Check Houston area power outages for Centerpoint and <strong>Entergy</strong>, see the outage map and how to report a power outage.

Houston power outages: Centerpoint, Entergy tracker; report an outage

Southeast Texas citizens are no strangers to inclement weather and lengthy power outages in the aftermath. However, with these upgrades that may be changing soon.

Entergy completes $5M project to upgrade Port Arthur's power infrastructure

Entergy Texas announced Thursday it recently completed a series of infrastructure upgrades that will help protect the City of Port Arthur from extreme weather events. The service reliability project, company officials said, consisted of upgrades designed to the newly developed extreme wind ...

Replacing 240 utility poles and aging power lines are part of Entergy upgrades in Port Arthur - Port Arthur News | Port Arthur News

Entergy improves PA electric grid to help weather strong storms. (photo credit: Entergy Texas)

Entergy improves PA electric grid to help weather strong storms | KFDM

Entergy Corp (NYSE:ETR) recently announced a dividend of $1.13 per share, payable on 2024-03-01, with the ex-dividend date set for 2024-02-08. As investors look forward to this upcoming payment, the spotlight also shines on the company's dividend history, yield, and growth rates.

Entergy Corp's Dividend Analysis

As tax season begins it’s important to know what you qualify for and know the people out there who can help make the process easier.

Entergy Arkansas offering customers free tax help | KARK

Utility giant Entergy’s plans to sell its gas distribution business is moving forward under a $484 million cash deal with Bernhard Capital Partners, a Baton Rouge-based private equity management firm. The two companies announced the pending deal on Monday.

Louisiana company set to acquire Entergy gas distribution business

This week marks the end of an era for Entergy, as the company agreed to sell off its gas distribution service to Bernhard Capital Partners, and with it, the more than 200,000 customers it served ... R

Entergy to sell gas distribution business to Bernhard Capital Partners in $484M deal

Utility giant Entergy’s plans to sell its gas distribution business is moving forward under a $484 million cash deal with Bernhard Capital Partners, a Baton Rouge-based private equity management firm.

Baton Rouge company set to acquire Entergy gas distribution business

Q3 revenues of $3,595.5 million missed the Zacks Consensus Estimate by 14.9%. The figure also declined 14.8% from the year-ago quarter???s level.

Entergy's (ETR) Q3 Earnings Beat Estimates, Revenues Fall Y/Y

Entergy

entergy

Have you ever been in the middle of an important task, only for your Mac to dim the screen or go to sleep at the worst possible moment? Whether you’re giving a presentation, running a long download, monitoring real-time data, or just stepping away briefly, there are plenty of situations where you might need your Mac to stay awake. Fortunately, there are multiple ways to prevent your Mac from sleeping — some built-in, and others requiring a little extra help. Let’s explore the best options. Adjust Energy Saver Settings Adjust Energy Saver Settings One way to prevent your Mac from sleeping is to modify the system settings: For macOS Ventura and later: For macOS Ventura and later: Open System Settings → Displays → Advanced
Enable Prevent automatic sleeping when the display is off Open System Settings → Displays → Advanced Enable Prevent automatic sleeping when the display is off For macOS Monterey and earlier: For macOS Monterey and earlier: Go to System Preferences → Energy Saver
Adjust the Turn display off after slider to “Never” Go to System Preferences → Energy Saver Adjust the Turn display off after slider to “Never” This method works well if you only need to keep the system awake while plugged in. However, it doesn’t always prevent applications from detecting inactivity. Use Terminal Command Use Terminal Command For those comfortable with the command line, macOS provides a built-in tool called caffeinate that prevents sleep temporarily. How to use it: Open Terminal
Type: caffeinate -d Open Terminal Type: caffeinate -d This keeps your Mac awake until you close the Terminal window. To stop it, press Control + C or close the Terminal window. You can also use additional options like caffeinate -t 3600 to keep the Mac awake for an hour. External Hardware Solutions Some users opt for external hardware-based mouse jigglers, which are small USB devices that move the cursor automatically. These are plug-and-play solutions that don’t require software installation. While effective, they’re less flexible compared to software options. Third-Party Apps for Keeping Your Mac Awake Sometimes, built-in settings aren’t enough, especially if certain applications check for activity beyond system-level sleep prevention. That’s where third-party apps come in. Mouse Jiggler — For Simulating Activity Mouse Jiggler — For Simulating Activity Unlike traditional sleep prevention apps, Mouse Jiggler works by simulating actual cursor movement, which some applications and policies require to detect activity. When this is useful: When this is useful: • Remote Desktop & Virtual Machines — Apps like Citrix, VMware, and Microsoft RDP may log you out if there’s no mouse or keyboard activity. • Web Applications & Security Portals — Some platforms enforce automatic logouts unless movement is detected. • Live Monitoring Dashboards — Stock trading platforms, analytics dashboards, or real-time monitoring tools often pause or time out without user interaction. • Software Testing & Automation — Developers and IT professionals can use it to simulate user input during testing. Mouse Jiggler also includes Silent Mode, which moves the cursor invisibly to avoid interfering with workflows, along with customizable deactivation options. Download for free from the App Store: https://apps.apple.com/app/id6740313656 Download for free from the App Store https://apps.apple.com/app/id6740313656 Which Method Is Right for You? The best way to keep your Mac awake depends on your needs: ✅ For presentations, sleep prevention or preventing logouts in remote sessions → Mouse Jiggler provides actual movement detection. ✅ For command-line users → Terminal’s caffeinate command is a quick and easy option. ✅ For a plug-and-play solution → A hardware jiggler might be a good choice. Keeping your Mac awake doesn’t have to be complicated. Whether you prefer a built-in option or a third-party tool, there’s a solution that fits your workflow. What’s your preferred way to keep your Mac awake? Let us know in the comments! 🚀

How to Adjust Your Mac's Energy Saver Settings to Keep It Awake For Longer

When OpenAI launched ChatGPT in late 2022, it sparked both delight and concern. Generative AI demonstrated remarkable potential—crafting essays, solving coding problems, and even creating art. But it also raised alarms among environmentalists, researchers, and technologists. The biggest concern? The massive energy consumption required to train and run Large Language Models (LLMs), prompting questions about their long-term sustainability. As LLMs continue to reshape industries like education and healthcare, their impact can't be ignored. This paper raises an important question: Can these intelligent systems optimize themselves to reduce power consumption and minimize their environmental footprint? And if so, how might this transform the AI landscape? We’ll break down the energy challenges of LLMs, from training to inference, and explore innovative self-tuning strategies that could make AI more sustainable. Understanding the AI Energy Challenge Understanding the AI Energy Challenge Training vs. Inference Training vs. Inference Google's training of large language models such as  GPT-4 or PaLM demands a huge amount of computational resources. For example, training GPT-3 took thousands of GPUs running for weeks, consuming as much energy as hundreds of U.S. households in a year. The carbon footprint depends on the energy mix powering data centers. Even after training, the inference phase—where models handle real-world tasks—adds to energy use. Although the energy required for a single query is small, when we consider that there are billions of such interactions taking place across various platforms every day, it becomes a significant problem. Why do LLMs  Consume So Much Energy? Why do LLMs  Consume So Much Energy? Model Size: Today’s LLMs are parameter sensitive; they have billions or even trillions of parameters that require a lot of resources to be processed, updated, and stored.


Hardware Constraints: The use of silicon-based chips is limited by their processing capacities and thus the need for clusters of GPUs or TPUs to increase energy use exponentially. Model Size: Today’s LLMs are parameter sensitive; they have billions or even trillions of parameters that require a lot of resources to be processed, updated, and stored. Model Size: Today’s LLMs are parameter sensitive; they have billions or even trillions of parameters that require a lot of resources to be processed, updated, and stored. Model Size: Hardware Constraints: The use of silicon-based chips is limited by their processing capacities and thus the need for clusters of GPUs or TPUs to increase energy use exponentially. Hardware Constraints: The use of silicon-based chips is limited by their processing capacities and thus the need for clusters of GPUs or TPUs to increase energy use exponentially. Hardware Constraints: Cooling Needs: Data centers supporting high computational workloads are warm and the cooling systems can consume as much as 40 % of the power if they are not energy efficient. Cooling Needs: Data centers supporting high computational workloads are warm and the cooling systems can consume as much as 40 % of the power if they are not energy efficient. Cooling Needs: Environmental and Economic Toll Environmental and Economic Toll The costs in terms of the environment include the carbon emissions as well as water usage in cooling while the operational expenses are a problem for the smaller AI companies. The annual costs may reach billions, which makes sustainability an important not only environmental but also economic issue. AI Model Energy Consumption Breakdown AI Model Energy Consumption Breakdown To understand how LLMs consume energy, let’s break it down: AI Operation

