Ash Rust

Space Is the Next Bitcoin Many people I meet complain they missed the boat on Bitcoin while others made a fortune. More than 10 years later, I believe we're now on the precipice of a similar boom in the space industry, but the fundamentals are much stronger. Humans were always going to operate in space. We have looked to the stars for our entire existence and nations remain captivated by the minute-by-minute actions of their astronauts as they boldly go. We were on vacation for the recent return of Artemis 2, but my wife insisted we watch the landing live. While Bitcoin has spent more than a decade trying to demonstrate use cases beyond 'store of value', the first commercial use cases for space are generating massive growth. Starlink revenue in 2025 was $12B. Reusable launch from SpaceX is driving down costs for the entire market and, in turn, we're seeing the use cases expanding. Varda has been manufacturing medications in space for over 2 years and at Sterling Road we've invested in data centers in space (Starcloud), accommodation on the Moon (GRU Space) and nuclear reactors for space power (Zephyr Fusion). The space market is still very speculative, offering both enormous potential and minimal competition. SpaceX is planning to IPO this year at $1.75 Trillion, almost a 100x multiple on sales. Starcloud became Y Combinator's fastest ever unicorn and reports suggest its valuation doubled again in the last 60 days. Skeptics will correctly point out that the math has a long way to go before profitability but that has never stopped incredible founders from building companies that push humanity forward. If you're looking for the next Bitcoin, it's here.

@HankShedwrecker I invest in founders who are getting started with their business ideas. Several of those companies have grown to $1B+ valuation and created numerous jobs.

I shit you not the *incredibly vital* company he owns is a consultant firm that helps people in tech get seed money and go public. This whole thing is such a self licking bullshit cone.

@ScotIrish1719 @n6532l I was born in Lancashire.

@AshRust @n6532l No where were you born? Then went to UK and now here? Your not American

My green card application took ~18 months and I worked in the Bay Area the entire time. This would have sent me back to the UK before I was able to start a company or a fund. It's really damaging.

@matt_cavanaugh_ The constitution, the cultural diversity and the opportunity for all.

@AshRust What are the three things you most love about America?

@ArmedAndGentle I have 3 CS degrees, 2 from Oxford and I came here on an O1 extraordinary ability visa.

@AshRust Well, your application for the adjustment of status was weak, and should've been denied. I know people who's petitions were approved in less than two weeks.

@JonahTreible I’ve paid taxes since I arrived 20 years ago.

@AshRust So for 18 months, you were just using the American system using our healthcare system not paying taxes just twiddling your thumbs good enough. We should send you back to wherever you came from for 18 months then

@DiscipulusMundi Before I moved to the USA, I was the fifth generation of British Military in my family.

@AshRust You're not even British. You're just stringing along a chain of fraudulent nationalities for economic gain. You should be denaturalized and returned to Tajikistan or Azerbaijan or wherever the hell you came from.

@leesteapleton Lol. I'm 5th generation British Military.

@n6532l I am an American.

@AshRust “…It's really damaging.” To whom? Not Americans.

@suzukisapers 1st yes

@AshRust were you here on a work visa?

@Shoestring_Lab I'm a US citizen, I am home.

@AshRust We don't care. Everyone ia going home.

@StBruh6969 I'm a citizen too, I'll wager we both care deeply about the constitution.

@AshRust Why should I as a natural born American give a fuck about you ?

@TimTimboFL I write first checks for people just getting started, not sure you have the right person.

@AshRust What will we every do without venture capitalist in our country driving up the price of everything and making it shittier

@FacingNorthX I became a citizen under Trump.

@AshRust Wow. 18 months. With the right leadership, we can denaturalize and deport you faster than that.

@robertmclaws @garrytan My green card application took ~18 months and I worked in the Bay Area the entire time. This would have sent me back to the UK before I was able to start a company or a fund. It's really damaging.

@garrytan @AshRust People who self-deport can come back in 2-4 weeks. Asking people to complete their permanent applications from outside the country for that period of time is not a huge ask. And they already said that certain types of visas will not have to do that.

