Karimulla Shaikh

Glad to have shared my thoughts around software, hardware, and our insatiable thirst for more energy with Scott McGrew on NBC Bay Area. https://lnkd.in/gYR-q9gS As data centers, and the energy to run them, represent a larger portion of enterprise P&L we need to start thinking about how we're building software, and whether it is efficient, not simply to make an app faster, but to spend the least amount of energy necessary. We're in a new era of tech companies buying nuclear reactors, it's time to get serious about leveraging more efficient hardware to get the most done with the fewest watts. Voltron Data #SQL on #GPUs

50

2 Comments

Geoffrey Hinton, the "godfather of AI" and 2024 Nobel laureate in Physics, has sounded the alarm on the technology he helped create. His concerns echo those of the Keurig cup inventor a decade ago, who regretted the environmental impact of his innovation. Both cases highlight a crucial lesson: groundbreaking technologies can evolve beyond their creators' intentions. Hinton warns that AI could pose an existential threat to humanity, potentially manipulating people and even replacing biological intelligence24. As he puts it, "It's quite conceivable that humanity is just a passing phase in the evolution of intelligence." These cautionary tales remind us of the importance of responsible innovation. As technology advances at breakneck speed, we must ensure it benefits rather than harms our future. It's time to heed the warnings of pioneers like Hinton and work towards aligning AI with human values and interests. Source - https://lnkd.in/gxH-KDFu

3

Had a great conversation with Aravind Krishnaswamy (Arvi) on the Pivotal Clarity podcast about how AI is shaping startups and product management. 🌐 Arvi is the Director of PM for Cisco's Contact Center and Experience, responsible for key AI projects in the last few years. Few things we covered - AI isn't magic. It's a tool to solve real problems. Many founders get caught up in the hype and forget this. - Timing matters as much as the idea. A great product at the wrong time will fail. Understand market readiness. - Distribution is critical. The best tech won't succeed if no one uses it. Think about this from day one. - Data is key for AI. Collect it carefully and use it to make decisions. Guessing won't cut it anymore. - Indian founders have a unique opportunity. The AI shift levels the playing field. Take advantage of it. If you're working on AI or a new startup, listen to this talk. Arvi has interesting ideas about where things are going. 🎧 Listen to the full episode available now! Links in comments below 👇

17

1 Comment

AI leaders have a crucial chance to address risks now and strengthen governance practices before playing catch-up to regulations, notes Kjell Carlsson, Ph.D. of Domino Data Lab. In a recent VentureBeat article, he shares his perspective on California's AI bill veto and why proactive governance is key to the future of AI innovation. Learn more about this addressing this opportunity: https://gag.gl/4WCTkP #AI #AIGovernance #AIInnovation #DataScience

3

AI leaders have a crucial chance to address risks now and strengthen governance practices before playing catch-up to regulations, notes Kjell Carlsson, Ph.D. of Domino Data Lab. In a recent VentureBeat article, he shares his perspective on California's AI bill veto and why proactive governance is key to the future of AI innovation. Learn more about this addressing this opportunity: https://gag.gl/4WCTkP #AI #AIGovernance #AIInnovation #DataScience

1

Exceptional article by Walter Frick on the evolving views of economists about the impact of technology on workers, the workplace, and the economy. This is our world at KYield, Inc. -- my core patent for the KOS is optimizing knowledge yield in the digital workplace, first developed nearly three decades ago while operating the world's leading learning network for thought leaders. The name of the company is based on that theorem. Automation is necessary but woefully insufficient to achieve knowledge yield as knowledge is the domain of complex humans, even if LLM bots can mimic behavior. The more transfer of knowledge that occurs without permission, the less knowledge will be shared. One of the first lessons learned in knowledge transfer. Evidence has shown that when exploitation is excessive, false information will be transferred, as we've seen in the poisoning of LLM training data among many other case studies. Optimizing knowledge yield in the digital workplace requires a holistic systems approach based on deep understanding of several disciplines, hence the KOS (EAI OS) as the manifestation of the theorem. One of those disciplines is economics, where I've invested a good part of my career. I moderated the global economics community in our learning network (GWIN) in the late 90s, which included leading economists, and where I have conducted a fair amount of research myself. This work is mostly in the domain of behavioral economics, which is the subspecialty that best understands related phenomenon such as the influence of realized moral hazard in causing financial crises, and the wealth gap influence that can help trigger economic depressions. The land grab currently underway by LLM firms and Big Tech is actually quite similar to patterns in other industries that caused major catastrophes, such as the subprime mortgage disaster when a handful of "market makers" would have caused the collapse of the global economy if not for record levels of financial engineering by the Fed, which did severe damage to the U.S., and is still causing severe repercussions. One of many lessons in human-caused catastrophes is never allow specialists in one narrow discipline who have billions of USD at stake set the terms that impact everyone else. Surprising numbers of people have proven willing to throw society under the bus in an attempt to satisfy infinite greed or an ideological obsession. In the case of LLMs, not only do we have both greed and idealistic cults driving behavior, the implications include no less than risking the entire knowledge economy.

