Discover Daily by Perplexity

IRS Acquiring Nvidia Supercomputer, ChatGPT Energy Use Overestimated, and Complex Brains Evolved Independently

Share episode

Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.

0:00 | 9:36

We're experimenting and would love to hear from you!

In this episode of 'Discover Daily', the IRS is making waves in government tech adoption with its upcoming acquisition of an Nvidia SuperPod AI supercomputer. This sophisticated system, featuring thirty-one servers with Blackwell processors, will be installed at their Martinsburg, West Virginia computing center. The system aims to revolutionize the agency's machine learning capabilities, particularly in fraud detection and taxpayer behavior analysis, marking a significant step forward in AI integration within government operations.

In a recent study on AI energy consumption, ChatGPT's latest model GPT-4 has been found to use significantly less power than previously thought. Research by Epoch AI reveals that each query consumes approximately 0.3 watt-hours, ten times lower than earlier estimates. This efficiency improvement is attributed to advanced hardware implementation, system optimization, and more accurate calculation methods, though the cumulative energy impact remains substantial given millions of daily queries.

Lastly, research published in Science has unveiled how birds and mammals have independently evolved complex brains through distinct pathways while achieving similar cognitive abilities. The study, led by Dr. Fernando García-Moreno, used advanced spatial transcriptomics and mathematical modeling to reveal fundamental differences in brain development between species. Despite using different genetic tools and developmental processes, both groups have achieved remarkable cognitive capabilities, challenging our traditional understanding of brain evolution and intelligence.

From Perplexity's Discover Feed: 

https://www.perplexity.ai/page/irs-acquiring-nvidia-supercomp-bQ9BUXJuSLOUpHZdCJ.vEA 

https://www.perplexity.ai/page/chatgpt-energy-use-overestimat-cn02azRBR2._eM_sH2n_Pw 

https://www.perplexity.ai/page/complex-brains-evolved-indepen-uPS546tuRJaWwidjGvrNCQ

**Introducing Perplexity Deep Research:**

https://www.perplexity.ai/hub/blog/introducing-perplexity-deep-research 

Perplexity is the fastest and most powerful way to search the web. Perplexity crawls the web and curates the most relevant and up-to-date sources (from academic papers to Reddit threads) to create the perfect response to any question or topic you’re interested in.

Take the world's knowledge with you anywhere.  Available on iOS and Android

Join our growing Discord community for the latest updates and exclusive content.

Follow us on:

Speaker 1

Welcome to Discover Daily by Perplexity, an AI-generated show on tech, science and culture. I'm Alex. Today we're exploring how birds and mammals have independently evolved complex brains, challenging our understanding of cognitive evolution. But first let's look at what else is happening across the world of technology and AI. Our first story today, the Internal Revenue Service is set to acquire an NVIDIA SuperPod AI supercomputer. This powerful system combines 31 servers with advanced Blackwell processors, each containing eight processing units. It's designed to enhance the IRS's machine learning capabilities for tasks such as fraud detection and taxpayer behavior analysis. The SuperPod will be installed at the IRS Computing Center in Martinsburg, west Virginia. While the exact cost isn't public yet, these systems typically start at $7 million. The acquisition is part of a broader AI transformation within the agency, which currently has 68 different AI-related projects in development. While this system may be smaller than the massive AI training centers used by tech giants like OpenAI and Meta, it represents a significant computing resource for a government agency. This acquisition marks a notable technological advancement for the IRS. The Research, applied Analytics and Statistics Division will operate the SuperPOD as part of their Compliance Data Warehouse project. The system will come with a substantial memory upgrade from NVIDIA enabling more sophisticated data analysis. Previous machine learning applications by this division have already shown promise in automated fraud detection and identity theft prevention. While the hardware hasn't been purchased or installed yet, the SuperPOD's advanced AI capabilities are expected to enhance the agency's ability to identify complex patterns indicative of fraudulent activities. This technology investment reflects a growing trend of AI adoption in government operations, particularly in areas requiring sophisticated pattern recognition and data analysis.

