Making Things Happen

So I want to pass along this skill that I think I have. Elon has been described as having the eye of sauron on a particular project, where the bottlenecks just disappear away when he pays attention to them. I've felt a similar thing -- and I think that a bottleneck on this happening is just executive time, and by chance / luck, I've got a lot of the executive time at CAIS.

It's pretty easy actually. You ask, how long is this going to take. Why. What are the key bottlenecks. Why can't they be solved with this 80-20. Ignore all the rules that say it must go slower. Figure whether you can throw money / time at the problem, and then make it go faster. Be very suspicious of things that look like planning meetings or "all we need is to hire someone to do that" or any other things that does not solve the exact thing. Rinse and repeat for all of the teams.

On a meta-level, zoom out and ask, what is on the critical path, the set of things that need to get correct for the final goal to be achieved. Make that path as short as possible, then ruthlessly execute until it has been achieved. Do sustainability things like relaxing in between achieving goals (but oftentimes the adrenaline and excitement you get from achieving your goals will be enough).

Your final goal should be something that is insanely ambitious, concrete though, and the upside so high as to it definitely not be satisfiable. Satisfice (do to the bare minimum) everything else. Make sure that you understand the power of exponentials, and pick the right ones to bet on. Everything is power-law distributed, make sure that of all of the things. Doing things early matters.

Joyce pointed to the lesswrong concept of steering versus rowing. I think almost everyone should do more steering -- thinking about which direction that they are pushing towards, and optimizing for that. You need to make sure that you are exposed to enough surface area so that you don't miss the obvious ways that the tides are moving. But the execution is often easier once you've got a clear vision.

As Thiel says in his book, the complex multi-step plans work. Try it! The rush of achieving one is incomparable.

There is the organizational version of this question that is more interesting. If the bottleneck to things happening within orgs is the exec time, making decisions, what can be done to enable a whole organization to move faster? The eye of sauron can only be spent in one direction for any given time. There can be cultural changes that make things easier, principles. For example:

The priority should be creating actionable proposals. These should be clearly written, and written quickly. The proposal writing should never be a bottleneck. If there is a bottleneck, the bottleneck should only be someone's proposal. Everyone should have practice spending one hour and then writing a proposal. Go round and check whether people have this ability. Teach them to use LLMs if they don't. Teach them to write clearly and concisely.

There should be a person for whom's full time job is to go through those proposals, chase all the ones that haven't been written, spend ~2 hours a day with the exec, perhaps between all the meetings and in the off-time, summarizing all the feedback + getting the decisions. This should be a chief of staff type role, and they should have both a really good strategic sense (because they need to be able to prioritize).

It should be clear what doesn't need to be approved (eg. all purchases below a certain amount -- eg. $500, hiring a contractor for a specific job), and then which things can be approved by someone lower in the org chart.

There should be a good amount of contact with the people in the process, and the exec should be able to set direction for projects to be organized, but there also needs to be a reasonableness check, so that side quests that are not worth the time are vetoed.

The approver of those proposals should be able to do many of the smaller things async. It may be the chief of staff's role to summarize the proposal into a signal chat, which can be thumbs up-ed or voice messaged.

The organizational priorities should be known to everyone. For ones where org priorities are changed, they should be communicated.

Systems should only serve the goal of achieving the strategic vision. The strategic vision should be multi-faceted, but everything must either directly work towards the critical path (which can be updated) or improve systems/people that are working towards the critical path. The story of how all the projects can improve the critical path, and by how much, should be common knowledge.

Bad people must be cut quickly. Their actual cost is not their salary, but the attention/emotional energy/counterfactual hire. Average competence matters a lot for a well-functioning org (think about why a 5 person startup can often perform as well as a 500-person team).

Pay very good people well, and treat them well. Understand the power-law for importance to the org. For the people who are critical to the organization, all the blockers on their time should be lifted to the extent that is reasonably possible.

Threads should be managed on very zoomed out level, and there should be willingness to drop everything for a strategic shift, and move quickly and confidently into a new area if it is important.

Leaders set a behavioral precedent for the rest of the org. They should work hard, be kind, be strategic.

The org should have a lot of surface area, for information flow, for hiring, for strategic insights. Relationship management with important people is important, but the idea should be for the org to contribute value -- everything follows from that.

The highest level goal must be defined and internalized by everyone at the org. Everyone should feel empowered to ask questions about the strategy.

There should be a strong norm for people completing tasks by when they said they will, or communicating that they will not. There is no excuse for not doing this, really.

Notice where you feel cautious or fearful of acting, see if you can diagnose it to organizational scar tissue (eg. nearly dying from lack of funding, and therefore adopting a scarcity mindset around spending money), and find people who don't have that organizational scar tissue. For example, it's not obvious at all that the default way of doing things / default prices that are paid / default speeds are optimal at all -- they should at most be considered a reasonable prior. Trust the private world and those with competitive pressures such as other functioning orgs (like high-levels of government etc.).

Leadership should be critically focused on self-improvement. What are the areas that they don't do well at, and how can they hire / change to compensate for that. Hire a professional who understands this. Mentors are potentially better. They can identify blind spots.