Profile avatar
harsimony.bsky.social
I write about opportunities in science, space, and policy here: https://splittinginfinity.substack.com/
520 posts 141 followers 383 following
Regular Contributor
Active Commenter
Posts 22 Comments 28

Taking testosterone as a performance-enhancing drug has made me sharper than I was a decade ago, and along the way improved basically every aspect of my life. I wrote about it on substack usefulfictions.substack.com/p/testostero...

submitted 31 days ago • 0 comments

james dyson just seems like an absolutely lovely old man, what a humble fun no-frills presentation of world best engineering www.youtube.com/watch?v=ve6J...

submitted 15 hours ago • 0 comments

Yes! This is the kind of AI company I want to see. One that gets their hands dirty trying to accelerate specific tasks. More practical and less x-risk compared to building general intelligence. nitter.poast.org/MechanizeWor... www.mechanize.work

submitted 19 hours ago • 1 comment

I think TC is obviously wrong here and should've issued a clarification rather than put up a wall of weird arguments. I'll keep reading MR, but I trust Tyler's reasoning less now. marginalrevolution.com/marginalrevo...

submitted 22 hours ago • 0 comments

second order optimization methods are having a bit of a moment/resurgence right now. suppose we solve those, what's next then? what is optimal optimization? I wrote a new small note on my thoughts on what infinite order optimization methods could look like (link in next post)

submitted 5 days ago • 1 comment

3D printing keeps improving. Commercially of limited importance, but interesting nonetheless. www.youtube.com/watch?v=Ek_7... www.youtube.com/watch?v=W1d3...

submitted 2 days ago • 0 comments

Part II of @toughsf.bsky.social's series on lasers. Takeaways thread with my own thoughts at the end. 1a. Tanks are safe-ish against laser weapons with some mitigations. 1b. Lasers can protect tanks from airplanes, drones 1c. Result: tanks make a comeback! toughsf.blogspot.com/2025/05/the-...

submitted 2 days ago • 1 comment

Great post by @toughsf.bsky.social. Thread of takeaways: 1. Lasers req. heavy energy/cooling systems making them less mobile and favoring defense. toughsf.blogspot.com/2025/05/the-...

submitted 2 days ago • 1 comment

Two ToughSF posts in a matter of days. So exciting!!

submitted 3 days ago • 0 comments

Taiwan remains a major flashpoint. China *might* be building this stuff for publicity reasons, but best to assume they are planning an invasion. There are weapons and support the U.S. and others can (quietly!) give to Taiwan to make invasion too costly for China. www.youtube.com/watch?v=_GJJ...

submitted 3 days ago • 1 comment

This o3 output was linked by Tyler Cowen on Marginal Revolution. There's definitely some self-dealing in the response but ... Most people haven't come around to the fact that frontier LLM's are fine tuned on a very particular subset of the internet. chatgpt.com/share/683335...

submitted 4 days ago • 3 comments

Wow, the proposal is actually doing pretty well! We might actually be able to do the more extensive test of OxA and OxB

submitted 9 days ago • 0 comments

The sleep need reduction proposal with @niplav.bsky.social and nomagicpill is up!! manifund.org/projects/ore...

submitted 9 days ago • 2 comments

The bizarre and illegal armed takeover by Administration officials of USIP, an *independent institution mandated by Congress*, has been reversed for now.

submitted 10 days ago • 0 comments

no link deboosting on here is the most premier feature — once again, deepfates was right. this is a great platform for that alone

submitted 10 days ago • 3 comments

Big post up! I discuss why sleep need reduction is feasible and why I think orexin agonists are the most promising place to start. More details to come on a self-experiment on this topic. splittinginfinity.substack.com/p/sleep-need...

submitted 11 days ago • 0 comments

Statins are probably the best life extension drug we have right now. www.alexkesin.com/p/mr-secreta...

submitted 13 days ago • 0 comments

Lets be a little careful here; while it seems to be the case that there have been real performance challenges, it is also not uncommon that other clean energy technologies (EVs, solar panels, etc.) take a year or two of operation to pay off their embodied emissions:

submitted 13 days ago • 9 comments

"Data Science at the Singularity" argues that frictionless reproducibility is the wild thing happening in ML: data sharing, code sharing, and competitive challenges with easy access. It's also a challenge to every field, to ask how they can become more like this assets.pubpub.org/9bk0194n/Don...

submitted 13 days ago • 2 comments

So in when talking about nuclear power in the long term, seawater extraction often comes up. A slightly less often discussed aspect is the Energy Return On Energy Invested (EROI) of this process. This tell us if the whole thing is even worth thinking about. 1/6

submitted 14 days ago • 2 comments

Although curmudgeonly, the actual scientists who do takedowns of hype-y things have a great record I think the Heinlein quote holds — scientists saying “that is impossible” is often wrong. But “this actually existing work is not as exciting as you think it is” is often right.

submitted 16 days ago • 1 comment

It's good when informed, consenting adults can try out medical treatments at their own risk. It's not likely to succeed, but they're benefiting everyone else. Regardless, freedom is good.

submitted 16 days ago • 0 comments