I also write at https://splittinginfinity.substack.com/
harsimony
Karma: 683
Personhood for digital minds is good
This is a really excellent response thank you.
Annual solar irradiance differs by about 2-3x between the darkest and sunniest inhabited places. So on paper, if the world only used solar, the darker places would have ~3x more expensive electricity (it’s a little higher because seasonal variation and clouds increase the required number of batteries).
https://globalsolaratlas.info/map?c=35.245619,-93.603516,3&s=30.448674,105.117188&m=site
If solar achieves $20/MWh in sunny places, $60/MWh in cloudy places isn’t too bad! That’s cheaper than wholesale electricity prices in most places today.
As for whether renewables can lower electricity prices, looking at wholesale electricity pricing is tricky because governments do a lot of weird things managing the grid and setting prices. Good review here:
https://janrosenow.substack.com/p/do-renewables-make-electricity-cheaper
Fair point. A different way of putting this: when the raw materials are a significant fraction of overall costs, overall costs really do move in proportion to raw material prices. Because in this case, the relative cost of raw material to equipment actually matters for the design of of the product.
But when idiot index is high, small changes in raw material cost have ~zero effect on the cost optimization of other components. So index 100 is too far of an extrapolation.
Perhaps for fusion I need to look at the capital cost and lifetime of superconducting magnets and radiation shielding instead.
Agreed. I need to write a post on this but longevity tech looks a lot like nanotechnology for repairing our bodies.
https://www.writingruxandrabio.com/p/autoimmune-therapies-an-inspiration
I’m pretty optimistic about the future in general! I’m hopeful that AI creates enough abundance that we can pursue all sorts of wild ideas without needing to justify it with profit. A world where countries start a Mars colony just for national pride, or one where civilians leave on a lightsail just because they can.
AI safely automating all tasks:
https://splittinginfinity.substack.com/p/on-ai-scaling
https://splittinginfinity.substack.com/p/training-on-aligned-data-mostly-solves
On flying and self-driving cars :
https://splittinginfinity.substack.com/p/links-29?open=false#%C2%A73
https://splittinginfinity.substack.com/p/new-cities-for-self-driving-cars
Hyperlogistics:
https://splittinginfinity.substack.com/p/delivering-a-revamped-mailbox
Recycling atoms:
https://splittinginfinity.substack.com/p/recycling-atoms-with-supercritical
Remaking our economy around solar, wind, batteries (and a little natural gas):
https://splittinginfinity.substack.com/p/net-zero-part-1-energy
Space tech:
https://splittinginfinity.substack.com/p/the-economics-of-space-tethers
https://splittinginfinity.substack.com/p/the-high-potential-of-satellites
https://splittinginfinity.substack.com/p/a-glass-oneill-cylinder-in-leo
Further in the future, I see building a Dyson sphere and whole brain emulation as feasible (I’m writing posts on these).
My thinking is that a single lane of close-following busses would be comparable in throughput to a single rail line? Trains pack people more densely per unit floor space and travel a little faster, but a lane of busses has a higher density of cars relative to a train (counting the empty track in the car density estimate).
Ah interesting. I’ll look though that podcast because it’s worth mentioning in the permitting section.
At first glance, I do see the argument for why permitting will limit building power near term. But long term I think the incentives to build AI datacenters will overcome this.
Absolutely. For example, my estimate on fusion has a large enough error bar that (with some luck) it could beat solar. Just being wrong there implies a totally different energy system.
This is why I’d prefer to see people taking lots of risky shots on goal. The upside is very high.
Tech I’m skeptical of and why
Interesting that an antagonist seems to reduce your sleep need. Maybe medications that increase REM are the way to go rather than stimulants like orexin?
On whether sleep need reduction is possible, it’s definitely an open question. I’m encouraged by the existence of short sleepers and my hand-wavy argument that sleep evolved to save energy. But entirely possible that we discover a deep reason sleep reduction is infeasible. https://splittinginfinity.substack.com/p/sleep-need-reduction-therapies
I’m excited that orexin agonists are getting more attention from pharma companies. Sleep need reduction is one thing, but helping people with narcolepsy escape their circadian prison is really important! https://autismcrisis.blogspot.com/2010/05/circadian-prison.html
This is neat. Is the dose of glycine supposed to happen the day following a night of sleep deprivation? Or the following day?
