/sci/ - Science & Math

>>15961566
Because this happened and still hasn't been corrected.

Because they can get away with it no matter how massive the consequences for everyone else.

>>15948917
Yes, that was at Imperial College London. Don't let the institution that supported and stood behind him get away without their shared responsibility in all that mess.

It's endemic.

>>14965633
>>14965673
Pic related for those not familiar with the context.

Yes, we should leave coding to the "real" scientists.

You're a sucker. When the day of the rope comes, those academics are going to make themselves scarce while you take all the blame.

Rapidly increasing hardware speed and capabilities have allowed software to become incredibly inefficient and of poor quality. Hardware increases flattening will force software to become efficient again, which will have a good side effect of increasing their quality too. There's an incredible amount of efficiency to be gain in software design and implementation but we've been rapidly going in the wrong direction due to hardware covering for bad software. Now that will be forced to change.

>>14672795

>>14644144
On the other hand, the ease with which these calculations can be performed can lead to very poor design and bad inputs. That in turn can lead to inaccurate results that are blindly accepted. When the costs of the calculations were higher, much more thought and analysis were put into what was actually being calculated and what were the proper inputs.
Beside pic related, there was the wave of austerity that flooded over western government policies a decade ago. These policies were based on an Excel spreadsheet with a simple error in it that showed the opposite of what it would have if the formulas were correct. Hundreds of billions in public money was misdirected and entire economies steered in a bad direction because of this error that would have quickly been caught using a more manual process involving large numbers of people.

>>14597068
>how scary covid could've been.
Why were we so wrong?

>>14591008
Next should be public executions, especially for the bozos at ICL whose negligent reckless faulty analysis was used to justify all of this excess. Death, in public, with international coverage, should be the first step on the road to changing how the world responds to this type of thing. Scientists should be allowed to make mistakes but they shouldn't be allowed to make reckless mistakes they don't try to correct. A painful gruesome death with billions watching is what's needed to get scientists off their pedestals of arrogance and back into the real world of those who are supporting them and their resources.

Reminder that John Carmack gave credibility to the ICL model that sent the world into a panicked lockdown. We now know that model was complete shit and the code was shit multiplied by bloody vomit. While he doesn't have the knowledge to validate a pandemic model, he has the expertise to know that the code was a complete dumpster fire. He knowingly misrepresented this in order to help scare the world into lockdown.