> I think CRISPR-Cas9 seems like the most exciting technology probably in all of science.
I tend to agree. However, having spoken at length about this with a friend doing her PhD, there are a few major problems between now and then.
- Scope: if you thought Big Data (relating to human behavior) was a massive endeavor (still very much not solved, not by a long shot), try genetics. We're talking orders of magnitude Bigger Data. We have the PoC but finding practical solutions remains a hard problem as we speak (needle in a haystack).
- Money: Bio-sectors don't pay software engineers enough to compete with the tech sector (almost no one does), and bio-experts are generally not good enough at it. So there's a huge lack in terms of dual [SE skills + domain knowledge] experts for this category of problems. Research funding is massive in big (private) pharma and comparably non-existent in basic research — and for now, CRISPR-Cas9 is mostly the latter.
The first problem (resources) will probably solve itself as time goes by (assuming some Moore's Law continuation however it's done), however the second problem (domain / education politics? Idk how to call it) could virtually be forever — academia and big pharma aren't exactly known for being fast movers or innovators, let alone disruptors. Especially when CRISPR-Cas9 is a direct threat to well-established revenues in the trillions — curing whatever is much less profitable than selling drugs to ease symptoms over a lifetime.
If this sounds like somptuous irony, it's probably because it is.
> We have the PoC but finding practical solutions remains a hard problem as we speak (needle in a haystack).
One of my former lab mates was doing his thesis on some fluorescent cancers screening stuff, somewhat similar. In his presentations, he'd use a slide explaining the order of magnitude issues in finding cancer cells this way. To illustrate this he'd explain it wasn't like finding a needle i a haystack, but more like trying to find a 20-gauge needle in a Walmart filled with 19-gauge needles.
>Bio-sectors don't pay software engineers enough to compete with the tech sector (almost no one does), and bio-experts are generally not good enough at it.
I've a fair amount of programming (enough to be really dangerous), so other students would come to me with help every once in a while. One of my friends getting his PhD in neuroscience traded a case of beer for an afternoon in helping him. He was doing some vision research with gerbil and was trying to time neuron spikes with some images on a screen. By the 32nd nested 'if statement', I requested another case of beer.
Generally, research-grade programming and software is, at best, spaghetti. At worst, you get answers that you think are right, but are wildly off. You can really lie to yourself, and the rest of the world, when you publish those errors as facts. Most grad students are learning programming by the seat of their threadbare pants, and it shows.
I tend to agree. However, having spoken at length about this with a friend doing her PhD, there are a few major problems between now and then.
- Scope: if you thought Big Data (relating to human behavior) was a massive endeavor (still very much not solved, not by a long shot), try genetics. We're talking orders of magnitude Bigger Data. We have the PoC but finding practical solutions remains a hard problem as we speak (needle in a haystack).
- Money: Bio-sectors don't pay software engineers enough to compete with the tech sector (almost no one does), and bio-experts are generally not good enough at it. So there's a huge lack in terms of dual [SE skills + domain knowledge] experts for this category of problems. Research funding is massive in big (private) pharma and comparably non-existent in basic research — and for now, CRISPR-Cas9 is mostly the latter.
The first problem (resources) will probably solve itself as time goes by (assuming some Moore's Law continuation however it's done), however the second problem (domain / education politics? Idk how to call it) could virtually be forever — academia and big pharma aren't exactly known for being fast movers or innovators, let alone disruptors. Especially when CRISPR-Cas9 is a direct threat to well-established revenues in the trillions — curing whatever is much less profitable than selling drugs to ease symptoms over a lifetime.
If this sounds like somptuous irony, it's probably because it is.