AI (aka: deer-in-headlights syndrome!) To do our part, we have instituted an AI research grant program (valued at $3,000 per award) to help researchers, professors and others in the Edu space to get rolling with our award-winning no-code / low-code AI platform. As I've often said: If you are not able to get rolling with your first AI pilot in just a few hours, we have FAILED! More details about the program here: https://lnkd.in/eSZG6rTG PS: If you know people in the Edu space, please tag them in the comments -- OR -- share this post. #ai #grants #research #education

19

2 Comments

Acquired nails it again: learning about Microsoft's early days is like sitting at the feet of the software startup gods. Two things really stood out beyond Acquired's playbook below: >> Team talent. An unnatural amount of talent that already was within their own social circle was a real catalyst. Also, I gained a real appreciation for Steve Ballmer and his contribution in building a real enterprise software company. >> Waiting for the right business model. The patience of Bill Gates in building the right, with an all you can eat DOC license for IBM to build a market is admirable. That gave Microsoft the opportunity to sell on individual licenses to the clone manufacturers, where they made their money. That was really playing the long game. Thanks again to David Rosenthal and Ben Gilbert for an outstanding episode. #innovation #startups #microsoft #software #startup

18

Loved bringing together a group of research-centric, product-minded technologists building the future of AI tonight in the 🅱️🍎, incl. product / eng leadership @ NVIDIA, Pinecone, Datadog; founding eng @ Abnormal Security, Modal, Counsel Health; applied AI @ OpenAI, Anthropic; research eng @ Meta (FAIR), Harvey, Google DeepMind; founders @ Sapien and some of NYC's greatest stealth AI co's ☄ We covered a lot of ground over 4 hours - some of the recurring themes in our conversations: - Agents:🕴Model inference is getting so much faster, which will unlock AI agents' potential. That said, agents should be trained to handle very narrow tasks, reliably, first vs. an open-ended approach many startups are attempting today - Robotics: 🤖 While the science is not there yet, the robotics market and its impact will be much bigger than LLM. The introduction of humanoids will fundamentally change how we live. That said, the U.S. is in a vulnerable position due to lack of supply chain independence, compared to rivals like China - Value creation: ✨ AI personalization is overrated as consumers mostly care about status (i.e., ‘whatever is the best’) vs. what's tailored to them. Delivering great UX is still hugely undervalued, for both enterprise and consumer use cases. Cost and reliability remain enterprise buyers' front-and-center priorities. AI can preserve endangered rare languages by documenting and creating digital archives / databases cc: our amazing sponsor Arc (Brendan Bettencourt, Erika Bricky, Eric Lo), Visible Ventures

136

5 Comments

Narratize is delighted to announce that we have been accepted into NVIDIA's Inception program, an accelerator designed to help startups evolve faster through cutting-edge technology, opportunities to connect with venture capitalists, and access to the latest technical resources from NVIDIA. Since the start of 2024, the team has: ✅ Recruited a seasoned and diverse engineering, data science, and narrative science product team ✅ Focused on building DEI best practices into our data architecture, ensuring our uses are always safe, equitable, secure, and human-centric ✅ Launched the world’s first generative AI white paper builder ✅ Welcomed new enterprise customers across science, tech, and medical industries ✅ Traveled the world sharing insight and inspiration on GenAI transformation This spring, Narratize is bringing 20+ new use cases to Narratize’s Core Product Library as well as new features designed to streamline communications and collaboration among product, R&D, and marketing teams.   ‍ Narratize combines the power of narrative science with cutting-edge generative AI technology to ensure that no great idea is left untold. From optimizing research & development processes to scaling product launch, Narratize empowers brands with transformative ideas to unleash their full potential. Narratize brings together computer science, data science, and narrative science expertise to develop AI in a uniquely human, story-led way. You’re invited to experience the magic yourself, Try Narratize’s Core Product Library free for 14 days 👉 https://lnkd.in/e8HEfwvR #narratize #genai #innovation #nvidiainception #techstartup

5

Back at it with another awesome #linkedinlive with Evan Hall! In this session, they covered topics like: ➡ What data is important in customer education ➡ Telling a story with your data ➡ Trends in how organizations are structuring customer education roles ➡ The relevance of AI in customer education We always learn a ton from thought leaders in this space and looking forward to exploring the world of customer education even more in the future.

7

Small models becoming smarter! Phi-3 reinforces the importance of data quality.

24

📜 A new blog post explaining the fundamentals of RL for LLMs—the perfect place to intuitively understand PPO and RLHF from first principles. 👇👇 Pre-trained LLMs are often compared to Lovecraft's Shoggoth, a monstrous, shape-shifting beast. PPO is the algorithm that brought it under control. Take a look! 🔗 https://lnkd.in/eXqDvrUV

29

1 Comment

"How does a vision LLM outperform OCR, and why?" I got this question from our customers a lot and we distill the answers into the blog: https://lnkd.in/guv8waCs Overall, Vision language models (VLM) address key OCR shortcomings by: 1. Accurately interpreting low-quality images and complex layouts 2. Understanding context and semantic meaning in documents 3. Handling multiple languages and scripts seamlessly 4. Reducing post-processing overhead #pdf #word #ocr #llm

36

3 Comments