Energy Consumption (%)



Training Phase

60%



Inference (Running Queries)

25%



Data Center Cooling

10%



Hardware Operations

5% AI Operation

Energy Consumption (%)



Training Phase

60%



Inference (Running Queries)

25%



Data Center Cooling

10%



Hardware Operations

5% AI Operation

Energy Consumption (%) AI Operation AI Operation AI Operation Energy Consumption (%) Energy Consumption (%) Energy Consumption (%) Training Phase

60% Training Phase Training Phase 60% 60% Inference (Running Queries)

25% Inference (Running Queries) Inference (Running Queries) 25% 25% Data Center Cooling

10% Data Center Cooling Data Center Cooling 10% 10% Hardware Operations

5% Hardware Operations Hardware Operations 5% 5% Key Takeaway: The training phase remains the biggest contributor to power consumption. Key Takeaway: The training phase remains the biggest contributor to power consumption. Key Takeaway: Strategies for Self-Optimization Strategies for Self-Optimization Researchers are looking into how LLMs can optimize their energy use, combining software work with hardware changes. Model Pruning and Quantization Model Pruning and Quantization Pruning: Redundant parameters that affect accuracy to a limited extent are removed, resulting in a reduction in the size of the model without compromising the accuracy.
Quantization: This reduces the precision (e.g., from 32-bit to 8-bit) of the data, which reduces the memory and computational requirements. Pruning: Redundant parameters that affect accuracy to a limited extent are removed, resulting in a reduction in the size of the model without compromising the accuracy. Pruning: Quantization: This reduces the precision (e.g., from 32-bit to 8-bit) of the data, which reduces the memory and computational requirements. Quantization: Quantization and Pruning are useful but when used with feedback loops where a model is able to determine which parts are crucial and which parts can be quantized then it becomes quite effective. This is a new area, but the potential exists in self-optimizing networks. Dynamic Inference (Conditional Computation) Dynamic Inference (Conditional Computation) The idea of conditional computation enables the models to use only those neurons or layers that are relevant to a given task. For instance, Google's  Mixture-of-Experts (MoE) approach divides the network into specialized subnetworks that enhance training and reduction in energy consumption by limiting the number of active parameters. Reinforcement Learning for Tuning Reinforcement Learning for Tuning Reinforcement learning can optimize hyperparameters like learning rate and batch size, balancing accuracy and energy consumption to ensure models operate efficiently. Multi-Objective Optimization In addition to optimizing for accuracy, LLMs can also optimize for other objectives: accuracy, latency, and power consumption, using tools such as Google Vizier or  Ray Tune. Recently, energy efficiency has become a crucial objective in these frameworks. Hardware Innovations and AI Co-Design Hardware Innovations and AI Co-Design Application Specific Integrated Circuits (ASICs): Special purpose chips to improve efficiency in the execution of AI tasks.
Neuromorphic Computing: Brain-inspired chips, still in development to minimize power consumption when performing neural network computations are under development.
Optical Computing: Computation using light could overcome the limitations of the electronic system to scale down the power consumption of the system. Application Specific Integrated Circuits (ASICs): Special purpose chips to improve efficiency in the execution of AI tasks. Application Specific Integrated Circuits (ASICs): Neuromorphic Computing: Brain-inspired chips, still in development to minimize power consumption when performing neural network computations are under development. Neuromorphic Computing: Optical Computing: Computation using light could overcome the limitations of the electronic system to scale down the power consumption of the system. Optical Computing: AI systems created through the co-design of hardware with software allow for the simultaneous adjustment of software algorithms and hardware resources. Comparing AI Energy Optimization Techniques Comparing AI Energy Optimization Techniques Technique

Energy Reduction (%)

Primary Benefit



Model Pruning

30%

Reduces unnecessary model parameters



Quantization

40%

Lowers computational precision



Conditional Computation (MoE)

25%

Activates only necessary model



Reinforcement Learning

15%

Dynamically adjusts power usage



Neuromorphic Computing

50%

Emulates brain efficiency



Hardware Co-Design (ASICs, Optical Chips)

35%

Develops AI-specific hardware for maximum efficiency Technique

Energy Reduction (%)

Primary Benefit



Model Pruning

30%

Reduces unnecessary model parameters



Quantization

40%

Lowers computational precision



Conditional Computation (MoE)

25%

Activates only necessary model



Reinforcement Learning

15%

Dynamically adjusts power usage



Neuromorphic Computing

50%

Emulates brain efficiency



Hardware Co-Design (ASICs, Optical Chips)

35%

Develops AI-specific hardware for maximum efficiency Technique

Energy Reduction (%)

Primary Benefit Technique Technique Technique Energy Reduction (%) Energy Reduction (%) Energy Reduction (%) Primary Benefit Primary Benefit Primary Benefit Model Pruning

30%

Reduces unnecessary model parameters Model Pruning Model Pruning 30% 30% Reduces unnecessary model parameters Reduces unnecessary model parameters Quantization

40%

Lowers computational precision Quantization Quantization 40% 40% Lowers computational precision Lowers computational precision Conditional Computation (MoE)

25%

Activates only necessary model Conditional Computation (MoE) Conditional Computation (MoE) 25% 25% Activates only necessary model Activates only necessary model Reinforcement Learning

15%

Dynamically adjusts power usage Reinforcement Learning Reinforcement Learning 15% 15% Dynamically adjusts power usage Dynamically adjusts power usage Neuromorphic Computing

50%

Emulates brain efficiency Neuromorphic Computing Neuromorphic Computing 50% 50% Emulates brain efficiency Emulates brain efficiency Hardware Co-Design (ASICs, Optical Chips)

35%

Develops AI-specific hardware for maximum efficiency Hardware Co-Design (ASICs, Optical Chips) Hardware Co-Design (ASICs, Optical Chips) 35% 35% Develops AI-specific hardware for maximum efficiency Develops AI-specific hardware for maximum efficiency Future AI models will likely combine multiple techniques to achieve 60-70% overall energy reduction. Future AI models will likely combine multiple techniques to achieve 60-70% overall energy reduction. Future AI models will likely combine multiple techniques to achieve 60-70% overall energy reduction. Challenges to  Self-Optimizing AI Challenges to  Self-Optimizing AI Accuracy Trade-offs: Some features, such as pruning and quantization, may compromise accuracy slightly.
Data Center Infrastructure Limits: We are still operating under the assumption of reliance on inefficient silicon chips.
Energy Performance Measures Gaps: There is currently no universal standard for tracking energy efficiency.
Government Regulation: Strict sustainability rules may force the adoption of efficient models. Accuracy Trade-offs: Some features, such as pruning and quantization, may compromise accuracy slightly. Accuracy Trade-offs Data Center Infrastructure Limits: We are still operating under the assumption of reliance on inefficient silicon chips. Data Center Infrastructure Limits: Energy Performance Measures Gaps: There is currently no universal standard for tracking energy efficiency. Energy Performance Measures Gaps: Government Regulation: Strict sustainability rules may force the adoption of efficient models. Government Regulation: Future Implications Future Implications Self-optimizing LLMs could reduce energy consumption by 20% or more for billions of queries, which would lead to enormous cost and emission savings. This is consistent with global net zero targets and impacts several sectors: Enterprise: Energy-efficient LLMs could increase uptake in customer service and analytics.
Research: Open source initiatives like Hugging Face may further speed innovation.
Policy: Standards on energy transparency could push self-optimization as a norm. Enterprise: Energy-efficient LLMs could increase uptake in customer service and analytics. Enterprise Research: Open source initiatives like Hugging Face may further speed innovation. Research Policy: Standards on energy transparency could push self-optimization as a norm. Policy Conclusion Conclusion LLMs have brought in a new level of sophistication in language processing but the problem of their energy consumption is a major concern. However, the same intelligence that gave rise to these models provides the solution. Techniques like pruning, quantization, conditional computation, and hardware co-design indicate that it is possible to design LLMs that manage their own energy consumption. As the research advances, the issue becomes less of whether sustainable AI is possible and more of how quickly the tech industry can come together to achieve it—without sacrificing innovation for the environment. References References Brown, T., et al. (2020). "Language Models are Few-Shot Learners." Advances in Neural Information Processing Systems, 33, 1877-1901. (Hypothetical source for GPT-3 training data.)
Strubell, E., Ganesh, A., & McCallum, A. (2019). "Energy and Policy Considerations for Deep Learning in NLP." Proceedings of the 57th Annual Meeting of the ACL, 3645-3650. (Illustrative source on AI energy costs.)
Fedus, W., et al. (2021). "Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity." arXiv preprint arXiv:2101.03961. (Basis for Mixture-of-Experts discussion.)
Patterson, D., et al. (2021). "Carbon Emissions and Large Neural Network Training." arXiv preprint arXiv:2104.10350. (Source for training energy estimates.)
Google Research. (2023). "Vizier: A Service for Black-Box Optimization." Google AI Blog. (Illustrative tool reference.) Brown, T., et al. (2020). "Language Models are Few-Shot Learners." Advances in Neural Information Processing Systems, 33, 1877-1901. (Hypothetical source for GPT-3 training data.) Advances in Neural Information Processing Systems Strubell, E., Ganesh, A., & McCallum, A. (2019). "Energy and Policy Considerations for Deep Learning in NLP." Proceedings of the 57th Annual Meeting of the ACL, 3645-3650. (Illustrative source on AI energy costs.) Proceedings of the 57th Annual Meeting of the ACL Fedus, W., et al. (2021). "Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity." arXiv preprint arXiv:2101.03961. (Basis for Mixture-of-Experts discussion.) arXiv preprint arXiv:2101.03961 Patterson, D., et al. (2021). "Carbon Emissions and Large Neural Network Training." arXiv preprint arXiv:2104.10350. (Source for training energy estimates.) arXiv preprint arXiv:2104.10350 Google Research. (2023). "Vizier: A Service for Black-Box Optimization." Google AI Blog. (Illustrative tool reference.) Google AI Blog

