Investing @EnergizeCapital | Over the last 15 years, I’ve worked in energy including startups, incumbents, & investing. I share that experience here.

DFW ➡️ CHI Subscribe 👉🏻
Maybe the coolest map of global power plants ever made.
32
346
1,737
192,416
Different version of fall baseball, but will always take 70 and sunny in Chicago this late in the year. @NUCatsBaseball
205
Pretty wild to me that Amazon violated Bezos’ law that customers will always want it better, faster, cheaper. “The customer is always beautifully unsatisfied,” he said at one point. In times of change, you can count on what stays the same and Amazon hasn’t done that here.
Today we’re announcing Personal Agent Protocol — an open standard @Meta and @SierraPlatform are developing along with industry partners at @Genesys, @instinct, @RocketOTD, @Shopify, @stripe, and @Walmart. It will help define how personal agents interact with businesses and is open for anyone to implement. Read more: sierra.ai/blog/introducing-p…
1
2
474
Here’s a scary thought, and the strongest argument for a better domestic supply chain. There are ~100 substations that are critical to the grid. If someone coordinated a physical attack to hit 1 at each critical interconnect, the grid could go down for 18 months.
252
This is an under-discussed point for industries that are traditionally slow to adopt software from @kwharrison13. All the headlines are free customer education. At the same time, they're raising the bar for what software is expected to do. This lowers the CAC for everyone and is a larger version of what the iPad did for Procore and other software used in the field. There's a massive subsidy right now for architecture and infrastructure. The money pouring into LLMs and the infrastructure build-out lets everyone from startups to incumbents build cheaper. Net, you can acquire customers and build at a more cost-effective rate right now. The key is finding large markets that are eventually profitable without those subsidies. That distinction matters. Large markets with big tailwinds aren't always profitable (hello, Climate 1.0) as he points out and I lived. And the last point is that you can build for the current and future state at the same time. After all, if you didn't build a company to survive in good times and in bad, you didn't build a company.
1
229
Indigestion kills more companies than starvation. But, I think you need this $ if Meta is the competition.
Scoop in @dealbook : Instinct, the hot AI assistant, is now worth $10 billion! With only 14 employees. It raised $1 billion from @benchmark, @sequoia and @coatuemgmt. nytimes.com/2026/09/28/busin…
338
There’s a section of my feed that suddenly discovered Brandon Flowers as if The Killers haven’t put out 5 great albums in the last 20 years.
Project Hail Emmy 🚀
24
44
759
122,829
Energy being WAY flatter than the other technologies is the other story here.
the price of AI has fallen faster than for any other transformative technology in history we really are getting intelligence too cheap to meter
1
2
490
Did Outlook push an update that makes Copilot appear aggressively on the side now? I hated the little bubble, but this is a new level of bad.
380
Kevin Stevens retweeted
So much incredible information packed into this ⚡️🤓👇🏻
Halcyon's Gas Power Plant Tracker now covers 505 planned US gas plants (~286 GW) after adding 38 new plants this month. New this month: Dominion's 2,511 MW Mt. Storm, Crusoe's 1,398 MW behind-the-meter Duroc Project in Abilene, and the tracker's first CPCN denial: Duke's $584M Sherwood H. Smith Jr. Energy Complex (Richmond County, NC), for failing SB 266's least-cost test amid stranded-cost concerns. Learn more! link in comments 👇
2
7
1,367
The model options continue to evolve. Both Anthropic and OpenAI reduced costs this week, Jev is taking off, and now a new model from Hugging Face.
Introducing Contrastive Language Model (CLM): an ultra-fast System One Model trained with a contrastive learning objective that connects states and actions. CLM-8B is pre-trained on internet-scale data and delivers up to 9× faster inference than Jev ⚡ while achieving comparable performance across computer-use, gaming, and tool-calling tasks. With lightweight fine-tuning, CLM-8B sets a new SOTA on challenging agentic coding benchmarks, such as DeepSWE (81.6%) and Terminal-Bench 2.1 (87.6%). In contrast, Jev fails to serve as an effective verifier for these long-horizon tasks. We also build an efficient training and serving infra for CLMs by disaggregating states and actions, allowing their embeddings to be cached and reused independently. This substantially reduces inference latency in settings where the state evolves continuously while the action set remains fixed. Finally, we establish scaling laws for CLMs and show that the test contrastive loss decreases predictably as a power law in training compute, model size, and dataset size. 📄 Blog: contrastive-lm.notion.site 💻 Code: github.com/Contrastive-LM/CL… 🗣️ Discord: discord.gg/5dAQEDJBs 🤗 Data & Models: huggingface.co/Contrastive-L… More details on CLM’s architecture, data recipe, and scaling laws in the thread below 🧵
395
The AI buildout is 3x bigger than any other infrastructure boom ever - with the exception of railroads - as a % of GDP. In nominal terms, GDP in the 1870-1890 period was about $ 350B. Today, it’s $31 TRILLION. Not only have we never seen anything like this, we’ve never seen anything close.
224
Me: AI will rapidly penetrate every industry. Also me: utilities adopt technology at a different pace. Exhibit A:
1
3
753
Maybe the most important quality an investor can have.
219
Great point by @benthompson and one I’ve been saying this a lot. Models are now good enough for most people. Muse is running on a model that’s not state of the art and it accomplishes a lot for the average consumer. This is going to be a distribution war moving forward.
1
14
17,675
One of the replies in original post here nailed it. Once you have kids the amount of things you want automated grows by a lot.
Not once have I enjoyed having to go thru all that info to make travel decisions, been offloading as much as I can to agents and once they learn my preferences rather them handle it end to end Only thing worth doing and enjoyable is hearing from friends you trust on their opinion. I don’t care to look through every Google review, trip advisor page, travel blog, etc
1
4
1,057
Incredibly shortsighted by $AMZN in my opinion. Consumers will use agents moving forward, and the easier it is to shop the better it is for Amazon. My guess is the argument is now you can search everywhere at once which kills their position as the default, but the advantages they have in selection and shipping speed don’t go away.
Amazon cuts off Muse. While I am bullish Meta and Muse, I think many people are overlooking the digital knife fight that’s about to occur Nobody wants to get commoditized or layered here. Let the games begin
1
7
1,708
The New York PSC is giving utilities 60 days to report all internal AI use cases and the oversight corresponding them. The intersection of critical infrastructure and AI is going to be messier than people realize. It’s a huge opportunity for incumbents who already have stamps of approval.
1
1
4
461
These stories are becoming more common and there’s a lesson in this one. It’s going to be really hard to build a data center where a lot of new infrastructure is required. Oracle ran into a lot of permitting issues around the power supply. This creates an opportunity for those that are located near generation or T&D. The less permitting and government approval you need to build, the better. The natural question is will there be micro sites for compute in the way there are for energy? Will we go towards smaller data centers for inference that are then driven by coordination software? I don’t know the answers, but it’s clear mega sites that need a lot of new infrastructure built will be difficult in the near term.
1
1
264
This is where the world is headed. I work with my email through Codex / Claude, I organize my data so that AI can search it more efficiently, etc. There's no reason to reason to do anything other than chat with an agent.
all of my imessages & email now run through a single agent interface. i still text, but i deleted every email client from my phone & canceled every email subscription or app i had. gmail is basically just a database to me now. increasingly so are the rest of google’s services. i.e. i never want to manually write an email, create a doc, build slides, organize a calendar, or manage files again. i just want to express intent & have the underlying systems of record updated correctly. i think now you’ll see apps disappear quickly. many will become databases at best or entirely obsolete at worst. the agent will be the default interface to your personal & professional life.
1
7
1,048