Jeanee Rose Andrewartha’s Post

Headlines about energy use from AI will only increase and increase. An AI boom doesn't necessarily mean a bust for the planet but the concerns are not a surprise. GPUs have been used for years to churn out Bitcoin before deployment to turbo-charge AI. Bitcoin alone is estimated to to consume 127 terrawatt-hours a year. That is more than Norway. As Innovation in Tech demands more energy, hopefully innovative Tech will provide solutions. This can range top to bottom from advances in the way the AI ecosystem is designed to exploring greater use of green, sustainable energy sources for power and cooling. At RSM UK we shared thoughts about the impacts of AI on Data in our Real Economy report on AI last year (link below). Watch out for our flagship report on Sustainability later in July. https://lnkd.in/eTyDRmy9 Thomas Pugh | Rich Hall | Kathy Hobbs David Blacher | Mandy Girder | Richard Heap | Kirsty Fraser

❗ Limitless appetite for #AI based solutions increases power consumption and further grows our dependence on electricity as the only source of energy. Together with a precipitous transition to #e-mobility, we may find ourselves in direct contradiction with #sustainability commitments and deteriorating #resilience. 💡 Attached study shows how much more electricity is consumed (ten times more) compared to for example a simple Google search. It means that while we struggle to reduce our carbon dioxide emissions, we are actively increasing them. It also means that we will need to look for more efficient, more ecological and effective use of modern technologies. #Activism is generally recognised as a top #risk for business corporations and the society. It is perhaps the time to replace it with rationality if not too late. Multinational corporations need to lead the way and pressure Governments (including the European Commission in the EU) to adopt the rule of reason instead of giving in to the crowd. Policy makers should also inquire into clearly irrational use of electricity consuming technologies such as #cryptocurrency mining and transactions. ❓ By the way - I wonder how #ESG reports will reflect effects of AI based solutions on value chains. Any ideas on this from the most knowledgeable sustainability experts in my vicinity, Veronika Doubnerova or Zuzana Holá? 🙂

𝐈𝐬 𝐀𝐈’𝐬 𝐞𝐧𝐞𝐫𝐠𝐲 𝐚𝐩𝐩𝐞𝐭𝐢𝐭𝐞 𝐚 t𝐡𝐫𝐞𝐚𝐭 𝐭𝐨 𝐭𝐡𝐞 𝐩𝐥𝐚𝐧𝐞𝐭? - The tech world is abuzz with NVIDIA's latest chip. However, amidst the gushing praise and hype, there’s a deafening silence about the 𝐞𝐧𝐞𝐫𝐠𝐲 𝐚𝐧𝐝 𝐰𝐚𝐭𝐞𝐫 𝐜𝐫𝐢𝐬𝐢𝐬 brewing thanks to #AI.   AI’s hunger for water and power could take the risk of climate change to another level.   At current course and speed, AI is on a trajectory that could see it devour more electricity than many countries do today. Discussions at the World Economic Forum's meeting in Davos hinted that AI's hunger for energy might soon outpace the world's capacity to supply it sustainably.     Some data points:  🌍 ChatGPT, pulls over half a million kWh of electricity daily - equivalent to powering 180,000 U.S. households.    🌍 Data centers, use up 1.5 percent of global electricity, with AI’s insatiable appetite rapidly pushing that number higher.    🌍 NVIDIA alone could ship 1.5 million AI server units annually by 2027. These servers consume at least 85.4 terawatt-hours annually—more juice than some entire countries consume!  🌍 Those processors need fresh water to cool down. While training its GPT model, Microsoft's data centers in Des Moines, Iowa, slurped up 11.5 million gallons of water (6% of the entire district’s water for the month).    🌍 Researchers estimate that each interaction with ChatGPT guzzles roughly 500ml of water (a 16-ounce bottle to translate for Americans).   Recent articles have pointed to potential blackouts in the US this summer from anticipated increases in air conditioning use. 𝐖𝐢𝐥𝐥 𝐰𝐞 𝐬𝐞𝐞 𝐛𝐥𝐚𝐜𝐤𝐨𝐮𝐭𝐬 𝐟𝐫𝐨𝐦 𝐀𝐈 𝐮𝐬𝐚𝐠𝐞?   🌱 As AI grows, so does its appetite for power and water. Let’s talk about AI’s environmental impact, not just the cool things it can do. LINK: https://lnkd.in/eiNHR2CT? #EnergyCrisis #Sustainability