AI’s Energy Dilemma: Can LLMs Optimize Their Own Power Consumption?

KI se energie-dilemma: kan LLM's hul eie kragverbruik optimaliseer?

AI’s Energía Dilema: ¿LLMs jupanakax jupanakan ch’ama apnaqawip sum apnaqapxaspati?

Das Energiedilemma der KI: Können LLMs ihren eigenen Stromverbrauch optimieren?

El dilema energético de la IA: ¿pueden los LLM optimizar su propio consumo energético?

דילמת האנרגיה של AI: האם LLMs יכולים לייעל את צריכת החשמל שלהם?

एआई की ऊर्जा दुविधा: क्या एलएलएम अपनी स्वयं की ऊर्जा खपत को अनुकूलित कर सकते हैं?

AI のエネルギージレンマ: LLM は独自の電力消費を最適化できるか?

AI-ийн эрчим хүчний асуудал: LLM-үүд өөрсдийн эрчим хүчний хэрэглээг оновчтой болгож чадах уу?

Dilema Tenaga AI: Bolehkah LLM Mengoptimumkan Penggunaan Kuasa Mereka Sendiri?

AI හි බලශක්ති උභතෝකෝටිකය: LLM වලට තමන්ගේම බලශක්ති පරිභෝජනය ප්‍රශස්ත කළ හැකිද?

Energetická dilema AI: Môžu LLM optimalizovať svoju vlastnú spotrebu energie?

Ingxaki yaMandla ye-AI: Ngaba ii-LLMs ziyakwazi ukusebenzisa amandla azo?

Mass and Energy Flow in Jupiter’s Magnetosphere

"The future is renewable." That’s what we’ve heard for years. So why does the present still run on fossil fuels? The problem isn’t technology. We have solar panels with record-breaking efficiency, smart grids, and storage solutions that can hold more energy than ever. And yet, fossil fuels are still the main option. Not because we lack the tools, but because we’ve failed to make the shift economically inevitable. It’s naive to believe that simply understanding the consequences of fossil fuel dependence will be enough to drive change. People don’t switch to better systems just because it’s the right thing to do. They switch when it makes financial sense, when the cost of staying in the past outweighs the benefits of moving forward. The Incomplete Transition From a pragmatic standpoint, generating your own energy is still a tough sell. It’s not just about the upfront cost, it’s also about the time, effort, and mental load that comes with it. The perceived benefits rarely outweigh the hassle, making personal renewable energy a luxury rather than a viable option. Installing a solar system — the most accessible route for small-scale producers — requires a significant investment, one that few are willing to make without a clear and immediate return. And while climate messaging and marketing campaigns are everywhere, good intentions don’t pay the bills. Without real, tangible incentives, the energy transition will remain an aspiration rather than a reality. People don’t switch to cleaner energy because they read an IPCC report. They switch when the numbers make sense, and when they see a better option. Right now, the system is designed so that: Consumers have no control → You pay whatever your energy provider decides.
Small producers are stuck selling cheap → If you generate surplus energy, you’re forced to sell it to the grid at a price set by utilities.
Grid inefficiencies benefit middlemen → Energy distribution is still controlled by large corporations, meaning friction, inefficiencies, and profit siphoning. Aligning Incentives Through Decentralization If the core issue is the lack of clear, immediate incentives, then the solution has to be a system where users see and feel the benefits of renewable energy firsthand. One way to dismantle the outdated energy model is through peer-to-peer (P2P) energy trading, powered by blockchain technology as a secure, transparent, and decentralized framework. This system shifts the power dynamic in the energy market. Small-scale producers can now trade their surplus energy on their own terms, without being at the mercy of centralized pricing structures. By giving producers control over their output and pricing, this model doesn’t just empower individuals — it creates the economic incentives necessary for new small-scale investments, making distributed generation an economically viable and scalable reality. On a technical level, blockchain ensures a fully transparent and tamper-proof record of energy transactions, tracking both production and distribution. Producers inject their surplus energy into the grid, while bidirectional smart meters facilitate real-time communication with the grid operator, ensuring seamless validation of transactions. If we want people to make better choices, we need to give them a reason that makes sense for them. Effective incentives don’t just drive adoption — they make it effortless and financially rewarding. The key to an energy transition isn’t imposing restrictions or guilt, it’s making renewable energy economically attractive. If you're building (or breaking) in this space, let’s talk. "The future is renewable." That’s what we’ve heard for years. So why does the present still run on fossil fuels? "The future is renewable." That’s what we’ve heard for years. So why does the present still run on fossil fuels? The problem isn’t technology. We have solar panels with record-breaking efficiency, smart grids, and storage solutions that can hold more energy than ever. And yet, fossil fuels are still the main option. Not because we lack the tools, but because we’ve failed to make the shift economically inevitable. economically inevitable. It’s naive to believe that simply understanding the consequences of fossil fuel dependence will be enough to drive change. People don’t switch to better systems just because it’s the right thing to do. They switch when it makes financial sense , when the cost of staying in the past outweighs the benefits of moving forward. They switch when it makes financial sense The Incomplete Transition From a pragmatic standpoint, generating your own energy is still a tough sell. It’s not just about the upfront cost, it’s also about the time, effort, and mental load that comes with it. The perceived benefits rarely outweigh the hassle, making personal renewable energy a luxury rather than a viable option. Installing a solar system — the most accessible route for small-scale producers — requires a significant investment, one that few are willing to make without a clear and immediate return. And while climate messaging and marketing campaigns are everywhere, good intentions don’t pay the bills. Without real, tangible incentives, the energy transition will remain an aspiration rather than a reality. People don’t switch to cleaner energy because they read an IPCC report. They switch when the numbers make sense, and when they see a better option. Right now, the system is designed so that: Consumers have no control → You pay whatever your energy provider decides. Small producers are stuck selling cheap → If you generate surplus energy, you’re forced to sell it to the grid at a price set by utilities. Grid inefficiencies benefit middlemen → Energy distribution is still controlled by large corporations, meaning friction, inefficiencies, and profit siphoning. Consumers have no control → You pay whatever your energy provider decides. Consumers have no control Small producers are stuck selling cheap → If you generate surplus energy, you’re forced to sell it to the grid at a price set by utilities. Small producers are stuck selling cheap Grid inefficiencies benefit middlemen → Energy distribution is still controlled by large corporations, meaning friction, inefficiencies, and profit siphoning. Grid inefficiencies benefit middlemen Aligning Incentives Through Decentralization If the core issue is the lack of clear, immediate incentives, then the solution has to be a system where users see and feel the benefits of renewable energy firsthand. One way to dismantle the outdated energy model is through peer-to-peer (P2P) energy trading, powered by blockchain technology as a secure, transparent, and decentralized framework. This system shifts the power dynamic in the energy market. Small-scale producers can now trade their surplus energy on their own terms, without being at the mercy of centralized pricing structures. By giving producers control over their output and pricing, this model doesn’t just empower individuals — it creates the economic incentives necessary for new small-scale investments, making distributed generation an economically viable and scalable reality. On a technical level, blockchain ensures a fully transparent and tamper-proof record of energy transactions, tracking both production and distribution. Producers inject their surplus energy into the grid, while bidirectional smart meters facilitate real-time communication with the grid operator, ensuring seamless validation of transactions. If we want people to make better choices, we need to give them a reason that makes sense for them. Effective incentives don’t just drive adoption — they make it effortless and financially rewarding. The key to an energy transition isn’t imposing restrictions or guilt, it’s making renewable energy economically attractive. If we want people to make better choices, we need to give them a reason that makes sense for them. Effective incentives don’t just drive adoption — they make it effortless and financially rewarding. The key to an energy transition isn’t imposing restrictions or guilt, it’s making renewable energy economically attractive. If you're building (or breaking) in this space, let’s talk. If you're building (or breaking) in this space, let’s talk.