We need to keep smart people in the country to build the future and build tomorrow’s businesses that employ millions of people This is bad and misguided policy

The Solo Engineering Team: My Claude Code Setup It has been more than 10 years since I was committing code regularly but over the last 4 months, I have custom built 80% of the startup investor stack, using only one app: Claude. All this without interfering with my regular job, as I also made 17 investments during the same time. Here's my exact setup and approach: 1. All Tasks in Claude Code. I use Claude's Code tab for almost all my tasks, from spreadsheets to coding. I also don't leave tasks in my list, all new ideas go straight into Claude Code for me to address when I can. 2. Easy Start. I don't want to think about the right approach to a task, so I wrote a Claude 'skill' that does the work for me. I begin every new session with /lets-start and let Claude select the right approach using the gstack skills package. Completely free and open source (link below). 3. All the Plans. Gstack offers brainstorming and planning modes before anything is built. I do everything. Office hours, CEO review, design review and engineering review; then adversarial reviews of those by ChatGPT's Codex (optional, gstack will set this up for you). I want the best plan possible, with minimal effort from me. 4. Concurrency is the Win. All this planning and effort may seem like overkill or a slower approach for many tasks. That is true but the productivity gains come from progressing on *all* your tasks concurrently not speeding up individual tasks. 5. Just Hit Recommended. The side effect of heavy planning is that Claude will have lots of questions. I choose the recommended route unless it would mean missing a gstack skill, in which case, I choose the gstack skill option to ensure we have done everything possible. 6. Verify before Assert. Stopping Claude from making assumptions is a constant battle. The AI knows it's smart but is often wrong, so /lets-start includes a global setting for Claude to ensure it verifies something is true before assuming it. However, you should still expect to see this behavior. When it occurs, ask Claude to save to memory that it must verify before asserting. I had to do this many times. 7. Parallelize Large Projects. If I'm building a medium or large application from scratch, the plan could take many hours or even days to implement. I don't like waiting, so /lets-start now includes a /parallelize skill that breaks the plan into multiple, distinct session prompts, allowing you to run them concurrently and complete the work faster. Be warned: this requires more attention as you're managing numerous sessions per project. 8. Full Testing. Agentic coding carries numerous security risks, and there are already high-profile examples of hacks and data leaks. I don't want to deal with any of that, so I wrote a skill to /autoclean the code for me. It will first audit the existing tests and create new ones if needed, then it will clean up any sloppy code and finally it will pass the code to gstack's security review. 9. Bug Fixing. Bugs are an unavoidable part of the engineering process. However, when bugs are fixed, I always ask Claude to add regression tests to ensure they can't recur. If you don't stop the app from making the same mistake twice, it will happily take the opportunity. 10. Persistence. Incomplete apps and frequent bugs make using Claude a potentially frustrating experience requiring significant iteration, but the potential rewards are worth your time. Once it works well for you, the productivity gains will be massive. Anyone can now do the work of an entire engineering team for ~$200/month. This is the time to build! Prompt to install lets-start: github.com/ashrust/lets-s…; lets-start will install gstack for you too.

One of the most important use cases for AI and already helping thousands of people #proudinvestor.

We’re launching the missing team layer for AI-powered development 🚀 The SDLC was built around coding. Coding has changed, and the SDLC must be updated. Lightsprint makes agentic development collaborative with visual plan mode, parallel cloud agents and PR preview environments. Try now: lightsprint.ai

Build While You Search for a Cofounder If you want to attract a truly incredible co-founder, it’s best to show progress first. 1. A Deck is not Enough. Decks are an essential part of fundraising but not cofounder recruitment. Plus you want a cofounder who is there because of the vision not the potential funding. 2. A Demo is Baseline. Agentic coding means anyone can build and thus the expectation is that you make good progress even before there’s a team. 3. Customers. Outside of building, you can also demonstrate progress with Customer discovery interviews, a waitlist and engaged customers on the beta.