Cititec Talent’s Post

The quick development of AI technology since ChatGPT's launch in November 2022 has given rise to serious social, economic, and environmental issues. While AI's heavy dependence on energy-intensive data centers increases worldwide electricity consumption and carbon emissions, European regulators and financial institutions are closely examining the technology's effects on inflation and privacy. Large tech companies like Microsoft, Meta, and Google have shown significant increases in emissions as a result of their investments in AI, underscoring the pressing issues with water and energy use. Addressing these environmental effects and enhancing the sustainability of data management are vital worldwide imperatives as AI deployment picks up speed. #AI #ArtificialIntelligence #TechGiants #DataCenters #CarbonFootprint #EnvironmentalImpact #Sustainability #ClimateChange #ClimateFinance #RenewableEnergy #EnergyEfficiency #GlobalBank #InnovationLab #DisasterRiskReduction #PrivacyIssues

Ever wondered what powers your favorite AI tools behind the scenes? It's not coffee or energy drinks; it's WATER, and lots of it. AI companions require hydration, with ChatGPT alone consuming 17 ounces of water to process 5-50 prompts, enough to fill your reusable water bottle. Here's why: The computers driving AI generate heat and rely on water for cooling. But this water consumption is skyrocketing; Microsoft reported a 34% increase, and Google saw 22% in 2022. This raises questions about the environmental impact. Can we sustain AI's hydration without harming local communities and wildlife? How does it align with business environmental pledges? While we await OpenAI's perspective, we can help by conserving energy. Turning off power-hungry tech when unnecessary is a small step toward a greener future. Has this changed your view of AI? #AI #Sustainability #EnvironmentalImpact https://lnkd.in/gE5QJsJ5

Powered By Coffee. That’s me. Ever wonder what’s powering your favorite AI tool be-hind the scenes? Well it’s not powered by coffee, sipping a latte or guzzling an energy drink, however it is drinking… WATER, and lots of it. You see, our AI friends require hydration. Lots of it. Research shows ChatGPT alone gulps down about 17 ounces of water to process between five and 50 prompts. That's enough to fill your reusable water bottle. Let me explain. The computers that power our beloved AI tools generate a huge amount of heat. To stop them overheating, water is used in a cooling system. It's like giving them a refreshing dip in the pool, but this pool party has some serious consequences. Microsoft reported a 34% year-on-year increase in water consumption for 2022, while Google noted a 22% rise. Imagine if humans increased our water consumption at that rate? The question is, can we afford to keep our AI tools so well-hydrated at the expense of lo-cal populations and wildlife? And are we prepared to change the course of rivers to keep our digital assistants cool and collected? Considering the water shortages in the Southwest, AI servers may not be welcome in certain parts of the United States. My feeling is that this is not going away any time soon, and in fact as AI becomes ubiquitous, this issue will grow. While we wait for OpenAI to weigh in on this dilemma, perhaps we could turn off the tap on our power-hungry tech when it's not totally necessary. After all, every little bit helps. Does this change your view of AI, or are you not surprised? #AI #Sustainability #WaterConservation https://lnkd.in/gYZMCdVS

We all remember the movie Minority Report. Tom Cruise is accessing data at his fingertips. The data is helping him predict and act on it before anything happens. The most impressive thing about this movie from 2002 was the effort put into creating the world. This was not explained in the book. Three years before making the movie, Steven Spielberg reunited experts. The goal was to imagine the future. A lot of the tech from the movie is available today. Those imagined technologies are available today. 20-years later, we got driverless cars, personalized advertisements and voice-controlled homes. Artificial Intelligence supports and helps the world function. It is the foundation of everything. Recommendation engines help people choose the next movie or the next buy. In the past year, AI has appeared as a tool in our daily lives and offered an opportunity to all, not a select few. AI has been controlling part of our lives for years already. The tech giants have been using it to predict what will be the most relevant for us. It could be the best route to take on Google Maps or what is on your social feed. ChatGPT release brought the democratisation of AI. Everyone has access to a new tool that can give them an advantage and a new power. Read more... By Phil Araujo https://lnkd.in/grq_uPsy #productdevelopment #productmanagement

How are YOU using AI? In my board advisory work, this question often leads to........SILENCE The reality is AI is here - embedded into Microsoft apps, Google, Apple - your day to day activities. Plus Gen AI - ChatGPT, Dalle, Firefly, Gemini, Perplexity are easily available, safe (with responsible use) tools everyone can and should use to kickstart, speed, check, validate or just give a different option to daily tasks. This does not mean - turn over the task to AI and walk away.. but requires human engagement - in defining the problem to be addressed with AI, creating the prompts, carefully reviewing the results, discerningly looking for errors or hallucinations and integrating the results with human ingenuity. People PLUS AI will co-create a better future for all. #future #AI #leadership #boards

The water is of little concern. The power consumption is costly. Very costly. And to scale up AI to replace all those pesky humans the AGI enthusiasts keep telling us about, they're going to need to process daily requests certainly several orders of magnitude greater than 200 million. Some mock the EV enthusiasts, asking where we will generate sufficient electricity to charge all the EVs they dream of, when we really ought to be mocking the AGI hypesters claiming their AI will replace millions of jobs. The question is where will the power come from when all those out of work humans are out enjoying a Sunday drive in their EV?