Clean Energy Is a UX Problem

Once, my little son asked me why people fear the dark. I told him that our brains constantly dream, even when we’re awake. When awake, this dream is in sync with the outside world we perceive (what we see, hear, etc.). When it’s dark, visual stimuli disappear, and our brains start filling the void. And why does it fill it with scary things? Perhaps there’s an evolutionary reason for this. If we assume there are frightening things in the dark, we become more cautious. Better safe than sorry. When I explained this to my son, I didn’t know that it was an existing theory with a name: the Free Energy Principle. According to the Free Energy Principle, our brain is fundamentally a prediction machine. Its most important function is to anticipate the future. The evolutionary reason for this is quite obvious: if you can predict a predator's attack, you can escape in time; if you can't, you die. Artem Kirsanov made an excellent video about the Free Energy Principle. For anyone interested in delving deeper into the topic, it’s highly recommended to watch it: https://www.youtube.com/watch?v=iPj9D9LgK2A&embedable=true According to the theory, every human brain contains an abstract model of reality (a world model) that it uses to make predictions. The brain continuously fine-tunes these predictions based on incoming stimuli from the environment, and this is how our perception of reality is constructed. Artem illustrates this with an example involving a tiger in his video. If we see yellow and black stripes in the distance, our brain can generate several predictions. It could be a person wearing a tiger-striped shirt, a toy tiger lost by a child, or even a real tiger. The brain, therefore, has multiple theories about how to interpret the sensory input. This "uncertainty" is referred to in the theory as free energy, and the brain's goal is to minimize this free energy—essentially reducing the chance of being surprised as much as possible. As more information comes in from the environment, the number of possible theories decreases. For example, if we see the tiger pattern moving, the toy tiger theory can be ruled out, leaving us with either a person or a real tiger. If we’re in a park, it’s most likely a person walking through the bushes. If we’re on a safari, however, it’s far more likely to be a real tiger. According to the Free Energy Principle, perceiving reality is not a passive process. The reality we experience is something we project rather than merely perceive. The perceived reality largely originates from within, while sensory input from the environment serves to fine-tune this internal concept. There’s an idea that has been on my mind for a long time: what would happen if we removed sensory input from the process and connected brains directly to each other? Imagine four people lying in isolation tanks, completely shutting out all external stimuli. Their brains are connected via advanced BCI (Brain-Computer Interface) systems. It’s far from certain that this could ever be achieved, and it’s obviously not trivial from a technological standpoint, but let’s entertain the thought experiment. At the moment of connection, without any external stimuli, the state would likely be quite chaotic (with a very high level of free energy). However, since each brain would be working to reduce this free energy, an equilibrium state would eventually be reached, and the four brains would synchronize. What would it feel like to experience this equilibrium state? We have experienced what it’s like when our brain generates "reality" without sensory input based on its world model. This happens when we dream. But what would it be like if four people were to dream the same dream simultaneously? Since four brains would be working simultaneously to minimize surprises (reduce free energy), this shared dream would likely feel much more real than traditional dreams—perhaps so real that it would be entirely indistinguishable from external reality. When we think about fully immersive virtual reality, we often assume it will require immense computational power, as simulating the physical world is an incredibly complex task. However, it’s quite possible that simply connecting people’s brains directly could be enough—the brain might handle the rest on its own. To create the true metaverse, we don’t need supercomputers capable of perfectly simulating the physical world—what we need are wires connecting brains. How human brains could be "wired together" is, of course, a very difficult question to answer. The most likely candidate for this would be some form of BCI (Brain-Computer Interface), such as Neuralink, where electrodes detect neural activity and stimulate neurons. Other possibilities include magnetic field-activated nanoparticles or light (I wrote an entire article on this topic a few years ago), but something entirely unconventional could also emerge—for instance, direct biological connections using specialized fungi. Imagine if people spent the majority of their lives in fully immersive VR—it could drastically reduce humanity's consumption and environmental impact. In a world where resource usage is minimized and existing resources are optimally distributed, there would be no scarcity, no inequality, wars would lose their purpose, and there would be no need to exploit nature. For those interested in exploring this topic further, you can read a full article about it here: https://hackernoon.com/why-brain-machine-interfaces-will-be-human-kinds-greatest-invention?embedable=true The question arises: if the fake reality created by connecting brains is completely indistinguishable from “real reality”, isn’t it possible that we’re already living in such a fake reality? Yes, yes—this brings us to the good old Simulation Hypothesis. Several studies have explored how we might prove if we are indeed living in a simulation. Without exception, these theories focus on identifying the limits of the computer running the simulation. They propose experiments designed to push the simulating computer into calculations it cannot perform. But what if there is no computer? What if the simulation runs directly on our brains? In such a case, it would be impossible to prove we are living in a simulation because no matter what experiment we conduct, we would always receive results that align with our expectations and confirm that our world is not a simulation. But if we are indeed living in a simulation, who created it? This is another common question related to the Simulation Hypothesis. However, there’s a very simple answer to this: perhaps we have always existed in this interconnected state, as a single mind, and the real illusion is that we are separate individuals. This idea is the focus of my recent article on the Thinking Universe. For those interested in exploring the topic further, I delve into this concept in detail there. The Free Energy Principle is a fascinating theory from the perspectives of neuroscience and artificial intelligence. However, the most exciting aspect is that it allows us to understand our brain’s functioning deeply enough to potentially "hack" it. Such a technology would surpass all others, as it wouldn’t just solve problems—it would enable us to create entirely new realities for ourselves where those problems don’t even exist! Once, my little son asked me why people fear the dark. I told him that our brains constantly dream, even when we’re awake. When awake, this dream is in sync with the outside world we perceive (what we see, hear, etc.). When it’s dark, visual stimuli disappear, and our brains start filling the void. And why does it fill it with scary things? Perhaps there’s an evolutionary reason for this. If we assume there are frightening things in the dark, we become more cautious. Better safe than sorry. When I explained this to my son, I didn’t know that it was an existing theory with a name: the Free Energy Principle . Free Energy Principle According to the Free Energy Principle, our brain is fundamentally a prediction machine. Its most important function is to anticipate the future. The evolutionary reason for this is quite obvious: if you can predict a predator's attack, you can escape in time; if you can't, you die. Artem Kirsanov made an excellent video about the Free Energy Principle. For anyone interested in delving deeper into the topic, it’s highly recommended to watch it: https://www.youtube.com/watch?v=iPj9D9LgK2A&embedable=true https://www.youtube.com/watch?v=iPj9D9LgK2A&embedable=true According to the theory, every human brain contains an abstract model of reality (a world model) that it uses to make predictions. The brain continuously fine-tunes these predictions based on incoming stimuli from the environment, and this is how our perception of reality is constructed. Artem illustrates this with an example involving a tiger in his video. If we see yellow and black stripes in the distance, our brain can generate several predictions. It could be a person wearing a tiger-striped shirt, a toy tiger lost by a child, or even a real tiger. The brain, therefore, has multiple theories about how to interpret the sensory input. This "uncertainty" is referred to in the theory as free energy, and the brain's goal is to minimize this free energy—essentially reducing the chance of being surprised as much as possible. As more information comes in from the environment, the number of possible theories decreases. For example, if we see the tiger pattern moving, the toy tiger theory can be ruled out, leaving us with either a person or a real tiger. If we’re in a park, it’s most likely a person walking through the bushes. If we’re on a safari, however, it’s far more likely to be a real tiger. According to the Free Energy Principle, perceiving reality is not a passive process. The reality we experience is something we project rather than merely perceive. The perceived reality largely originates from within, while sensory input from the environment serves to fine-tune this internal concept. There’s an idea that has been on my mind for a long time: what would happen if we removed sensory input from the process and connected brains directly to each other? Imagine four people lying in isolation tanks, completely shutting out all external stimuli. Their brains are connected via advanced BCI (Brain-Computer Interface) systems. It’s far from certain that this could ever be achieved, and it’s obviously not trivial from a technological standpoint, but let’s entertain the thought experiment. At the moment of connection, without any external stimuli, the state would likely be quite chaotic (with a very high level of free energy). However, since each brain would be working to reduce this free energy, an equilibrium state would eventually be reached, and the four brains would synchronize. What would it feel like to experience this equilibrium state? We have experienced what it’s like when our brain generates "reality" without sensory input based on its world model. This happens when we dream. But what would it be like if four people were to dream the same dream simultaneously? Since four brains would be working simultaneously to minimize surprises (reduce free energy), this shared dream would likely feel much more real than traditional dreams—perhaps so real that it would be entirely indistinguishable from external reality. When we think about fully immersive virtual reality, we often assume it will require immense computational power, as simulating the physical world is an incredibly complex task. However, it’s quite possible that simply connecting people’s brains directly could be enough—the brain might handle the rest on its own. To create the true metaverse, we don’t need supercomputers capable of perfectly simulating the physical world—what we need are wires connecting brains. To create the true metaverse, we don’t need supercomputers capable of perfectly simulating the physical world—what we need are wires connecting brains. To create the true metaverse, we don’t need supercomputers capable of perfectly simulating the physical world—what we need are wires connecting brains. How human brains could be "wired together" is, of course, a very difficult question to answer. The most likely candidate for this would be some form of BCI (Brain-Computer Interface), such as Neuralink, where electrodes detect neural activity and stimulate neurons. Other possibilities include magnetic field-activated nanoparticles or light (I wrote an entire article on this topic a few years ago), but something entirely unconventional could also emerge—for instance, direct biological connections using specialized fungi. entire article Imagine if people spent the majority of their lives in fully immersive VR—it could drastically reduce humanity's consumption and environmental impact. In a world where resource usage is minimized and existing resources are optimally distributed, there would be no scarcity, no inequality, wars would lose their purpose, and there would be no need to exploit nature. For those interested in exploring this topic further, you can read a full article about it here: https://hackernoon.com/why-brain-machine-interfaces-will-be-human-kinds-greatest-invention?embedable=true https://hackernoon.com/why-brain-machine-interfaces-will-be-human-kinds-greatest-invention?embedable=true The question arises: if the fake reality created by connecting brains is completely indistinguishable from “real reality”, isn’t it possible that we’re already living in such a fake reality? Yes, yes—this brings us to the good old Simulation Hypothesis. Several studies have explored how we might prove if we are indeed living in a simulation. Without exception, these theories focus on identifying the limits of the computer running the simulation. They propose experiments designed to push the simulating computer into calculations it cannot perform. But what if there is no computer? What if the simulation runs directly on our brains? In such a case, it would be impossible to prove we are living in a simulation because no matter what experiment we conduct, we would always receive results that align with our expectations and confirm that our world is not a simulation. But if we are indeed living in a simulation, who created it? This is another common question related to the Simulation Hypothesis. However, there’s a very simple answer to this: perhaps we have always existed in this interconnected state, as a single mind, and the real illusion is that we are separate individuals. This idea is the focus of my recent article on the Thinking Universe . For those interested in exploring the topic further, I delve into this concept in detail there. Thinking Universe The Free Energy Principle is a fascinating theory from the perspectives of neuroscience and artificial intelligence. However, the most exciting aspect is that it allows us to understand our brain’s functioning deeply enough to potentially "hack" it. Such a technology would surpass all others, as it wouldn’t just solve problems—it would enable us to create entirely new realities for ourselves where those problems don’t even exist! Such a technology would surpass all others, as it wouldn’t just solve problems—it would enable us to create entirely new realities for ourselves where those problems don’t even exist!

