What a great moment to be part of a future defining industry! The week at #Q2B in Silicon Valley was incredible, meeting so many brilliant and ambitious people, this quantum thing could just work. The people at Google tend to think so, https://lnkd.in/dUEN3DjE Big thanks to Quantum Delta NL Infinity QD QDNL Participations University of California, Berkeley
Big thanks to you for helping define the future also 🔥
#Fraud lessons from 45 years of audit and studies of past frauds – 216 Continued from 215 Learning from #shortsellers-13 Super Micro-4 Shell companies/dubious/unusual deals: (always beware of (bomb) shells which can be deadly) · In February 2024, Super Micro made an undisclosed investment in Lambda Labs, a technology startup and a key customer, and signed an unusual $600 million lease for a data center in California, sub-leasing part of this space to Lambda. · There are also undisclosed financial ties and transactions with other companies, including Leadtek and a Turkish shell company, raising further concerns about the company’s transparency and business practices. · Products of Leadtek were almost identical to Super Micro products. Dubious products/customer concerns: Genesys cloud was touted as a ‘success story’ on Company website But employees called it catastrophic. The key partner Nvidia recently praised competitor Dell and endorsed it. To be continued All views are personal
International SEO Consultant & Founder at Orainti. Maker of the SEOFOMO + MarketingFOMO Newsletters and the Learningseo.io Roadmap. Sharing insights and resources about SEO & Growth Marketing.
Google’s CEO Sundar Pichai remarks from the Q1 earnings call: * “We’re starting to bring AI overviews to the main search results page… focusing on areas where gen AI can improve the Search experience, while also prioritizing traffic to websites and merchants.” * “Based on our testing, we’re encouraged that we’re seeing an increase in Search usage among people who use the new AI overviews, as well as increased user satisfaction with the results” * “Since introducing SGE about a year ago, machine costs associated with SGE responses have decreased 80 percent from when first introduced in Labs, driven by hardware, engineering and technical breakthroughs.” More in the link in comments 👇
First Principles
CEO @ Groq, the Most Popular API for Fast Inference | Creator of the TPU and LPU, Two of the World’s Most Important AI Chips | On a Mission to Provide more than Half the World's AI Inference Compute by 2027
I often get asked about how the TPU chip came to be at Google? There were actually three attempts to build an AI accelerator chip at Google that went on about the same time. One of them was actually led by a Turing Award winner (I think everyone expected that to be the one that Google would use), there was the TPU (the project my team worked on), and one other team. The VP that funded the project had funded it out of his “slush fund,” which was the leftover money from canceled projects that he had discretion over. One of the most important things about how the project came to be was this… BECAUSE IT WAS JUST LEFT OVER MONEY, THERE WAS NO STARTING POINT OF “OH, THIS HAS TO WORK.” There was this freedom where we could say, “Okay, we'll just build whatever we think makes sense” and we went at it completely de novo with a clean slate. In my case, I didn't even know chip design. That turned out to be an edge because we were not confined by the curse of criticality, where the someone says, “Thou shalt be successful, and you must follow this playbook to limit risk.” My favorite part is the ending - After the TPU architecture was selected (and ultimately went on to run half the compute at Google), I remember the Turing Award winner giving a talk where he said, “I don't know who came up with this systolic array in the TPU, but they must be really old because no one uses systolic arrays.” I remember thinking, “Oh, that’s called a systolic array? It just seemed to make sense.”
This story deeply resonates with what i've observed working in technology innovation. I truly think we need different playbook for different kinds of innovation. For incremental innovation using adhered standards, benchmarking etc as a method for identifying angles of improvement make sense. but for First of its' kind type of innovations, its perhaps worthwhile to have a braver approach with little to no KPI, apart from successful PoC. it leads perhaps to a better exploration of the hypothesis space and consequentially to much better long term outcomes.
CEO @ Groq, the Most Popular API for Fast Inference | Creator of the TPU and LPU, Two of the World’s Most Important AI Chips | On a Mission to Provide more than Half the World's AI Inference Compute by 2027
I often get asked about how the TPU chip came to be at Google? There were actually three attempts to build an AI accelerator chip at Google that went on about the same time. One of them was actually led by a Turing Award winner (I think everyone expected that to be the one that Google would use), there was the TPU (the project my team worked on), and one other team. The VP that funded the project had funded it out of his “slush fund,” which was the leftover money from canceled projects that he had discretion over. One of the most important things about how the project came to be was this… BECAUSE IT WAS JUST LEFT OVER MONEY, THERE WAS NO STARTING POINT OF “OH, THIS HAS TO WORK.” There was this freedom where we could say, “Okay, we'll just build whatever we think makes sense” and we went at it completely de novo with a clean slate. In my case, I didn't even know chip design. That turned out to be an edge because we were not confined by the curse of criticality, where the someone says, “Thou shalt be successful, and you must follow this playbook to limit risk.” My favorite part is the ending - After the TPU architecture was selected (and ultimately went on to run half the compute at Google), I remember the Turing Award winner giving a talk where he said, “I don't know who came up with this systolic array in the TPU, but they must be really old because no one uses systolic arrays.” I remember thinking, “Oh, that’s called a systolic array? It just seemed to make sense.”
