Hi folks - author here! Happy to answer any questions about the post, hopefully it's useful for folks following in Palantir's footsteps.
236 karma · joined June 26, 2011
I write at ssaraf.com and partly.work. You can reach me at sanjay[at]withpartly.com
* Webflow (in case the client needs something relatively unsophisticated)
* Django / HTMX / AlpineJS / Django Ninja (in case the client needs a full CRUD application with more powerful capabilities).
I also want to shamelessly plug - if you are looking to do freelance or contract work, and you want an easy way to manage the contract, engagements, milestones, and invoices, consider using withpartly.com.
I'm the founder and CEO, and looking for freelancers / contractors to play around with the alpha product (for free).
I'm creating software to make it easy for freelancers to manage their engagements, get paid, and handle all of the pieces of paper involved in that (NDA, IP assignment, etc). As part of that, I'm working with companies looking to hire contractors, specifically software engineers.
If you're looking to work with Series A / B companies as a software engineer for at least 10 hours a week (either in contract to hire or in permanent freelancer setups), please reach out! The main cost to you is you'll have to use my software and give me feedback on it. Thanks!
Email: sanjay@withpartly.com
While CRISPR is often used in cultured cells (eg for CAR T therapies) there are CRISPR therapeutics in live human patients today in phase 3 trials for sickle cell disease (among other indications).
This is an underrated comment. These are good threads to pull on in future posts - thanks.
Thanks for this comment - I think the economic point is spot on. I think the fear of losing control is a fair summary - unfortunately there is some evidence of a selection effect that might mean employers do have something to fear. https://www.fractional.work/p/showing-face-is-more-important...
To the Synth team, awesome product! Great to see that more tools are getting built to help testing / QA workflows. I think this a huge area for the future. Welcome to the competition. :)
[Disclaimer] I'm the (solo, bootstrapped) founder of Sudopoint
My initial guess would've been dynamic programming (levenstein distance), but if this is API-wide, did you make a suffix tree?
EDIT: To your second point: Even though it's 300 kg total, it looks like that's allocated to 2 75 kg objects and 6 25 kg objects. I would assume that means much less weight/space per unit to work with, given the dead weight to go along with it.
Still, I agree with you. It's a very difficult target.
> Ask journal editors and scientific peers to review study designs and analysis plans and commit to publish the results if the study is conducted and reported in a professional manner (which will be ensured by a second round of peer review).
If the study design is peer reviewed, I would hazard a guess that there would be less bad science, not more. Currently study design takes second priority to significant results, which is perhaps why we have the problems of inconsistent research in the first place.
I'm criticizing Time magazine for making the claim that salt has no effect on blood pressure (or implying it), which is demonstrably not true.
My point is: if you're healthy, no, it isn't surprising that your body can regulate salt intake. If you aren't, it isn't surprising that doctors will want to keep track of how much salt you're eating.
> For a med STUDENT you show quite a lot arrogance.
I wish you hadn't said this, because I think it's unnecessary and diminishes the rest of your comment.
The study may have its own problems, but draws a conclusion I agree with, that the relationship between salt and blood pressure in healthy people may be overstated.
As for your point 3: I disagree. I think the news article makes a case that salt and blood pressure have no relationship, which isn't true. It is only true if, almost by definition, your salt / water regulation system is healthy.
Your point about maintaining NaCl levels is probably roughly correct, though I don't think it has been demonstrated per se.
Salt does have an immediate effect on blood pressure, as you said, because of osmotic effects. Increased salt in the intravascular space will draw water from other areas into the vessels, thus increasing blood pressure. The kidney has both osmotic and pressure sensitivity.
1) All clinically treated hypertensive patients were excluded - that means if you have kidney or vascular disease (hypertension) and are being treated for it, you aren't counted here. Doctors WILL care about your salt intake if you have these conditions, as they should! It is very common to prescribe something called the DASH diet to lower sodium intake and increase potassium intake.