KEY FACTORS AND STRUCTURES OF THE DEVELOPMENT OF THE DIGITAL TRANSFORMATION OF THE ECONOMY #1avatarresearcher The article explores the impact of technological innovations on the economy and business in the context of digital transformation. It is investigated that key technological trends, such as 5G, Wi-Fi, energy-efficient processing, advanced data recognition, and intellectual data processing, cloud, and edge computing, converge to form the technological foundation necessary for successful digital transformation. As a result, digital transformation is often accompanied by significant losses, and costs exceeding the planned ones, and many projects lead to changes that do not justify expectations from the use of information technologies. Investments in analytics are necessary but do not guarantee a global restructuring of the business, its core transformation, during which even a minor error can halt the supply chain, destroy a product, or even put an end to the business. It is noted that to achieve success, it is necessary to prioritize strategy over technology. Under these conditions, the task of goal-setting comes to the forefront – it is necessary to start with defining the ultimate goal of the business (profit, perspective, strengthening market position, etc.) and only then move on to the search for technology that will allow achieving it. As businesses focus on providing greater mobility for employees, moving most corporate workflows to cloud storage, digitization provides staff with access to what they need and when they need it, supporting a high level of productivity regardless of location. The article also examines the constructs of economic development in the context of information technologies, focusing on the development of ecosystems of crypto assets, metaverses, and non-fungible tokens (NFTs), central bank digital currencies (CBDC), institutional investments in the digital economy, and tokenization of traditional funds. It is emphasized that these technological and economic transformations define new directions of development and require increased attention to cybersecurity, technological infrastructure, and collaboration between business and technical teams. Noosphere Laboratory European University #digital #ai https://lnkd.in/d4p3iPu2

⚡ It seems like balancing electricity demand with decreased emissions just can't catch a break time and time again. ⚡ Way back, when we all first started talking about climate issues, one of the easiest (at least to say) ways to reduce emissions was to reduce consumption, either due to efficiency improvements or flat out reduced use. But then along came data centres, with their need to power the internet and put the answer to every question at our fingertips. Round 1 goes to increased absolute demand. ⚡ Then along, almost out of nowhere, came cryptocurrency mining, essential to the creation of these financial instruments that don't exist without significant energy consupmtion. Round 2 goes to increased absolute demand. ⚡ Round 3 might be the parallel energy transition, which some might argue was a sideways issue because the move to electricity in some areas (like electric vehicles) replaces decreased fossil fuel use and is therefore a wash, or even an improvement. Alright, even if I give you a foul ball on that one, we're not done yet. ⚡ As if these were not bad enough, again out of nowhere, artificial intelligence and its super consumption of power and information sent the hydro meters spinning like the scene in Christmas Vacation when Chevy Chase finally gets the Christmas lights to work and the emergency generator has to kick in (https://lnkd.in/g2RKAZmM). ⚡ Whether you believe this is a baseball game analogy and you can't recover from three strikes, or it's a boxing match and you get a ten count before you are out, it does make it more difficult to achieve the same net result when you are adding new consumption avenues to the objective of a net decrease. Unless we can use AI to reduce the energy consumption in data centres and cryptocurrency mining?

https://lnkd.in/e3cagZpf I am not going to argue that we need to abandon AI. That would be pointless. Just like it would be to argue that about crypto and cloud computing. What I do want to argue is that we need IMMEDIATE solutions for the huge energy drain that these technologies create. So far, we've been using the earth as a credit card to pay for the energy for all these cool new technologies. We must change to a pay-as-you-go system. We must ONLY use additional renewable energy, not draw existing renewable energy from the grid. And every new technology must have this requirement built in. No new tech unless it adds 0 carbon to the environment. Otherwise, we are continuing to gleefully saw off the branch we are sitting on. And if you don't need to use AI, like to write a 5 sentence post, don't use AI!