This was a fantastic event earlier today from C2C at the Google Atlanta offices! The theme was “AI optimized infrastructure - foundations to accelerate future innovation” I met some great people and learned a lot from the insightful discussions around AI. If you want to hear more about events like this please consider joining my newsletter linked in the comments below. 👇
❄️ Are we heading towards a #WebRTC winter? 🎯 The large vendors are going proprietary 🎯 Google, Microsoft, Amazon, Cisco, Meta, likely others They are putting their own innovations inside libWebRTC our around it But not contributing most of it back to the open source miracle that is libWebRTC I don’t blame them. It is a purely commercial decision (or a competitive one?) We are in for a huge cycle of innovation in WebRTC 👉 AV1 is coming 👉 Opus 1.5 with ML was just released (not inside libWebRTC yet) 👉 ML based codecs are around us (Lyra, Satin, …) 👉 We’re trying to figure where ML and #GenAI fits in WebRTC Lots of places to research and work on, but how much of it will find its way into the open source libWebRTC implementation, and how much of it will be kept proprietary? I am not optimistic on this front
Ahead of the budget tomorrow, I gave evidence to Parliament today on the opportunities and challenges facing the UK technology sector, alongside Alex Kendall, Erin Platts and Michal Holmes. Despite a lot of good news, we can't underestimate the scale of the challenge. NVIDIA is worth more than the FTSE 100 combined. Amazon's R&D budget is bigger than the entire UK private sector. X.ai built a datacenter in 19 days, which in the UK could have taken more than 19 months. The list goes on. If we want to be anything other than an also ran in AI, the UK needs more entrepreneurs and the resources to help them scale. Hopefully the budget will do that, and at the very least recognise our founders, engineers, SME owners and the self-employed as 'working people'... You can see the whole thing here: https://lnkd.in/eU9Jma9j
Executive Vice President Identity Verification Solutions at Regula. Entrepreneur, Advisor, Investor, Speaker, Thought Leader, Consultant, Company Director at various.
More of my thoughts have been published in a Forbes expert panel article on: "SMBs’ Pain Points In Working With Big Tech (And How To Fix Them)." https://lnkd.in/gtvKaXSy
The Justice Department and the Federal Trade Commission agreed to divide responsibility for investigating three major players in the artificial intelligence industry. Here at Humaina we have always been very vocal when it comes to Data Privacy and consumer rights and we see this a positive step in the right direction. However, we believe this is just a small band aid to address the growing concerns from the public with regards to how the data we generate using common productivity tools is being unapologetically used to train closed-source AI models for companies that are on a cult-like mission to build AGI (artifical general intelligence). If you refuse to agree to the updated T&C's and share your ideas, business plan, designs, customer's data with these platforms, you lose your access; meaning all the work you have stored on their servers and the network of clients and suppliers you have built over years. This investigation conducted by the US federal government is not likely to yield any measurable results, because those bodies rely on Big Tech to act as their advisors and fund them via lobbying. With the help of experienced consultants who understand the latest AI tools as well as the regulatory landscape you can optimize your business processes whilst staying true to your values and remaining fully in control of your data and AI safety. Contact us for an informal chat.
Founder & CEO of KYield. Pioneer in Artificial Intelligence, Data Physics and Knowledge Engineering.
Big news from the NYT this morning -- the AG will lead an antitrust investigation into the AI market with a focus on Microsoft, OpenAI and Nvidia. Although Nvidia may be a surprise to some, it would be difficult to investigate the AI market without also investigating Nvidia, with particular interest on supply of chips. If for example companies have been acquiring large amounts of specialty chips like H100 GPUs to prevent competition rather than for generating revenue, then that could be a prosecutable behavior. It's recently been reported that Microsoft is the largest customer of Nvidia, and Nvidia's CFO said that about 45% of their datacenter business is from Big Techs. Since Big Techs have more money than anyone else, much of which comes from monopolies, if they are using that market power to prevent competition, whether by attempting to buy all the chips, specialty talent, or startups, then that's clearly anticompetitive behavior and probably illegal. There is also no question in my mind and many others that's precisely what's been occurring for many years now. I think it's probably good news that the AG Antitrust Division is leading the investigation as they have a better track record than the FTC. The real question is how much the presidential election year is influencing this. I don't have enough evidence to comment on that other than to say I think the Big Tech antitrust issue is bipartisan, and appears to be becoming a much larger concern across society, industries, and countries.
Best-selling Author. Public Speaker. Founder of WOW
1wwow! can it beat you and me as the most succesful coaches in hockey?