Speaker 1

Moving on to our second story, a new study has revealed that ChatGPT's energy consumption is much lower than previously estimated. The non-profit research institute EPIC-AI found that ChatGPT's latest model, gpt-4-0, consumes approximately 0.3 watt-hours per query ten times less than earlier estimates. But it's worth noting that these figures are approximations, as OpenAI hasn't publicly disclosed detailed energy consumption data. This significant reduction in energy use can be attributed to advancements in AI technology and more accurate calculation methods. While some have compared ChatGPT's energy use to other digital services, it's important to understand that different measurement methodologies make direct comparisons challenging. To put this into perspective, typical household activities consume significantly more energy, with many common appliances using hundreds of watt-hours per hour of operation. These comparisons help illustrate the scale, though exact figures can vary based on specific devices and usage patterns. Several factors contribute to this improved energy efficiency, including the implementation of more advanced hardware like NVIDIA H100 chips, optimization of AI systems and updated calculation methods that better reflect current technology. These measurements specifically refer to standard text-based queries and don't include energy-intensive features like image generation or processing extremely long conversations, but the cumulative effect of millions of daily queries still results in substantial energy consumption. The AI industry is investing in energy-efficient hardware and renewable energy sources to mitigate environmental impacts. As models become more sophisticated and handle more complex tasks, their energy requirements may increase, making these efficiency improvements even more crucial.

Speaker 1

Now let's dive into our main story of the day. Recent research has revealed that birds and mammals have independently evolved complex brains through distinct developmental, structural and genetic pathways. This showcases convergent evolution, achieving similar cognitive abilities despite profound differences in brain architecture and molecular mechanisms, and the University of the Basque Country used advanced techniques in spatial transcriptomics and mathematical modelling to uncover fundamental differences in brain development between birds and mammals. Published in Science, this research actually comprised two complementary studies that together paint a fascinating picture of brain evolution. One key finding is that the pallium, a critical brain region responsible for cognitive functions, develops differently across species, despite sharing similar functional outcomes. In birds and mammals, while the circuits perform comparable tasks, the neurons form during embryonic development at distinct locations and times. This divergence extends to the genetic tools employed for establishing cellular identity, resulting in unique cell types across species. What's particularly interesting is that inhibitory neurons have remained remarkably similar across species for hundreds of millions of years, while excitatory neurons, responsible for information transmission, have evolved along dramatically different pathways. In fact, only a small number of excitatory neurons in birds share genetic similarities with their mammalian counterparts, specifically in brain regions called the clostrum and hippocampus. The evolutionary paths of birds and mammals diverged millions of years ago, resulting in distinct sets of genes being used for sensory processing neurons in birds and mammals. Yet nature has shown us that there isn't just one blueprint for building an intelligent brain. Despite these differences, both birds and mammals have achieved remarkably similar cognitive abilities through convergent evolution. Many bird species demonstrate impressive problem-solving abilities and complex social behaviors, showing us that advanced cognition can emerge from very different neural architectures. The study challenges the traditional understanding of brain evolution, demonstrating that complex neural circuits can arise through diverse developmental pathways. It provides crucial insights into the genetic and cellular mechanisms underlying the independent evolution of complex brains.

Speaker 1

That's it for today. Thanks for tuning in. Don't forget to subscribe on your favorite platform For more info on anything we covered today, check out the links in our episode description and don't forget you can now access Perplexity's AI-powered knowledge base on the go with the mobile app, available for both Android and iOS. There's also the Perplexity desktop app for macOS. In other Perplexity news, the company just launched Deep Research, a powerful new feature that digs deep into hundreds of sources to create thorough research reports. The tool, which is free to try and unlimited for pro users, can tackle complex research questions in just minutes. Early tests show it's remarkably accurate, making it a promising tool for anyone needing in-depth research assistance. We'll be back with more stories that matter. Until then, stay curious.