On a related-different axis, I've consistently seen on-prem GPUs running identical workloads ~35% faster than the same workloads on the same cloud hardware, regardless of intermediate infra stack layering/versioning choices. Weird but I'm not complaining!
Are you referring to Bruno Latour or the author of the post? If you're referring to Latour, I would strongly urge you to read "Laboratory Life", which is [as] clear and easy to read today as it was when I first encountered it nearly 20 years ago.
Tadalafil (Cialis) is available as a generic in the United States. If you think "big pharma" is served by drugs going off-patent, you may want to re-evaluate your perspective.
If you want to get more reasonably unhappy with market exclusivity, look at Celgene's grip on thalidomide/lenalidomide/pomalidomide for the treatment of multiple myeloma.
How, in your view, does dilution change the therapeutic index of a chemical? If a drug is effective at 20 milligrams, but at 60 milligrams has a steep rise in the incidence of hypotension, how does taking that same 60 milligrams in a larger volume of filler (e.g. water, or spread across more physical pills) change the fact that you've just taken 60 milligrams?
> how does taking that same 60 milligrams in a larger volume of filler (e.g. water, or spread across more physical pills) change the fact that you've just taken 60 milligrams?
It increases tolerance to mismeasurement and mistake. Same reason many pharmaceuticals require multiple pills for minimum dosing despite a concentrated form existing. A child eating a single pill, or you missing that you popped an extra pill into your hand, causes less damage.
That doesn't change the difference between effective and toxic doses, it changes the potential scale of off-by-one user error. If 20 milligrams is delivered in 1 tablet, then 3 tablets is the toxic dose. If 20 milligrams is delivered in 5 tablets, then 15 is the toxic dose. A single daily tablet is far superior to multiple daily tablets in terms of patient adherence and hence disease management.
Can you name specific examples where the number of pills/capsules/tablets has been increased to improve drug safety? Increased pill counts historically reduces patient adherence, which worses disease management. This is just the first example I found which measured it: https://pubmed.ncbi.nlm.nih.gov/31298592/ This review explicitly states it in the abstract even: https://pubmed.ncbi.nlm.nih.gov/30561486/
Idk what to say, this is established medicine [1]. When you have a small TI you dilute to make measurement tolerances wider in absolute terms. This is a motivating factor behind prescription-strength medicine.
I don't see dilution anywhere on that page, and increasing volume of administration at a lower concentration to achieve the same effective dose does not alter the dose itself. You are not interpreting TI correctly.
Edit: further, to your comment about "prescription strength" nomenclature, look at section 14 of the Cialis/tadalafil prescribing information, IIRC, table of clinical studies, where they have the second two outcomes of the clinical studies broken down by dose. Efficacy increases pretty directly with increasing dose, and these are where the observed side effects show up. It seems like patients may well self-escalate. Maybe the OTC countries have public data on this?
NTI drugs, a/k/a critical dose drugs, can avoid titration requirements through dilution. It’s harder to fuck up a 500 mL difference than a 1 mL difference. Again, this is why most OTC versions of prescription drugs are different in only one way: concentration.
Fair. This holds a place in my mind as "basically the same as alkyne - azide" since Tetrazzine is "more or less azide" and dbco is "more or less cyclooctyne". More to the point of the thread, I don't think the Tetrazzine/dbco is orthogonal to alkyne-azide -- they will cross-react -- am I wrong about that?
Apologies -- tetrazine reacts principally with trans-cyclooctenes. So azide/DBCO would be the orthogonal group. To the best of my knowledge, those two pairs do not cross-react.
I feel like fingerprint/iris/retinal scanners would be a reasonable stop-gap here if only there was some way to deploy them cheaply. Does anyone know if such add-ons exist that might be made compatible with a 2FA system?
This is a fabulous convenience! The reach of this ready-to-go data will be much larger (in some directions) than the model and CASP results themselves.
Yes, but photo-voltaic also implies energy conversion for power operations, like to run appliances. This is energy conversion for detection, like a sensor.