I made a big mistake this week. I assumed general awareness of this crucial data point around GenAI. The rise of accessibility of this technology now poses one of the biggest known threats to our natural resources. Research has shown that ChatGPT (and other models) consume 16oz of water (1 bottle) for every 5-50 prompts. Microsoft has publicly committed to working to find alternate energy solutions to mitigate this crisis. And we all have a responsibility to take action. What can we do? 1. When designing AI strategy, let’s include our ESG teams as key stakeholders. 2. Carefully consider licenses and users- the level of access to this technology is widespread in public domain. Does that make it necessary to provide access to all in our enterprise as well? Engage in robust decision making to determine the best path forward 3. Go in ‘eyes wide open’ about what the investment in AI means from an offset perspective in addition to the task at hand. We cannot leave the responsibility of our future to the tech giants to sort out. We all have a role in defining our impact on our future. #genaiethics

The dark side of AI. 2 facts: - 20 ChatGPT prompts use 1 liter of water (to cool down servers) - Annually, AI's carbon footprint is approaching 1% of global emissions A recent study projects that by 2027, NVIDIA’s new AI servers will be consuming over 85.4 terawatt-hours annually, exceeding the energy usage of countries such as Sweden and Argentina. https://lnkd.in/eQBZTxDC Shall we use AI trying to solve climate change or keep usage to the minimum? What do you think?

We asked ChatGPT about its carbon footprint. It responded in an unexpectedly candid manner, shining light on both its developer and the AI industry's lack of transparency on the subject. As a general-purpose technology, AI promises to deliver enormous benefits, including in terms of climate action. For instance, AI is already assisting in making air travel more fuel-efficient, in developing faster electric batteries, or in climate adaptation by improving forecast models for extreme weather. Recently, AI helped solve a major plasma confinement challenge, bringing us one step closer to safe fusion power. But how much energy is needed to deliver these breakthroughs? We know that large generative AI systems will soon consume more resources than entire nations (as much electricity as Japan and half of the UK’s freshwater use by 2027, according to recent reports). However, the exact environmental footprint of AI remains a closely guarded secret. In the fast-paced AI industry, companies fiercely compete by developing increasingly sophisticated and efficient algorithms. Disclosing the details of the infrastructure that underpins these breakthroughs (such as the location and efficiency of data centres or the size and optimization of AI models) could give rivals the blueprint they need to replicate success or even surpass the original innovators. Aware of the problem (a group of Facebook researchers called it the “elephant in the room”), AI companies have started making carbon-neutral pledges and releasing best practices that promise to reduce the footprints of AI systems: developing sparse models with fewer parameters (the evidence so far is that models are becoming larger and more complex); and relying on more efficient processors and data centres. The issue is that it is difficult to know whether they’re implementing them since AI companies disclose very limited information. But without a better understanding of just how much energy AI systems consume, AI companies and developers will have a hard time reducing it or driving the development of more energy-efficient technologies. Ultimately, if you don’t measure it, you can’t improve it. The development of AI holds lots of promises for climate action, but AI companies’ lack of accountability in terms of their environmental footprint cannot be addressed with voluntary measures alone. Enforcing clear environmental reporting requirements for AI companies is a necessary step to steer the industry towards a sustainable future. #sustainability #carbonfooptrint #GHGemissions #AI #LLMs #ChatGPT #OpenAi

My ChatGPT-4o (omni) comments: - It is free and faster to generate text; I wonder how many AI Startups will fail due to that unfair competition (free Microsoft copilot, Google Gemini, too). I would surely not want to compete. But if you check Producthunt, 80% of new products listed for the past 6 months are "AI." - NVIDIA is happy selling the hardware. Electricity consumption is boosted, and I wonder about CO2 and the environment. - the Italian speech in the demo video with their CTO sounds like a fake accent; only Italians could tell that, but I am one. - described exactly a few "cat on chair" photos, including hints of partial objects and people in the backgrounds. This works. PS: No AI was used to write this post. I am open to working for you on what AI can't do yet; click on my profile for my story.