The Free Energy Principle and the Simulation Hypothesis

AI’s Boom vs. Energy Crisis Community: What the 2030 Apocalypse Could Look Like

New York City plans to support electric vehicle (EV) adoption with a $15 million initiative to install 600 new curbside chargers. Can the city’s current grid infrastructure keep pace? New York City is ramping up its efforts to support electric vehicle (EV) adoption as it plans to install 600 new curbside EV chargers. Mayor Eric Adams recently announced that the initiative is part of a $15 million federal grant being awarded to the city to bolster its EV charging infrastructure — with a focus on underserved and low-income communities. “We’re building a cleaner, greener, and more affordable city for all New Yorkers, and through initiatives like this, we’re making sure that no New Yorkers are left behind,” wrote Mayor Adams in a blog post. “This $15 million grant will let us build the nation’s largest EV charging program, focused on low-income and disadvantaged areas so people living in all five boroughs can afford to drive and charge electric vehicles.” By 2050, New York City aims to get 20% of all new vehicles registered as electric. Thanks to incentives like the Inflation Reduction Act tax credits and Drive Clean rebates, a growing number of NYC drivers are switching to EVs. Despite these efforts, the city's grid is already struggling to keep up with the power and infrastructural requirements of current EV owners. Can NYC’s Electrical Grid Keep Up with the Growing Demand for EV Charging? NYC is currently grappling with two key challenges to scale its charging infrastructure — space and available power. For commercial transportation, charging hubs require significant space for infrastructure and larger electric trucks. In dense urban areas, especially NYC, real estate is limited and expensive, making it difficult to find suitable locations to accommodate charging stations and vehicles without disrupting city traffic. “Most of the federal EV infrastructure grants are for chargers deployed along highways (not buildings), limiting monetary incentives available to building developers deploying EV chargers,” Akshay Singh, Industrial and Automotive Industries Principal at PwC Strategy&, told me. “City planners need to work closely with utilities to make sure upgrades are being planned & performed timely, especially to supply dense urban areas with the highest EV adoption.” New York’s grid is already strained as the city attempts to electrify not only its transportation but also its buildings. Singh asserts that developers often do not understand the level of grid upgrades required due to a lack of available information. Moreover, these challenges are expected to become more prevalent with charging sites located in high-rise buildings that have not had capacity upgrades in years. “Major challenges with EV charging today are around how to make the economics of operating EV chargers feasible. DC fast chargers that operate at 100kW+ power output (to charge EVs within 30-45 mins) can cost upwards of $50k per unit which makes upfront costs of a multi-unit charging site very expensive,” added Singh. Electric Vehicles as Untapped Grid Assets? As New York pushes toward its ambitious goal of having 400,000 EVs on the road by 2030 as part of the "Green Rides" initiative, it’s clear that the city needs more fast-charging sites. Conglomerates like Revel are already making strides with its "superhub" in Brooklyn, boasting 25 high-speed chargers. But even with their success, the company faces a significant challenge: finding enough grid power to keep these large fast-charging stations running smoothly. While the city grapples with its grid limitations, some startups are stepping in to offer innovative solutions through A.I. and edge computing. Brooklyn-based Synop, for instance, aims to help stabilize the grid by integrating data from grid operators, EV charging stations, and EV fleet operators. “We focus on edge computing to enhance charging management,” Gagan Dhillon, CEO of Synop, told me. By bringing cloud-based functionality closer to the charging site, the company says it can increase reliability and reduce latency, “which is critical in areas with constrained grids or complex microgrid setups.” Dhillon added that one overlooked aspect of EV adoption is that electric vehicles are essentially batteries on wheels, and with the right infrastructure, they can serve as valuable assets rather than just power consumers. “EVs can act as backup power sources during times of high grid demand, providing energy back to the grid or even powering homes when needed,” he said. “With the integration of smart charging technology, EVs can be programmed to charge during off-peak times when energy is more abundant. This flexibility makes them less of a strain on the grid than often imagined.” With more people driving EVs and a grid under a lot of strain, NYC's EV goal success will hinge on how open it is to new technologies, speeding up infrastructure projects, and investing in sustainable energy solutions. “There will be a need for higher powered EV chargers that will further increase the grid upgrade requirements. Developers need to work together with local utilities from the start to successfully deploy EV projects,” Singh told me. “Along with that, additional [capital expenditure] needed to sustain the grid for more demand needs to be laid out and discussions on who will bear the cost burden.” New York City plans to support electric vehicle (EV) adoption with a $15 million initiative to install 600 new curbside chargers. Can the city’s current grid infrastructure keep pace? New York City plans to support electric vehicle (EV) adoption with a $15 million initiative to install 600 new curbside chargers. Can the city’s current grid infrastructure keep pace? New York City plans to support electric vehicle (EV) adoption with a $15 million initiative to install 600 new curbside chargers. Can the city’s current grid infrastructure keep pace? New York City is ramping up its efforts to support electric vehicle (EV) adoption as it plans to install 600 new curbside EV chargers. Mayor Eric Adams recently announced that the initiative is part of a $15 million federal grant being awarded to the city to bolster its EV charging infrastructure — with a focus on underserved and low-income communities. “We’re building a cleaner, greener, and more affordable city for all New Yorkers, and through initiatives like this, we’re making sure that no New Yorkers are left behind,” wrote Mayor Adams in a blog post . “This $15 million grant will let us build the nation’s largest EV charging program, focused on low-income and disadvantaged areas so people living in all five boroughs can afford to drive and charge electric vehicles.” By 2050, New York City aims to get 20% of all new vehicles registered as electric. Thanks to incentives like the Inflation Reduction Act tax credits and Drive Clean rebates, a growing number of NYC drivers are switching to EVs. Despite these efforts, the city's grid is already struggling to keep up with the power and infrastructural requirements of current EV owners. blog post blog post Inflation Reduction Act Inflation Reduction Act switching switching Can NYC’s Electrical Grid Keep Up with the Growing Demand for EV Charging? NYC is currently grappling with two key challenges to scale its charging infrastructure — space and available power. For commercial transportation, charging hubs require significant space for infrastructure and larger electric trucks. In dense urban areas, especially NYC, real estate is limited and expensive, making it difficult to find suitable locations to accommodate charging stations and vehicles without disrupting city traffic. “Most of the federal EV infrastructure grants are for chargers deployed along highways (not buildings), limiting monetary incentives available to building developers deploying EV chargers,” Akshay Singh , Industrial and Automotive Industries Principal at PwC Strategy&, told me. “City planners need to work closely with utilities to make sure upgrades are being planned & performed timely, especially to supply dense urban areas with the highest EV adoption.” Akshay Singh Akshay Singh New York’s grid is already strained as the city attempts to electrify not only its transportation but also its buildings. Singh asserts that developers often do not understand the level of grid upgrades required due to a lack of available information. Moreover, these challenges are expected to become more prevalent with charging sites located in high-rise buildings that have not had capacity upgrades in years. “Major challenges with EV charging today are around how to make the economics of operating EV chargers feasible. DC fast chargers that operate at 100kW+ power output (to charge EVs within 30-45 mins) can cost upwards of $50k per unit which makes upfront costs of a multi-unit charging site very expensive,” added Singh. Electric Vehicles as Untapped Grid Assets? As New York pushes toward its ambitious goal of having 400,000 EVs on the road by 2030 as part of the " Green Rides " initiative, it’s clear that the city needs more fast-charging sites. Conglomerates like Revel are already making strides with its " superhub " in Brooklyn, boasting 25 high-speed chargers. But even with their success, the company faces a significant challenge: finding enough grid power to keep these large fast-charging stations running smoothly. Green Rides Green Rides superhub superhub While the city grapples with its grid limitations, some startups are stepping in to offer innovative solutions through A.I. and edge computing. Brooklyn-based Synop, for instance, aims to help stabilize the grid by integrating data from grid operators, EV charging stations, and EV fleet operators. “We focus on edge computing to enhance charging management,” Gagan Dhillon , CEO of Synop, told me. By bringing cloud-based functionality closer to the charging site, the company says it can increase reliability and reduce latency, “which is critical in areas with constrained grids or complex microgrid setups.” Gagan Dhillon Gagan Dhillon Dhillon added that one overlooked aspect of EV adoption is that electric vehicles are essentially batteries on wheels, and with the right infrastructure, they can serve as valuable assets rather than just power consumers. “EVs can act as backup power sources during times of high grid demand, providing energy back to the grid or even powering homes when needed,” he said. “With the integration of smart charging technology, EVs can be programmed to charge during off-peak times when energy is more abundant. This flexibility makes them less of a strain on the grid than often imagined.” With more people driving EVs and a grid under a lot of strain, NYC's EV goal success will hinge on how open it is to new technologies, speeding up infrastructure projects, and investing in sustainable energy solutions. “There will be a need for higher powered EV chargers that will further increase the grid upgrade requirements. Developers need to work together with local utilities from the start to successfully deploy EV projects,” Singh told me. “Along with that, additional [capital expenditure] needed to sustain the grid for more demand needs to be laid out and discussions on who will bear the cost burden.”

