Reza Rooholamini

Organizations have insufficient technical capabilities for the speed and scale of their ML model implementations. My colleague Sumit Agarwal just published a very helpful research with a capability view and sample vendors for defining a modular, flexible and scalable MLOps solution. https://lnkd.in/gnB_T9bq #MLOps #LLMOps #AI

149

17 Comments

I've contributed to a groundbreaking white paper titled "The AI Revolution No One is Talking About: Why Most of Your Employees Will Be AI Agents." This important work, published by SDL Advisors, led by Chad Andrews, is now available at https://www.sdla.ai/ Key highlights of the paper: - Provides evidence for the imminent and severe disruption of labor markets by AI - Offers a step-by-step guide for preparing enterprises to benefit from an AI-powered workforce - Explores the concept of Agentic AI as a new class of employee This is a must-read for anyone interested in the future of work and enterprise technology. It provides actionable insights on how to boost your workforce's productivity with AI agents, which I believe will be the next evolution of Generative AI capabilities. If you're even remotely interested in what Generative AI is capable of through AI agents, please take a moment to download it. Chad and his team are also available to consult with your enterprise, and walk you and your team through their research, and how it may apply to your business. #AIRevolution #FutureOfWork #AIAdvisory #WorkforceTransformation #EnterpriseTech #AIProductivity #AgenticAI

23

2 Comments

Beyond single-score measurement and rankings this article provides more comprehensive approach to LFM (Large Fundamental Model) evaluation from end user perspective. https://lnkd.in/gSZbrqEb

20

“Despite militaries’ desire to use LLMs and other AI-enabled decision-making tools, there are real limitations and dangers. Above all, those militaries that rely on these technologies to make decisions need a better understanding of how the LLM works and the importance of differences in LLM design and execution. This requires significant user training and an ability to evaluate the underlying logics and data that make an LLM work. The result should be that a military user is just as familiar with an LLM as the user is with the radar, tank, or missile that it enables. This level of training and expertise will be easier to accomplish in peacetime and with advanced militaries, meaning it is the wartime use by militaries already strapped for labor, technology, and weapons where these systems may create the most risk. Militaries must realize that, fundamentally, an LLM’s behavior can never be completely guaranteed, especially when making rare and difficult choices about escalation and war”

9

1 Comment

Generative AI is a popular topic today. AGMA is providing a free webinar tomorrow at 1:00 pm Eastern to help you optimize your use of this new tool. Join us! Intelligent Manufacturing Augmentation: The Copilot Model for Generative AI with Craig B. Hodges and Frank Braski Unlock the future of manufacturing with cutting-edge insights into Generative AI, Large (and small) Language Models, and RAG!

3

I enjoyed this article about how generative AI can be incorporated into Computer Science progrems in colleges and universities. I think almost everything in here is true for adoption within enterprises as well. I especially like the quote from Johnny Chang of Stanford University -- "We should be making AI a copilot -- not the autopilot -- for learning" Thanks to my colleague Edward Kiledjian for sharing this with me.

65

2 Comments

Whew! Everything is moving so quickly, I can barely keep up lol Planning to tune in to this Applitools webinar to see how AI is being used to help with testing https://lnkd.in/eiHV3ztg

39

Current Time Series Anomaly Detection Benchmarks are Flawed and are Creating the Illusion of Progress: https://lnkd.in/gFKSqrH6

1

AI model providers are updating their usage policies. Anthropic has changed their policy to classify insurance - including underwriting, claims processing, and coverage decisions as "High Risk", with the policy mandating human review of all AI output. Like many such topics, there's some nuance. Human oversight is a good thing. But insurance is a heavily regulated industry that already has policies covering how AI is used. Restrictions from AI providers are another challenge to navigate for insurers, and evolving use policies create uncertainty for insuretechs who are building off of AI platforms. Philip Dawson and I discuss this further in the linked article. While it's not always practical for businesses, updates like this are one of the reasons I'm a big fan of running your own AI models and not depending on big providers. https://lnkd.in/e5fVs6df

10

Facinating. Identifying concept representations in LLM. By Anthropic. This Hard Fork episode has a segment on it: https://lnkd.in/eT6wEghz

8

After attending a Kellogg Executive Scholar class at Northwestern University, I've come across fascinating AI articles and use cases from various research areas in the school. One such article discusses the transformative power of AI in healthcare, aiming to "truly save lives." The article is a use case from a favorite AI as a Service of mine: Weights and Biases. #AI #Healthcare #Kellogg #NorthwesternUniversity #WeightsAndBiases https://lnkd.in/gSyVFZ7j

10

It seems every day presents a new innovation in the field of AI with constant speculation on when AI will surpass human intelligence. It seems to me we should get a better grasp on human brain physiology and consciousness before we profess to be on the cusp of AGI (Artificial General Intelligence). A recent project at the Lichtman laboratory at Harvard University and the Connectomics at Google team have brought us one step closer to that understanding with the release of the “H01” dataset and companion paper. "The dataset comprises roughly one cubic millimeter of imaging data, including tens of thousands of reconstructed neurons, millions of neuron fragments, 183 million annotated synapses, 100 proofread cells, and many additional subcellular annotations and structures — all easily accessible with the Neuroglancer browser interface." Source: MIT Technology Review #AI #AGI #Neuroscience #CognitiveScience

12

"Discover the groundbreaking robotics advancements unveiled in the last 24 hours, highlighting the impressive leaps in AI integration, cooperative robotics, and innovative public demonstrations:" - Innovative Robotics Competition: The Curiosity Open: Robotics Challenge, hosted by the Great Lakes Science Center, features high school robotics teams competing in a music-themed game called CRESCENDO, where robots must score notes, amplify speakers, and harmonize on stage, showcasing advanced engineering and teamwork. - Robot Block Party Demonstrations: At the 7th Annual Robot Block Party in Boston, the public can interact with cutting-edge robots, including drones, collaborative robots, humanoids, robot dogs, and flying bionic birds, with hands-on activities such as flying through a drone obstacle course and controlling an underwater robot. - Advanced Robot Interactions: The Robot Block Party includes a robot parade and opportunities for the public to drive a robot dog, control a collaborative robot arm, and engage with various robotic technologies, highlighting the latest innovations in robotics. - Cooperative Robotics: In the CRESCENDO game, robots can earn cooperation points by working together with opposing teams, demonstrating advanced cooperative robotics where teams repurpose notes to reduce scoring thresholds and earn bonuses. - Robotics and AI Integration: The events at RoboBoston and the Curiosity Open highlight the integration of AI and robotics, with robots performing complex tasks such as autonomous navigation, note scoring, and amplifying speakers, showcasing the growing capabilities of AI in robotics. #RoboticsInnovation #AIIntegration #TechAdvancements #STEMEducation #RoboticsCompetition #CooperativeRobotics #RobotBlockParty #EngineeringExcellence #FutureOfAI #InteractiveRobotics

1