Venky Rao

Lots of data being captured by the current set of tools What we need is AI, Unification of that data and conversational insights. Jeff Wellstead talks about the intelligence layer AI can bring in. #AI

43

2 Comments

Excellent book on Enterprise AI with recipes and reference cases published by #GridDynamics and written by #IlyaKatsov Advancements in deep learning, reinforcement learning, and generative AI have dramatically extended the toolkit of machine learning methods available to enterprise practitioners. This book provides a comprehensive guide to how marketing, supply chain, and production operations can be improved using these new methods, as well as their use in conjunction with traditional analytics and optimization approaches. ISBN 979-8218291624

3

2 Comments

IT leaders look beyond LLMs for gen AI needs Not all enterprise use cases are best suited for large language models. New multimodal generative AI models and smaller models show promise for niche needs. #llms #generativeai #aidmodels

1

1 Comment

I've worked in and around systems management for over 25 years, Back in the good old days, we talked about something we called, "automatic remediation". This was the idea that, through some kind of automation, we could automatically clear errors on networked equipment, servers, routers, switches, etc. However, this never became the norm. Mainly because it was too complex. The idea was that we would have some kind of logic tree that would take some action based on an error and then fix or "remediate" the error, reducing the mean time to recovery. Anyhow, with the new advances in AI, I think this is finally going to become a feature that's likely going to be put into practice. One of my partners, Pulumi, yesterday released a beta of their Pulumi Copilot, which expects to provide a level of automation that I think could finally make this wish a reality. I like Pulumi because their base technology is open source, and they have stayed the course even when companies like HashiCorp changed their license, resulting in confusing forks and ultimately a huge disruption to their user base. While Pulumi's AI component is part of a service (currently in beta), it's a good model where the software can remain open source but rely on higher value solutions that have real costs, like usage from hosted AI models. I also saw that Amazon's CloudTrail was doing something similar with a Natural Language Processing (NLP) interface. I think it's the beginning of a trend. Bringing AI into the data center as an administrative aid. Because the infrastructure limits the context of the model and they can make use of a ton of incoming data to provide training to AI agents or copilots, I believe these are excellent use cases for AI. Highly skilled systems and networking professionals will have a copilot or junior administrative AI agents at their beck and call to reduce the mean time to recovery and improve the reliability of infrastructure. Anyhow, I was jazzed to see my old world of systems integration start to adopt my new passion for AI. I'd be curious to hear from any of my systems management friends if they are leveraging AI to manage systems and that's working out. Let me know if you have any use cases in the comments below.

572

4 Comments

Google Cloud Vertex AI is rolling out multiple updates to make AI accessible for every enterprise. Up to 85% lower input costs and 80% lower output costs, and 100+ new languages coming to Gemini 1.5. #GoogleCloud #VertexAI #Gemini

147

7 Comments

What happens when a market disruptor teams up with an establishment tech company? Thanks to Madhumita Murgia of Financial Times for sharing my insights from 5 years at Microsoft (first as General Manager of AI Products and Search, then later as Chief AI Technology Officer in the business application group). As we all saw with the marriage of OpenAI and Microsoft, ways of working can clash and lead to crisis. The truth is that behind the scenes, aligning communications approaches and tolerance for risk are huge challenges when a start-up joins a long-time industry leader. The public typically isn’t aware of how disparate teams with unique operating cultures - even at the same company! - don’t always see eye-to-eye on product development, rollout, and sustainment. The road to innovation is never without bumps along the way, but I think both companies (and their leaders) can put the drama behind them to continue pushing AI into its next phase. Learn more about how Microsoft adjusted their strategy after 2023’s drama of OpenAI - and the excellent perspectives from fellow industry insiders - here: https://t.ly/WGfqP Stephen Morris, Navrina Singh, @Jan Leike, Eric Boyd, Reid Hoffman, Ayanna Howard, Fei-Fei Li #AI #FutureTech #GenerativeAI #GenAI #DigitalTransformation #TechTrends

23

6 Comments

Give your AI models an edge. Leaders across industries are racing to get generative AI off the ground—and get ahead of the competition. How can your enterprise unlock the full potential of gen AI? Achieve more accurate and nuanced results by moving beyond generic training sets and grounding your AI models in real-time and enterprise data. At Google Cloud, we call this approach “enterprise truth”. Grounding foundation models in enterprise truth—like your databases, enterprise applications, and Google Search—enables the real-world integration, data authenticity, and data relevance today’s business demands.

2

Amit Bleiweiss explores how re-ranking, powered by machine learning, is revolutionizing enterprise search by aligning results with user intent for enhanced accuracy and relevance. This leads to a significantly more effective semantic search experience. #AI #MachineLearning #ReRanking #SemanticSearch #EnterpriseSearch #this_post_was_generated_with_ai_assistance #responsibleai https://lnkd.in/ei9_kjGn

1

Exciting Times Ahead: The Convergence of Generative AI and Live Streaming. The worlds of generative AI and live streaming are colliding, and the results are nothing short of game-changing! I can't help but be thrilled by the potential this convergence holds for enhancing the user experience in live events. Imagine a live stream that not only broadcasts the event but also adapts and personalizes the content for each viewer in real-time. That's the power of generative AI! By leveraging machine learning algorithms, live streaming platforms can now analyze viewer engagement and dynamically modify the stream to keep users hooked. But that's not all! Generative AI is also revolutionizing the way we interact with live streams. Picture virtual hosts and interactive elements that respond directly to individual users, creating a truly immersive and engaging experience. It's like having your own personal guide through the event! Moreover, the integration of AI-driven analytics is providing unprecedented insights into viewer behavior and preferences. This data goldmine allows businesses to fine-tune their live streaming strategies, deliver targeted content, and maximize monetization opportunities. The applications are endless, from sports events and concerts to educational webinars and product launches. Generative AI is transforming the live streaming landscape, making it more dynamic, personalized, and profitable than ever before. As industry leaders like Google Cloud pioneer groundbreaking AI solutions for live streaming, it's clear that this is just the beginning. The future of live events is here, and it's powered by the incredible synergy of generative AI and live streaming! Are you ready to embrace this exciting new era? Let's dive in and explore the boundless possibilities together! #GenerativeAI #LiveStreaming #UserExperience #LiveEvents #FutureOfStreaming