A clip here where #Peter Thiel is stating that AI is to to be dominated by #San #Francisco. The point he makes about crypto is valid, and I do not have access to the whole #interview, but I cannot see how #AI, which heavily relies on data will be centralised in one location. The Internet boom was focused there, and so have all of the majority of the applications based on general compute. With the next wave of tech applications requiring #accelerated compute infrastructure that creates ubiquitous compute environments. coupled with #ubiquitous connectivity, the need for localised development and #localised innovation is more crucial. While the #innovation, through research and #development, from #Silicon #Valley has been an overall positive #contribution to #society, it has not been without its issues. The move to localising compute and enabling the localising of services maintains cultures and values not only creates jobs locally, but also, I think, make a #contribution to other parts of the world as more people get to experiment, test and ideate.

When I was a kid, I read a book called "Tas and the Postal Rocket", it's a story about a boy who accidentally stows away on a rocket used for delivering mail. Written in the 1950's the fictitious postal rocket was a machine that had the potential to deliver mail in 90mins to anywhere in the world. In the 50's this would have been revolutionary. To me, the 90's kid, this seemed silly. The postal rocket was made obsolete by the internet and email. During the industrial revolution steam powered machines replaced jobs, and many predicted that people would soon be made obsolete. Eventually machines would replace us. Instead, throughout the 19th and 20th century machines, and now electronics, are becoming more and more pervasive. Instead of replacing us, they're increasing our work output. Instead of working fewer hours, we expect higher standards of output. We can observe this in the workplace and domestically. Before the iron and the washing machine, clothes were washed more infrequently. The advent of this technically, increased our standards. We expect a sparkling clean crispy, well ironed shirt to be waiting for us in our cupboard everyday. When it comes to new technology, it's fun to try and predict how it'll change our lives. But, like Tas's postal rocket, sometimes we'll miss the mark. And that's fine. It's impossible to predict the future. Instead of asking how technologies like AI and Blockchain will change our lives it's often more helpful to ask, "What will remain the same?" What is fundamental about the human psyche that will never change? Read more here: https://rossgriffin.com/

So I was wandering: what forms of payment will be used in the 31st century? Some theories claim that money will become meaningless by the 24th century, so what if it does not? Below are some of the predictions; Biometric Payments: Future societies might use advanced biometric verification methods, such as fingerprint scans, retina scans, or even DNA verification, to authorize transactions. This could minimize fraud and make transactions incredibly fast and secure. Blockchain and Distributed Ledger Technologies: Beyond cryptocurrencies, these technologies could underpin most future financial transactions, offering unprecedented levels of transparency, security, and efficiency in transactions across the globe. Quantum Currency: With the advent of quantum computing, we might see the development of quantum currencies that are impossible to counterfeit and can be securely transferred without the need for traditional banks. Resource-based Economy: It's also possible that future societies could move towards a resource-based economy, as theorized by some futurists, where goods and services are freely available without the need for currency transactions, based on advanced production and distribution technologies. Social Credit Systems: Inspired by speculative fiction and some emerging real-world systems, transactions might not solely depend on monetary exchange but also on social credits earned through one's behavior or contributions to society. Interstellar Currency: If humanity expands beyond Earth, we might develop standardized currencies for interplanetary or even interstellar trade, possibly based on energy units or universally valued resources. I think quantum currency is exciting. I will look into this further and see what information I can find. I appreciate science fiction shows like Star Trek because they expand one's thinking to endless possibilities. This has an impact on pivoting because your imagination and instincts are all you have in the event that data is insufficient to make certain decisions. And, of course, you'll encounter this frequently in product management. When data is insufficient, you must rely on your instincts and imagination. That is what gave birth to our contemporary world, and it will continue to do so for many years. Image source: Star Trek Discovery

Artificial intelligence requires a lot of power. The kind of machine learning that produced ChatGPT relies on models that process fantastic amounts of information, and every bit of processing takes energy. It’s been estimated that ChatGPT is responding to something like two hundred million requests per day, and, in so doing, is consuming more than half a million kilowatt-hours of electricity. (For comparison’s sake, the average U.S. household consumes twenty-nine kilowatt-hours a day.) A.I. could potentially be used to improve the efficiency of renewable-energy systems, which could reduce emissions from server farms.