That would be neat, though I’m not sure I want to use up even more of my collaborators time/effort doing standardized exercise!
Anecdotally, the sleep deprivation doesn’t seem to affect our performance on cardio-type tasks but does seem to cause trouble for strength training. But perhaps that’s a psychological effect. I’m sure there’s some literature on this too, might be useful for pinning down the benefits of sleep.
This was my hypothesis going in, that mild/acute sleep deprivation does little to cognition. This is a more controversial statement than you might expect! People have gotten mad at me for suggesting as much.
What muddies the waters is that everyone feels like things are way worse with sleep deprivation even if there’s little measurable difference. I felt it too, I’d be grumpy about how I didn’t sleep much last night and then I would pause and review my alertness and productivity and realize that there wasn’t much difference to the day before. Is mild sleep deprivation all in our head?
Are there vigilance tests besides the PVT that you think would be good to add here?
Yeah this is a good point though I think we need more data to conclusively say “sleep deprivation doesn’t make you dumber on these tests”.
If you can think of other tests to try here I’m happy to take suggestions!
One thing to note is that even if the cognitive tests can’t tell us much, we can still have a successful trial. If orexin allows you to sleep slightly less while feeling just as good and being as productive in your job I would consider that a success.
That’s a great post! It did more to popularize the idea than I ever could. I’ve been thinking about this for a while and my first writing on the topic was in 2021. I’m going to refrain from linking to it because I’m planning on depreciating that blog soon though.
Re storage and handling: this part was tricky, we opted to dissolve in sterile water (not saline) and froze the batch after mixing. So doses were only exposed to room temperature for ~minutes. We also used a C-terminal NH2 peptide that is less susceptible to degradation.
There’s still no guarantee that peptide is stable under these conditions and we don’t have a good way to check. This is a big reason why we thought there was a ~60% chance of a null on this trial (and the next one too perhaps). But hope springs eternal!
Thank you for offering a more constructive comment.
We did a power analysis to set the total number of trials (iirc assumed d=0.5, alpha=0.05, 80% power, so ~30 total test weeks and 10 weeks/person). However, the design proved unsustainable for us and the Fitbit dropped one persons data.
Though in some sense it worked out, we can pursue a better trial now.
Thanks for this background, it’s super helpful!
For dosing, I think we based the dose on the Deadwyler et. al. study in monkeys as well as user experiences.
Unfortunate that a 10-20x dose in humans seems to have small effects? I would have expected them to use a C-terminal -NH2 modification on their Orexin-A to prevent degradation, but it doesn’t look like it (but maybe they did and I missed it). If that’s the case, might reduce the effective dose gap between ours and theirs somewhat.
I’m excited to see our results at a higher dose, though part of me is frustrated by how difficult peptides are to work with. But hopefully Takeda or someone else will perfect small-molecule orexin agonists!
Yeah we used C-terminal NH2 modified orexin to prevent degradation but its possible it simply wasn’t effective.
Interesting that orexin sprays haven’t been working, I’ll have to look into this. Do you know the names of any off the top of your head?
Love that post! Made me realize that sleep need reduction therapies have to be pretty specific in what receptors they hit. A stimulant like modafinil that hits orexin + other stuff doesn’t reduce long-term sleep need in healthy individuals right? So a sleep need therapy needs to stimulate for just the right window, while enabling efficient sleep at night.
This is great thank you! One missing piece of info is the land footprint of a BRT lane versus a subway. If (say) 2 BRT lanes fit into the right-of-way of the subway that changes the math a bit. At first glance BRT lanes seem comparable in width to subway lines though.
And I agree on comparing to a driver-less electric bus system. A fair comparison between busses and trains will assume both have high levels of automation. Current AV’s are a proof of concept that both can be automated.