Halcyon

At the #deploy24 conference, many speakers talked about the rising demand for #electricity from electrification and #datacenters. Increasing demand would be a welcome change of pace for many in the industry. When I started working with power in 2008, the US consumed about as much electricity as it does today (see picture). To help analysts, consultants, and investors get back into a growth mindset, Halcyon is developing tools using the latest #AI technology. From notifications of docket changes to easily-accessible catalogs of IRPs, the tools allow access to information faster. Click below to find out more.... https://lnkd.in/gsqv8HE5

It’s here! The 2025 World Positive Report has arrived and covers the incredible impact our portfolio companies are making on the world. https://obvious.com/wpr/ We’re all in on generative science—AI trained on data in fields like chemistry and biology. We believe this game-changing technology will become humanity’s greatest invention because it is, for the first time, an invention that can invent. Join us as we share our vision of the future and the world positive companies working at this new AI frontier: Halcyon, Zanskar Geothermal & Minerals, Inc., Diamond Foundry, Synop, Arbor, Inceptive, GenHealth.ai, Pi Health, Virta Health, Synteny, Brightside, Incredible Health, Prism Data, and Dexterity, Inc.

Data center theory, and data center practice. This week, both Meta and Amazon have announced big plans to use nuclear power to meet their growing demand for data center energy. These two companies, and their hyperscaler peers Google and Microsoft, have $200 billion of capex planned this year, most if it constituting an incredibly dense build-out of electricity-intensive infrastructure. It's news release and press release stuff - important, but theoretical. What is the practice, then? It's less flashy; it's less easy to get; it's both smaller and more precise. Data center electricity demand starts with grand announcements and flashy press releases; the real work moves through proceedings, testimonies, plans, forecasts, and filing after filing after filing. I did some digging, at that paperwork level, into just one new US data center growth area, in the Rappahannock Electric Cooperative, where data center electricity demand could go from zero to 3,000 megawatts in the next 15 years (and figuratively or maybe literally ruin the utility business model in the process). It's why REC is exploring an entirely new contractual structure in order to meet this demand, protect consumers, and plan for growth. Discovery, search, and investigation through Halcyon querying, and lots of useful external references too. I learned something the true state of data center plans in one market in the process; you can learn the same by reading. https://lnkd.in/gY22P42F

I cannot think of how many thousands of pages of decarbonization documentation I have read in the past two decades. It's probably enough thousands to reach into the millions. That much important stuff to read can feel almost like a well-intentioned denial of service attack on getting work done. But, information is one of the energy sector's connecting layers, just like pipes and wires, and it it has its own known unknowns. Any energy professional knows what sort of information will inform their decisionmaking. They often know where it will come from, but not when; they know that there's importance lurking in that third response to a request for proposals, but not who the important party is. They can identify authorities, but not every authoritative source. Awareness of an information challenge is high; discoverability of information is very low.  Solving for this challenge is a longtime goal of mine - and so I'm very happy to tell you what Halcyon has been working on: an AI-tuned notifications system directed, in the first instance, at US regulatory publication streams. If a public service commission is a feed, and every document in it is ready for AI-driven classification and summarization, then the best way to discover information is not to look for it, but to have it served directly to you - pre-processed, contextualized, linked to source, simultaneously simple enough to pique interest and robust enough to enable further inquiry. The public beta is out today. Please click through to see what's in place now, and what's coming (not just markets, but topics and themes), and to start receiving notifications directly. It is exciting - and hopefully, it is the gateway to utility (no pun intended) in a great unstructured data energy transition challenge. https://lnkd.in/gCvpMAX3

About two months ago I received a note from Azeem Azhar, wondering which professional market researcher or forecasting group would take a wager on future solar power installations. If you are reading this post you probably know that forecasts have....somewhat undershot reality in the past. I wrote back a long note, not about who might take the wager, but about why forecasts for solar in particular have historically done a poor job of predicting reality. It was such a long note, in fact, that it became the basis for an essay for Exponential View. It's about observed reality and theories of change, descriptive near terms and proscriptive long terms, purposes, and priorities. Quote: "I do not believe that analytical skills are the missing ingredient in thinking about the future. Rather, I believe that imagination about the future and a theory of change that helps to describe it clearly, are what is needed to look ahead in a more compelling way. It is not a lack of computing power, so to speak, but a lack of programming." I enjoyed working on it with Azeem, Marija Gavrilov and Nathan Warren. I hope you enjoy reading it, too. https://lnkd.in/geJXjaxS

I'll be in DC for the U.S. Department of Energy (DOE) #deploy24 conference Dec 4&5 to discuss the impact of #AI on energy development. If you're around, let me know if you'd like to see what Halcyon can do to accelerate your business. https://lnkd.in/ei7RjRuv

https://lnkd.in/gWbEbvZN Yesterday, I went on Bloomberg TV to talk about AI, Generative Science, and Climate Tech under Trump 2.0. There are reasons to be optimistic! The Inflation Reduction Act has already deployed over $200B dollars, with $165B going to red states. That's delivering good green jobs in states like Georgia, North Carolina, and Tennessee. At Obvious we're excited about #GenerativeScience to solve unsolvable problems in climate like discovering geothermal and inventing new and better batteries.