NYC's $15 Million EV Charger Initiative: Can the City Grid Handle Rising Energy Demands?

It’s no secret we've stumbled upon a formidable obstacle to our AI-powered future: the staggering energy consumption of our current models. Industry leaders are scrambling to come up with short-term answers to ensure they don’t miss the wave, with ambitious efforts like Microsoft’s reopening of nuclear reactors on Three Mile Island and Google working on “first-of-its-kind” geothermal projects. And while that plays out at the big kid’s table, a flock of new startups are building on the progress made in recent years, re-thinking the fundamentals to see if there are solutions that could serve as the long-term solution. One that doesn’t require hundreds of millions of dollars in infrastructure investment. AI’s Power Demands Are a Ticking Time Bomb Operating large language models in their current iteration is an energy-intensive process that's rapidly approaching unsustainable levels. Training a single AI model can emit as much carbon as five cars over their entire lifetimes. It's not just an environmental concern; it's a scalability nightmare threatening to derail the AI revolution before it fully takes flight. Consider these sobering facts: GPT-3, with its 175 billion parameters, reportedly required 1,287 MWh for a single training run.
The carbon footprint of training a large NLP model is estimated to be around 626,000 pounds of CO2 equivalent.
As models grow, so does their energy consumption – often at a superlinear rate. And as the industry pushes for more advanced AI capabilities, this energy consumption is set to skyrocket. This is not only a problem on an operational level but also in the greater picture, as industry leaders like Google have pledged to achieve net-zero carbon emissions through the purchase of carbon credits from firms doing things like plugging orphaned oil and gas wells - a market where demand is already far out-pacing supply. Efficiency & Minimal Parameter Activation: The New North Star of AI Innovation The solution may be rather simple: smarter, smaller, more efficient models built for a set of specific purposes. Narrowing the scope, if you will. One such example is the open-source Aria model built by Rhymes, which employs minimal parameter activation. While the Aria model boasts a total of 25.3 billion parameters, it only activates a mere 3.9 billion for any given task. Traditional, mainstream models like GPT-3 activate all of their parameters for every task, regardless of complexity, while Aria's approach is like a surgeon using only the necessary instruments for a specific procedure. Many surgeons would tell you they don’t need to deploy the entire operating room's worth of equipment for every operation. Rhyme has deployed this in practice with BeaGo, which they call “a smarter, faster AI search.” Based on my tests, BeaGo’s results were indistinguishable from competing products from Perplexity and other, most energy and time-intensive products. But it’s more than simply narrowing the scope: the startup has built a multimodal open-source mixture-of-experts model that sorts and intelligently manages large, long-context data of all types, including text, video, and images. Rhymes’ solution may be lighting the pathway for AI in 2025 and beyond, all without the hundreds of millions of dollars in infrastructure spending. The Reminder: Innovation Never Stops In the end, the work of companies like Rhymes serves as a reminder that just because we found something that works, doesn’t mean the task of innovating is over. While Microsoft and Google run with our existing large language models, working to productize and bring AI to the mass-market, others cannot stop working on building something even better. I’m heartened by the startup-driven approach that I see here at the end of 2024 – combining multimodal capabilities, selective parameter activation, and open-source collaboration – offers a blueprint for how we can achieve a vision of an AI that both works and works for the planet. It’s no secret we've stumbled upon a formidable obstacle to our AI-powered future: the staggering energy consumption of our current models. Industry leaders are scrambling to come up with short-term answers to ensure they don’t miss the wave, with ambitious efforts like Microsoft’s reopening of nuclear reactors on Three Mile Island and Google working on “first-of-its-kind” geothermal projects . Microsoft’s reopening of nuclear reactors on Three Mile Island Microsoft’s reopening of nuclear reactors on Three Mile Island Google working on “first-of-its-kind” geothermal projects Google working on “first-of-its-kind” geothermal projects And while that plays out at the big kid’s table, a flock of new startups are building on the progress made in recent years, re-thinking the fundamentals to see if there are solutions that could serve as the long-term solution. One that doesn’t require hundreds of millions of dollars in infrastructure investment. doesn’t AI’s Power Demands Are a Ticking Time Bomb Operating large language models in their current iteration is an energy-intensive process that's rapidly approaching unsustainable levels. Training a single AI model can emit as much carbon as five cars over their entire lifetimes. It's not just an environmental concern; it's a scalability nightmare threatening to derail the AI revolution before it fully takes flight. Consider these sobering facts: GPT-3, with its 175 billion parameters, reportedly required 1,287 MWh for a single training run. The carbon footprint of training a large NLP model is estimated to be around 626,000 pounds of CO2 equivalent. As models grow, so does their energy consumption – often at a superlinear rate. GPT-3, with its 175 billion parameters, reportedly required 1,287 MWh for a single training run. The carbon footprint of training a large NLP model is estimated to be around 626,000 pounds of CO2 equivalent. As models grow, so does their energy consumption – often at a superlinear rate. And as the industry pushes for more advanced AI capabilities, this energy consumption is set to skyrocket. This is not only a problem on an operational level but also in the greater picture, as industry leaders like Google have pledged to achieve net-zero carbon emissions through the purchase of carbon credits from firms doing things like plugging orphaned oil and gas wells - a market where demand is already far out-pacing supply. plugging orphaned oil and gas wells Efficiency & Minimal Parameter Activation: The New North Star of AI Innovation The solution may be rather simple: smarter, smaller, more efficient models built for a set of specific purposes. Narrowing the scope, if you will. One such example is the open-source Aria model built by Rhymes , which employs minimal parameter activation. While the Aria model boasts a total of 25.3 billion parameters, it only activates a mere 3.9 billion for any given task. Traditional, mainstream models like GPT-3 activate all of their parameters for every task, regardless of complexity, while Aria's approach is like a surgeon using only the necessary instruments for a specific procedure. Many surgeons would tell you they don’t need to deploy the entire operating room's worth of equipment for every operation. Aria model built by Rhymes Rhyme has deployed this in practice with BeaGo, which they call “a smarter, faster AI search.” Based on my tests, BeaGo’s results were indistinguishable from competing products from Perplexity and other, most energy and time-intensive products. But it’s more than simply narrowing the scope: the startup has built a multimodal open-source mixture-of-experts model that sorts and intelligently manages large, long-context data of all types, including text, video, and images. Rhymes’ solution may be lighting the pathway for AI in 2025 and beyond, all without the hundreds of millions of dollars in infrastructure spending. The Reminder: Innovation Never Stops In the end, the work of companies like Rhymes serves as a reminder that just because we found something that works , doesn’t mean the task of innovating is over. While Microsoft and Google run with our existing large language models, working to productize and bring AI to the mass-market, others cannot stop working on building something even better. works I’m heartened by the startup-driven approach that I see here at the end of 2024 – combining multimodal capabilities, selective parameter activation, and open-source collaboration – offers a blueprint for how we can achieve a vision of an AI that both works and works for the planet.