20

5 Comments

It was great to hear Robert Wong Julie Willem Angie McAllister, Ph.D. and Markus Kunesch on the recent episode of the @Ed-Technical podcast -- the first of a three-part series about how we're applying Google AI to our work in education and learning: https://lnkd.in/e7hQi6a2. This episode delves deep into LearnLM, how we're making Gemini tuned for pedagogy and how we're integrating this technology into our mainstream products to improve learning outcomes, and evolve how people learn on Google products. It explains inspiring new product ideas, and the steps we take across the company to understand and enhance the journey of learning in a responsible and innovative way. #LearnLM #AI #Education #Google

64

Have you ever heard of a $20 cup of coffee? The Ain't Normal Cafe in Oakland is serving up just that. This reminded me of Madhavan Ramanujam's talk on value-based pricing. He asks the audience if they like coffee, then how about at $20 to reinforce the idea. According to a recent article in the San Francisco Chronicle, there's one lingering question: Can a $20 cup of coffee ever be worth it? Well, that depends on who you ask :) Check out the article and let us know your thoughts! #Pricing #ValueBasedPricing #Coffee #AintNormalCafe

5

5 Comments

AI leaders have a crucial chance to address risks now and strengthen governance practices before playing catch-up to regulations, notes Kjell Carlsson, Ph.D. of Domino Data Lab. In a recent VentureBeat article, he shares his perspective on California's AI bill veto and why proactive governance is key to the future of AI innovation. Learn more about this addressing this opportunity: https://gag.gl/4WCTkP #AI #AIGovernance #AIInnovation #DataScience

1

Is AI making product managers obsolete? Brad Eiben reveals the truth about the future of our profession. In our conversation, Brad Eiben, the Executive Director for the MS Product Management program at Carnegie Mellon University - Tepper School of Business, sheds light on the impact of AI on product management practices. -> Efficiency Through Automation. AI is revolutionizing how product managers work by automating routine tasks, allowing them to focus on high-value strategic activities. This shift is enabling teams to be more productive and creative. -> Human-Centered Skills Remain Irreplaceable. While AI can handle many technical tasks, the human elements of product management—like understanding user needs and crafting a compelling vision—are irreplaceable. Product managers must continue to hone these skills to stay relevant and effective. -> Strategic Organizational Change. Successful AI integration requires more than just adopting new tools. It demands a cultural, organizational, and process shift within organizations, driven by clear leadership and a willingness to experiment. I could not agree more with Brad. Full interview: https://lnkd.in/gzRFzhHq #PMtransformation #PMproductivity #CPO

10

5 Comments

Great discussion on societal benefits of AI with Microsoft and PG&E CEO. Patti Poppe reinforcing grid modernization and decarbonization at a lower cost due to load growth from data centers. “the growth that AI and computing and data centers are driving, will in fact fund the rebuilding of our national grid, it will fund the decarbonization of our economy, without that new load growth, we would just be paying for it twice because it’s gotten old, - now we get to modernize it, - decarbonize it, and - we can do it at a lower cost, thanks to the growth driven by all of this computing”

14

1 Comment

Exciting times ahead as we witness technologies reshaping our world, akin to the transformative impact of electricity. Just as electricity paved the way for innovations and applications like ovens, washers, and refrigerators—things we interact with daily—generative AI is now the foundation for building the future. AI, like electricity, operates behind the scenes, powering the tools and applications we are starting to rely on, creating a new layer where folks interact with it in countless ways, from content creation to customer support and this is just the start. Recently, leveraging generative AI, we achieved remarkable efficiency, reducing a 200-hour task to 90% completion in just a couple of hours. Reflecting on historical inventions with my son, we discussed how the steam engine revolutionized transportation and the Bessemer process enabled mass steel production for skyscrapers. These innovations, like elevators, opened doors to new possibilities, much like how generative AI is transforming industries today. Amid economic fluctuations, the potential of technologies like generative AI is immense, reminding us of our place in a continuum of human ingenuity. I’m excited for the possibilities ahead for us and future generations.

