5 ms·
That’s absurd. Yes, I still test my code. But when I write it, I do it with the lessons I’ve learned over decades, and it’s usually close to working before th
by lolwut987 5y ago
That’s absurd.
Yes, I still test my code.
But when I write it, I do it with the lessons I’ve learned over decades, and it’s usually close to working before the first test comes into the picture.
That’s where we are with vaccine production: we’ve been cranking them out for decades, and the process is so robust that failures at in the six plus nines range of outliers.
- whatshisface 5y agoI think you might be overestimating the success rate of drug discovery and underestimating the complexity of the situation. Starting from 10,000 compounds chosen on the basis of reasonable mechanistic or chemical arguments, one or two working drugs are typically discovered. https://www.researchgate.net/figure/Drug-discovery-and-development-timeline-The-current-drug-approval-pipeline-can-take-15_fig1_308045230 https://www.researchgate.net/figure/Drug-discovery-and-devel...
- lolwut987 5y agoLike the rest of engineering, you very rarely start with a novel target. You’re building things from a library of tested techniques. Vaccines aren’t drugs. They’re a method of delivering material to the immune system for it to do its thing. We do that by finding the thing we want to present, finding a pipeline to extract that, and finding a delivery mechanism that is tested and compatible. We don’t “discover” a new flu vaccine ever my season, we put together this seasons release. That’s why it didn’t take years to design and release a new vaccine for covid: we have the technology already, we just needed to pick the payload and select the rocket best suited for the job.