The AI Energy Crisis & A Newfound Push for Efficiency

Die KI-Energiekrise und ein neuer Drang nach Effizienz

La crise énergétique de l'IA et une nouvelle dynamique en faveur de l'efficacité

משבר האנרגיה של AI ודחיפה חדשה ליעילות

La crisi energetica dell’intelligenza artificiale e una nuova spinta verso l’efficienza

De AI-energiecrisis en een nieuwe drang naar efficiëntie

A crise energética da IA e um novo impulso para a eficiência

It’s a real shame that I couldn’t get a ticket to one of the landmark events that solidified London’s status as a hub of concentrated creative energy. This year, those who harness this energy – the ones who still burn with the desire to make the world a better place – used it to raise public awareness on a wide range of issues. At the London Design Festival, themes of synergy between nature and humanity took center stage, embodied in the work of Marjan van Aubel. Hot on the heels of her installation, the industrial design team at Diez Office encouraged an unspoken dialogue about ecological responsibility with their project Vert. In general, the program featured numerous innovative solutions, breaking down the boundaries of stereotypical thinking. And we crave that kind of intellectual stimulation! In this piece, I’ll talk about the playful, mesmerizing, light-hearted and literally warming design on display this year! London and design go hand in hand. It is part of our story. London Design Festival is a platform for hundreds of design stories to be told. Each of them speaks to an expanding audience hungry for design ideas and enjoying the quality and diversity of what’s on offer. It all confirms London’s status as the global center of design – says Ben Evans CBE. It’s hard to disagree with the words of this passionate advocate and activist for all-encompassing creativity, which can be found in everything that surrounds us. Today, we’ll spot it in glass, but in a broader sense, we’ll explore the role materials play in design – and in parallels. Parallels not only between the red oak beams of the exhibit at Chelsea College of Arts but also between the architecture of play and real life. The Sun, My Heart The energy within us has the power to create incredible things, especially when we learn to recognize it and channel it in the right direction – one where we create something extraordinary that captivates the world. Energy is the central theme in the work of Dutch designer Marjan van Aubel. Her creations can hardly be described as merely "sunny" although, in reality, each one of them is filled with light. Without this element, we can no longer imagine her work. Energy, as an aesthetic element, has become an integral part of every object crafted by her hands: from glassware, which reflects the flow of matter, to entire installations. We, in turn, have become vessels filled with her revelation: They were transparent and colored. It was the first time I saw that solar energy could be beautiful. From that moment, I started implementing it into different objects: glassware, tables, light installations, buildings. Now, I see all surfaces as a potential to harvest energy Speaking of installations, we can now explore her piece “The Sun, My Heart” a highlight of the LDF collection. The title becomes somewhat rhetorical when you consider van Aubel’s relationship with the primary source of life, with roots that delve deep into Buddhist philosophy. In the language of Vietnamese monks, it calls on us to "open our hearts to new connections". These connections come to life through 77 self-charging Sunne lamps (some of which represent nothing more than rays of light), symbolizing the complete solar cycle, accompanied by harmoniously selected sounds that echo the sun's vibrations. Working in unison, this masterpiece immerses the viewer in a state of total relaxation, stimulating the brain's theta waves. We receive the sun in such abundance. It’s obvious that we need to transition to solar energy, but the current approach is too technical. We shouldn’t only talk about numbers and efficiency. The sun is something holistic; we can feel it. We need beauty and design to make us want to change things. We need to feel it inside – the designer shares. Pavilions of Wonder A fresh perspective with a historical undertone, embodying cultural heritage – the true American dream, brought to life in a trilogy of works by designer Nina Tolstrup. This project has become a symbol of vivid imagination and aesthetics, driven by a spirit of fun and playfulness. The installation consists of three objects, harking back to the time when Palm Springs became the true capital of modernism in the USA, owing to its ability to "grow rapidly" with affordable housing after World War II. This area, dotted with buildings of the corresponding architectural style, would later become a prototype of inspiration for future designers. Among those designers is Nina Tolstrup, who infused the project with the same playful spirit reflected in Barbie’s DreamHouses. It is through her vibrant qualities, captured like film on a reel, that we recognize the essence of the designer. We looked at the evolution of the DreamHouse, informed by Greater Palm Springs architecture. We examined the visual language and iconic elements common to both Vert And here are the very same "parallels" made from red oak, presented in the collaborative work of the American Hardwood Export Council (AHEC), the Munich-based Diez Office, and Frankfurt's Office for Microclimatic Cultivation (OMC°C). This project is far from a cry for help, but rather a message of mindful care for the environment. Vert, which signifies something "green" is an undeniably bold statement, offering a tangible and effective solution to a global issue. It represents a solution through a small, carefully planned, and meticulously thought-out noble endeavor. It can't be achieved just by planting trees or through individual façade and rooftop greening projects, so additional solutions are urgently needed. We need fairly large structures to generate the necessary amount of shade and biomass – explains Nicola Stattmann, co-founder of OMC°C. The architectural installation consists of several triangular structures, each standing 10 meters tall. These are all covered with plants such as morning glory, cheese plant, black-eyed Susan, and night-blooming moonflower, which will act as "killers" of carbon dioxide released into the atmosphere. Interestingly, the modularity of the designed structure allows for Vert to be produced in series in the future. This once again highlights the brilliance of the minds behind the project, brought to life with the support of AHEC, who have long been advocating for the use of durable engineered materials. Red oak glulam's natural beauty, combined with its structural capabilities, makes it a promising choice both functionally and aesthetically – says Stefan Diez, Diez Office founder. The London Design Festival 2024 masterfully demonstrated the power of design in tackling global issues – from sustainable energy solutions to supporting biodiversity and reimagining urban spaces. Through innovative projects, the festival highlighted the crucial role of creativity in shaping a more sustainable and harmonious future. As these projects blur the lines between functionality and beauty, they invite us to envision a world where technology, nature, and human ingenuity coexist seamlessly, encouraging a more thoughtful and responsible approach to life. The London Design Festival week was one of the most anticipated events for us. Our fields are closely intertwined – that much is clear. We keep track of the latest trends to make our 3D visualizations even more realistic, filled with the smallest details that reflect the quality of bringing the 'imagined' to life. We’re interested in everything: from the materials often used in designers' projects to the very concepts behind these works, because this knowledge embodies our commitment to staying current. We don't want to fall behind. We aim to build trust with every client interaction, ensuring that the result meets and exceeds their needs, leaving them fully satisfied – says Alexander Kasperovich, CEO of 3D visualization and rendering studio, Studio57. It’s a real shame that I couldn’t get a ticket to one of the landmark events that solidified London’s status as a hub of concentrated creative energy. This year, those who harness this energy – the ones who still burn with the desire to make the world a better place – used it to raise public awareness on a wide range of issues. It’s a real shame that I couldn’t get a ticket to one of the landmark events that solidified London’s status as a hub of concentrated creative energy. This year, those who harness this energy – the ones who still burn with the desire to make the world a better place – used it to raise public awareness on a wide range of issues. At the London Design Festival , themes of synergy between nature and humanity took center stage, embodied in the work of Marjan van Aubel. Hot on the heels of her installation, the industrial design team at Diez Office encouraged an unspoken dialogue about ecological responsibility with their project Vert. In general, the program featured numerous innovative solutions, breaking down the boundaries of stereotypical thinking. And we crave that kind of intellectual stimulation! In this piece, I’ll talk about the playful, mesmerizing, light-hearted and literally warming design on display this year! London Design Festival London Design