4

Usage-based revenue models are on the rise. BoostUp's co-founder, Amit Sasturkar breaks down everything you need to know about how to set up an effective forecasting cadence.

10

Stability AI, the company behind the popular image generation model Stable Diffusion, is undergoing a leadership change: New Captain at the Wheel: Stability AI appoints Prem Akkaraju, former CEO of Weta Digital, as their new CEO to navigate financial challenges. Financial Turbulence: The company reportedly faced low revenue and high cloud computing costs, leading to employee layoffs and a need for restructuring. Reboot for Stability: The hope is that Akkaraju's experience and a strategic financial backing can stabilize the company and ensure its future success. AI Startup Struggles: Stability AI's situation highlights the challenges faced by AI startups in commercializing their innovative products. The Future of Stable Diffusion: With new leadership and resources, Stability AI has the potential to overcome its hurdles and further develop Stable Diffusion and other AI tools. This leadership change signifies a critical juncture for Stability AI. Hopefully, it paves the way for a more stable and successful future! #AWS #datascience #AI #StabilityAI #Entrepreneurship https://lnkd.in/gzgU7xEt

2

New this week: our latest Google Cloud[TAG] Marketplace video series – all about the transformative power of #genAI for enterprises – launches with a video from our partner Weights & Biases. See how simple it is to fine-tune a large language model using Weights & Biases and Vertex AI; Watch it here on demand → [ https://lnkd.in/gP7mU_vm]

1

1 Comment

Ai : Hunanity's greatest creation for Good and Bad. To secure ourself #zerotrustai In the journey, MS brings up VASA-1 will AI model which essentially spins a selfie image and turn it into a talking clip of you. All you have to do is upload a photo along with a voice note and let the AI model do the talking for you.

2

𝗜𝘀 𝗜𝗕𝗠 𝗔𝗜 𝗺𝗼𝗿𝗲 𝗼𝗽𝗲𝗻 𝘁𝗵𝗮𝗻 𝗢𝗽𝗲𝗻𝗔𝗜? Today IBM has kicked off its annual 𝗧𝗛𝗜𝗡𝗞 conference in Boston with a keynote by CEO Arvind Krishna. https://lnkd.in/dfHEMyWM The event, preceded by a dedicated day for Ecosystem Partners, focuses on AI. Among the announcements, and expanding partnerships with top AI players in the industry, IBM released a family of open-source models called 𝗜𝗕𝗠 𝗚𝗿𝗮𝗻𝗶𝘁𝗲, including a Code model that supports over 100 programming languages. These models are licensed under 𝗔𝗽𝗮𝗰𝗵𝗲 𝟮.𝟬, allowing unrestricted use and development of derivative models. In comparison, Meta’s 𝗟𝗹𝗮𝗺𝗮 𝟯 is labeled as an “o͟p͟e͟n͟ ͟m͟o͟d͟e͟l͟,” but it falls short of being a true “o͟p͟e͟n͟-͟s͟o͟u͟r͟c͟e͟ ͟m͟o͟d͟e͟l͟.” Their Community License Agreement imposes limitations — for instance, it prohibits using Llama 3 to improve other competitive models. On the other hand, 𝗢𝗽𝗲𝗻𝗔𝗜 maintains a closed approach with models like GPT-4 or GPT-4o, that was announced last week. Despite the “Open” in their name, OpenAI prioritizes control and security for responsible AI development, as emphasized by CEO Sam Altman. Interestingly, Altman was a guest speaker during 𝗠𝗶𝗰𝗿𝗼𝘀𝗼𝗳𝘁 𝗕𝘂𝗶𝗹𝗱 which coincidentially also takes place this week, and both IBM and Microsoft CEOs delivered opening keynotes today.
 IBM demonstrates the ability to create trusted models while adhering to open-source principles. The recently announced 𝗜𝗻𝘀𝘁𝗿𝘂𝗰𝘁𝗟𝗮𝗯, sponsored by Red Hat and IBM, aims to effectively manage truly open-source LLMs, mitigating the challenge of uncontrolled forks. 𝗜𝗕𝗠 started in 1911 as proprietary, today their AI models are 𝗼𝗽𝗲𝗻. 𝗢𝗽𝗲𝗻𝗔𝗜 started in 2015 as open, today their models are 𝗽𝗿𝗼𝗽𝗿𝗶𝗲𝘁𝗮𝗿𝘆. Interesting, isn't it? #ai #Think2024 #ibm #opensource #llm #watsonx #ibmgranite #genai #openai #llama3

43

1 Comment

Download the IBM white paper. Generative AI (gen AI) and large language models (LLMs) are revolutionizing the data engineering capabilities required to solve complex data problems This white paper provides a perspective on the opportunities for improving data services using generative AI to augment data practitioners. https://lnkd.in/e9qy4R3W #ibm #GenAi #watsonx

7

Gen AI is moving from POCs to Projects in 2024...

8