23

3 Comments

In new work at UCSC announced recently, researchers demonstrate how to dramatically improve the energy efficiency of LLMs while maintaining performance by eliminating resource-intensive matrix multiplication operations and leveraging custom accelerator hardware. At the core of the work is a scalable LLM approach that uses additive operations in dense layers and element-wise products for self-attention functions. For their GPU implementation, which employs CUDA kernel optimizations on standard Nvidia GPUs, they saw significant acceleration and reduced memory consumption for training, and for inferencing they saw far greater performance improvements and memory usage reductions, especially when the model is scaled to 13B parameters. The dramatic energy savings is seen on a custom FPGA accelerator which exploits lightweight operations not available on today's GPUs. The work suggests the potential for impressive advancement of lightweight LLMs, as well as a new direction for future LLM hardware accelerators. UCSC Newcenter article: https://lnkd.in/gqzqqF_w

16

1 Comment

Communities are vital for startups to achieve PMF, CMF, PLG, and support OSS. In AI startups today, communities drive innovation and product adoption by addressing diverse user needs and complex product demands. Read this post for more details. Digital Data Design (D^3) Institute at Harvard Harvard Business School Harvard Innovation Labs #MTLC #BostonTech #TechHotTopicSeries

11

1 Comment

AI leaders have a crucial chance to address risks now and strengthen governance practices before playing catch-up to regulations, notes Kjell Carlsson, Ph.D. of Domino Data Lab. In a recent VentureBeat article, he shares his perspective on California's AI bill veto and why proactive governance is key to the future of AI innovation. Learn more about this addressing this opportunity: https://gag.gl/4WCTkP #AI #AIGovernance #AIInnovation #DataScience

1

Honeywell’s CEO, Vimal Kapur, recently raised a perspective on AI that’s resonant with the direction we’re taking at Fatigue Science. It’s something I’ve been saying for a while now. He highlighted that the real value of AI isn't just in productivity boosts.  Kapur’s take emphasizes that AI copilots and predictive technologies aren't about replacing workers but empowering them to work more effectively and safely—a vision we share for high-hazard industries like mining and transportation.  Predictive fatigue management, for instance, can boost productivity, absolutely. But, more importantly, it can improve worker health and safety. For industries like ours, AI leaders move beyond productivity gains and anticipate other challenges—from labor shortages to equipment maintenance to fatigue risk. AI tools allow us to both mitigate risks and seize profitability opportunities with agility. As Kapur suggests, the future belongs to companies that understand and leverage their unique datasets to build AI models tailored to their industry’s needs. At Fatigue Science, we’re doing just that by creating solutions that turn our client’s sleep and fatigue data into actionable insights, ensuring heavy industry leads the way in safety. AI isn’t just a tool for optimizing today’s workflows; it’s an essential driver for tomorrow’s growth. Leaders who embrace this shift can navigate complexity with confidence and lead their companies (and industries as a whole) forward with foresight and precision. #AI #PredictiveAnalytics #FatigueManagement #Leadership Here’s the full post: https://lnkd.in/gnKem3sw

21

AI leaders have a crucial chance to address risks now and strengthen governance practices before playing catch-up to regulations, notes Kjell Carlsson, Ph.D. of Domino Data Lab. In a recent VentureBeat article, he shares his perspective on California's AI bill veto and why proactive governance is key to the future of AI innovation. Learn more about this addressing this opportunity: https://gag.gl/4WCTkP #AI #AIGovernance #AIInnovation #DataScience

5

This is another huge step forward for Appvance IQ. Since we introduced our AI Coverage Map in 2020, it has become a key feature in how AI Automation Engineers and their Dev counterparts strategically deploy unlimited AI testing resources targeting the most business-critical and vital user flows. The new coverage maps in 5.2 bring even more insight into how enterprise applications are really working and where defects are lurking, where performance is lagging, and where risks are undetected. Let me know if you'd like to learn more.

8