Festival London and design go hand in hand. It is part of our story. London Design Festival is a platform for hundreds of design stories to be told. Each of them speaks to an expanding audience hungry for design ideas and enjoying the quality and diversity of what’s on offer. It all confirms London’s status as the global center of design London and design go hand in hand. It is part of our story. London Design Festival is a platform for hundreds of design stories to be told. Each of them speaks to an expanding audience hungry for design ideas and enjoying the quality and diversity of what’s on offer. It all confirms London’s status as the global center of design – says Ben Evans CBE. – says Ben Evans CBE. Ben Evans It’s hard to disagree with the words of this passionate advocate and activist for all-encompassing creativity, which can be found in everything that surrounds us. Today, we’ll spot it in glass, but in a broader sense, we’ll explore the role materials play in design – and in parallels. Parallels not only between the red oak beams of the exhibit at Chelsea College of Arts but also between the architecture of play and real life. Chelsea College of Arts The Sun, My Heart The Sun, My Heart The energy within us has the power to create incredible things, especially when we learn to recognize it and channel it in the right direction – one where we create something extraordinary that captivates the world. Energy is the central theme in the work of Dutch designer Marjan van Aubel . Her creations can hardly be described as merely "sunny" although, in reality, each one of them is filled with light. Without this element, we can no longer imagine her work. Marjan van Aubel Marjan van Aubel Energy, as an aesthetic element, has become an integral part of every object crafted by her hands: from glassware, which reflects the flow of matter, to entire installations. We, in turn, have become vessels filled with her revelation: They were transparent and colored. It was the first time I saw that solar energy could be beautiful. From that moment, I started implementing it into different objects: glassware, tables, light installations, buildings. Now, I see all surfaces as a potential to harvest energy They were transparent and colored. It was the first time I saw that solar energy could be beautiful. From that moment, I started implementing it into different objects: glassware, tables, light installations, buildings. Now, I see all surfaces as a potential to harvest energy Speaking of installations, we can now explore her piece “The Sun, My Heart” a highlight of the LDF collection. The title becomes somewhat rhetorical when you consider van Aubel’s relationship with the primary source of life, with roots that delve deep into Buddhist philosophy. In the language of Vietnamese monks, it calls on us to "open our hearts to new connections". These connections come to life through 77 self-charging Sunne lamps (some of which represent nothing more than rays of light), symbolizing the complete solar cycle, accompanied by harmoniously selected sounds that echo the sun's vibrations. Working in unison, this masterpiece immerses the viewer in a state of total relaxation, stimulating the brain's theta waves. We receive the sun in such abundance. It’s obvious that we need to transition to solar energy, but the current approach is too technical. We shouldn’t only talk about numbers and efficiency. The sun is something holistic; we can feel it. We need beauty and design to make us want to change things. We need to feel it inside We receive the sun in such abundance. It’s obvious that we need to transition to solar energy, but the current approach is too technical. We shouldn’t only talk about numbers and efficiency. The sun is something holistic; we can feel it. We need beauty and design to make us want to change things. We need to feel it inside – the designer shares. – the designer shares. Pavilions of Wonder Pavilions of Wonder A fresh perspective with a historical undertone, embodying cultural heritage – the true American dream, brought to life in a trilogy of works by designer Nina Tolstrup. This project has become a symbol of vivid imagination and aesthetics, driven by a spirit of fun and playfulness. The installation consists of three objects, harking back to the time when Palm Springs became the true capital of modernism in the USA, owing to its ability to "grow rapidly" with affordable housing after World War II. This area, dotted with buildings of the corresponding architectural style, would later become a prototype of inspiration for future designers. Among those designers is Nina Tolstrup , who infused the project with the same playful spirit reflected in Barbie’s DreamHouses . It is through her vibrant qualities, captured like film on a reel, that we recognize the essence of the designer. Nina Tolstrup Barbie’s DreamHouses Barbie’s DreamHouses We looked at the evolution of the DreamHouse, informed by Greater Palm Springs architecture. We examined the visual language and iconic elements common to both We looked at the evolution of the DreamHouse, informed by Greater Palm Springs architecture. We examined the visual language and iconic elements common to both Vert Vert And here are the very same "parallels" made from red oak, presented in the collaborative work of the American Hardwood Export Council (AHEC), the Munich-based Diez Office, and Frankfurt's Office for Microclimatic Cultivation (OMC°C). This project is far from a cry for help, but rather a message of mindful care for the environment. Vert, which signifies something "green" is an undeniably bold statement, offering a tangible and effective solution to a global issue. It represents a solution through a small, carefully planned, and meticulously thought-out noble endeavor. It can't be achieved just by planting trees or through individual façade and rooftop greening projects, so additional solutions are urgently needed. We need fairly large structures to generate the necessary amount of shade and biomass It can't be achieved just by planting trees or through individual façade and rooftop greening projects, so additional solutions are urgently needed. We need fairly large structures to generate the necessary amount of shade and biomass – explains Nicola Stattmann, co-founder of OMC°C. – explains Nicola Stattmann, co-founder of OMC°C. The architectural installation consists of several triangular structures, each standing 10 meters tall. These are all covered with plants such as morning glory, cheese plant, black-eyed Susan, and night-blooming moonflower, which will act as "killers" of carbon dioxide released into the atmosphere. Interestingly, the modularity of the designed structure allows for Vert to be produced in series in the future. This once again highlights the brilliance of the minds behind the project, brought to life with the support of AHEC, who have long been advocating for the use of durable engineered materials. Red oak glulam's natural beauty, combined with its structural capabilities, makes it a promising choice both functionally and aesthetically Red oak glulam's natural beauty, combined with its structural capabilities, makes it a promising choice both functionally and aesthetically – says Stefan Diez, Diez Office founder. – says Stefan Diez , Diez Office founder. Stefan Diez Stefan Diez The London Design Festival 2024 masterfully demonstrated the power of design in tackling global issues – from sustainable energy solutions to supporting biodiversity and reimagining urban spaces. Through innovative projects, the festival highlighted the crucial role of creativity in shaping a more sustainable and harmonious future. As these projects blur the lines between functionality and beauty, they invite us to envision a world where technology, nature, and human ingenuity coexist seamlessly, encouraging a more thoughtful and responsible approach to life. The London Design Festival week was one of the most anticipated events for us. Our fields are closely intertwined – that much is clear. We keep track of the latest trends to make our 3D visualizations even more realistic, filled with the smallest details that reflect the quality of bringing the 'imagined' to life. We’re interested in everything: from the materials often used in designers' projects to the very concepts behind these works, because this knowledge embodies our commitment to staying current. We don't want to fall behind. We aim to build trust with every client interaction, ensuring that the result meets and exceeds their needs, leaving them fully satisfied The London Design Festival week was one of the most anticipated events for us. Our fields are closely intertwined – that much is clear. We keep track of the latest trends to make our 3D visualizations even more realistic, filled with the smallest details that reflect the quality of bringing the 'imagined' to life. We’re interested in everything: from the materials often used in designers' projects to the very concepts behind these works, because this knowledge embodies our commitment to staying current. We don't want to fall behind. We aim to build trust with every client interaction, ensuring that the result meets and exceeds their needs, leaving them fully satisfied – says Alexander Kasperovich, CEO of 3D visualization and rendering studio, Studio57. – says Alexander Kasperovich, CEO of 3D visualization and rendering studio, Studio57.

Stories, News & Mentions about ENTERGY

Entergy’s Stories on HackerNoon

Entergy's News & Mentions