2) If you're healthy, it means your kidneys are healthy (simplifying). One of the purposes of the kidneys is to maintain homeostatic blood pressure by excreting a combination of salt and water. To put it simply, if you eat a huge load of NaCl, say in a dominoes pizza, your blood pressure WILL NOT be affected for a very long time. I'd guess 20 minutes or so after the salt enters your bloodstream. Which means these folks can't measure that blood pressure change by testing blood pressure yearly.
3) Either medical doctors or the press that listens to them have a nasty habit of taking treatments that often work for very sick people - morbidly obese, type 1 diabetics, bed-ridden centenarians - and applying it to normal healthy folks. I will go out on a limb and say that these unilateral recommendations are almost all BS, and should be ignored. If you're healthy: eating eggs will not affect your cholesterol, eating cheesecake will not give you diabetes, sitting in a chair will not misalign your spine. It would be prudent for you to be skeptical when you hear these claims, often on daytime television or on the internet.
Every few years people seem to think we've found all of the significant small molecules for human disease - they keep getting proven wrong.
It's also hard to get the virus to target the cells you need to reach, as they often only target one specific cell type.
Genetic modification is one of the few ways we can keep the bananas we like while adjusting them for more recent diseases and pests. The only other way I know is to go back and find the ancestral species, and try to breed a fertile banana cousin.
While it is true the banana would use more energy, there is no evidence for lower crop yields.
The banana has to take the Iron from the soil, unless it’s a very rich Iron zone, so you need to use a fertilizer to replace the Iron. This also makes the bananas more expensive.
The gene they are using is from the soybean plant. Since there is a significant amount of iron in soil to begin with, I doubt fertilizer is required to supplement the soil with iron.
I'm also not sure that extrapolating to Chicago makes sense. I live there. Since anecdotal observations seem to matter, it seems that the people who would order an Uber car would call a cab if Uber wasn't available. It's hard for me to imagine Uber providing such an addition in service that it would change DUI rates.
As others have pointed out in this thread, the incentives behind folks pushing the "learn to code" idea are varied: some people want more engineers to hire and some people want to sell to those learning to code. These seem like reasonable criticisms, but I think they miss the point.
Many of you, if not most, are in the software industry. You try to solve interesting problems every day, or, at least, you'd like to. Every industry and every field has the need to solve them, at least with my definition of 'interesting'.
Solving such a problem is difficult. It requires knowing very clearly about the rules of engagement, and mapping out strategies for tackling the issue. It requires an understanding of the lay of the land, but also the capacity to think outside the box. In short, it is exactly what all of you would expect from a 'good' programmer - the ability to think actively.
You'll notice that I make no claims of this sort about physicians. The best doctors may do these things, but most doctors don't. And at the very least, doctors certainly are not trained in ways of solving problems effectively. Instead, we are mostly trained in pattern recognition.
So what? Do I think a doctor would benefit from learning to program? Yes. Because figuring out how to writing a complex program is not so different from doing a complex surgery. Understanding algorithms is not so different from producing a good differential diagnosis. Most importantly, if we want to solve big problems we can't always rely on an engineer from the outside looking in and solving it for us.
Programming forces you to explain your understanding to something that is very stupid (in a way). If you don't understand what you're doing, your program doesn't work. That basic check doesn't exist for doctors. Ask your doctor what a graph of y = 1/x looks like. Then ask him to graph alveolar CO2 vs ventilation. You should be really afraid of how many doctors only know the latter. I would guess some of them even wouldn't know that! (but they probably know what to do with a patient who's hyperventilating, because they'd memorized it)
Programming is the means to the end, not the end. All this talk of APIs and IDEs and C vs Python vs Javascript really seem to miss the bigger picture here. The best way we have of teaching thinking is with programming (IMHO) and there has to be a way of teaching programming by getting all of the annoying crap out of the way. And we desperately do need more people who have the capacity to think and solve problems. Maybe not will.i.am or Chris Bosh or even software junkies in the valley, but definitely in every other industry.
EDIT: small grammar edit, no content change.